WO2002084960A9 - Data transport acceleration and management within a network communication system - Google Patents
Data transport acceleration and management within a network communication systemInfo
- Publication number
- WO2002084960A9 WO2002084960A9 PCT/US2002/011779 US0211779W WO02084960A9 WO 2002084960 A9 WO2002084960 A9 WO 2002084960A9 US 0211779 W US0211779 W US 0211779W WO 02084960 A9 WO02084960 A9 WO 02084960A9
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- ofthe
- congestion window
- round trip
- trip time
- timer
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/0001—Systems modifying transmission characteristics according to link quality, e.g. power backoff
- H04L1/0015—Systems modifying transmission characteristics according to link quality, e.g. power backoff characterised by the adaptation strategy
- H04L1/0017—Systems modifying transmission characteristics according to link quality, e.g. power backoff characterised by the adaptation strategy where the mode-switching is based on Quality of Service requirement
- H04L1/0018—Systems modifying transmission characteristics according to link quality, e.g. power backoff characterised by the adaptation strategy where the mode-switching is based on Quality of Service requirement based on latency requirement
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/0001—Systems modifying transmission characteristics according to link quality, e.g. power backoff
- H04L1/0002—Systems modifying transmission characteristics according to link quality, e.g. power backoff by adapting the transmission rate
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/12—Arrangements for detecting or preventing errors in the information received by using return channel
- H04L1/16—Arrangements for detecting or preventing errors in the information received by using return channel in which the return channel carries supervisory signals, e.g. repetition request signals
- H04L1/18—Automatic repetition systems, e.g. Van Duuren systems
- H04L1/1809—Selective-repeat protocols
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/12—Arrangements for detecting or preventing errors in the information received by using return channel
- H04L1/16—Arrangements for detecting or preventing errors in the information received by using return channel in which the return channel carries supervisory signals, e.g. repetition request signals
- H04L1/18—Automatic repetition systems, e.g. Van Duuren systems
- H04L1/1867—Arrangements specially adapted for the transmitter end
- H04L1/1874—Buffer management
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/12—Arrangements for detecting or preventing errors in the information received by using return channel
- H04L1/16—Arrangements for detecting or preventing errors in the information received by using return channel in which the return channel carries supervisory signals, e.g. repetition request signals
- H04L1/18—Automatic repetition systems, e.g. Van Duuren systems
- H04L1/1867—Arrangements specially adapted for the transmitter end
- H04L1/188—Time-out mechanisms
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/12—Arrangements for detecting or preventing errors in the information received by using return channel
- H04L1/16—Arrangements for detecting or preventing errors in the information received by using return channel in which the return channel carries supervisory signals, e.g. repetition request signals
- H04L1/18—Automatic repetition systems, e.g. Van Duuren systems
- H04L1/1867—Arrangements specially adapted for the transmitter end
- H04L1/1887—Scheduling and prioritising arrangements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/19—Flow control; Congestion control at layers above the network layer
- H04L47/193—Flow control; Congestion control at layers above the network layer at the transport layer, e.g. TCP related
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/20—Traffic policing
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/28—Flow control; Congestion control in relation to timing considerations
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/28—Flow control; Congestion control in relation to timing considerations
- H04L47/283—Flow control; Congestion control in relation to timing considerations in response to processing delays, e.g. caused by jitter or round trip time [RTT]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/37—Slow start
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/16—Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
- H04L69/161—Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/16—Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
- H04L69/161—Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields
- H04L69/162—Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields involving adaptations of sockets based mechanisms
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/16—Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
- H04L69/163—In-band adaptation of TCP data exchange; In-band control procedures
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/40—Network security protocols
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/43—Assembling or disassembling of packets, e.g. segmentation and reassembly [SAR]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/16—Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/50—Reducing energy consumption in communication networks in wire-line communication networks, e.g. low power modes or reduced link rate
Definitions
- the present invention generally relates to data communication protocols, and more particularly, to systems and methods for data transport acceleration and management within a network communication system.
- Transport Control Protocol has become a common end-to-end data transport protocol used in modern data communication networks.
- Communication networks employing a TCP architecture offer significant advantages in terms of connectivity by enabling applications and users deployed on different physical networks to communicate with one another using a common communications protocol.
- the recent increase in the number and diversity of applications, users and networking environments utilizing TCP architectures has exposed many ofthe limitations associated with a single, ubiquitous design. Because these architectures were primarily intended to provide reliable, sequenced transmission of non-real-time data streams over relatively high bandwidth wireline channels, these TCP architectures tend to exhibit sub-optimal performance when employed in applications or networking environments having different or incompatible characteristics.
- Typical TCP flow control mechanisms utilize an acknowledgement-based approach to control the number and timing of new packets transmitted over the communication network.
- a sender maintains a congestion window parameter that specifies the maximum number of unacknowledged packets that may be transmitted to the receiver.
- the congestion control mechanism increases the size ofthe congestion window (and decreases the number of unacknowledged packets), thereby enabling the flow control mechanism to immediately transmit additional packets to the receiver.
- the network employs symmetric communication channels that enable data packets and acknowledgements to be equally spaced in time.
- communication networks such as wireless communication networks, that employ asymmetric uplink and downlink channels, where the available bandwidth towards the receiver is significantly higher than the available bandwidth towards the sender, the receiver may be unable to access the uplink channel in order to transmit acknowledgement signals to the sender in a timely manner.
- This initial delay in the transmission of acknowledgement signals may cause the sender to suspend transmission of additional data packets until additional acknowledgement signals are received, and then transmit a large burst of packets in response to the sender receiving a large group of acknowledgement signals.
- This bursty nature of data transmission may under-utilize the available bandwidth on the downlink channel, and may cause some applications requiring a steady flow of data, such as audio or video, to experience unusually poor performance.
- the congestion control and error recovery mechanisms typically employed in TCP architectures may also cause the communication network to exhibit sub-optimal performance.
- the congestion control and error recovery mechanisms are used to adjust the size of the congestion window (and therefore the number of new packets that may be transmitted to the receiver) based on the current state ofthe congestion control and error recovery algorithm.
- the sender In the initial "slow start" state, for example, the sender rapidly probes for bandwidth by increasing the size ofthe congestion window by one for each new acknowledgement received from the receiver until the congestion window exceeds a particular congestion window threshold. Once the congestion window exceeds the congestion window threshold, the algorithm enters a "congestion avoidance" state, where the congestion window is increased by one whenever a number of acknowledgment signals equal to the size ofthe current congestion window is received.
- the algorithm enters a "fast retransmit” state in which the sender decreases the congestion window to a size equal to one half of the current congestion window plus three, and retransmits the lost packet.
- the algorithm enters a temporary "fast recovery” state that increments the congestion window by one for each duplicate acknowledgement received from the receiver. If an acknowledgement for the lost packet is received before a retransmit timeout occurs (which is typically based on the average and mean deviation of round-trip time samples), the algorithm transitions to the "congestion avoidance" state.
- the sender if an acknowledgement for the lost packet is not received before a retransmit timeout occurs, the sender resets the congestion window to one, retransmits the lost packet and transitions to the "slow start" state.
- the problem with the foregoing approach is that the congestion avoidance and error recovery mechanisms assumes that packet loss within the communication network was caused by congestion, rather than a temporary degradation in the signal quality ofthe communication channel. Although this assumption may work adequately for many wireline communication networks that have a relatively low occurrence of random packet loss, random packet loss due to fading, temporary degradation in signal quality, signal handoffs or large propagation delays occur with relatively high frequency in most wireless and other bandwidth constrained networks.
- TCP performance may also be adversely impacted by signal handoffs and fades that typically occur in wireless networks. Handoffs and fades can cause multiple data packet losses, which can lead to failure of TCP's fast recovery mechanism and result in prolonged timeouts. If the handoff or fade lasts for several round trip times, failure of multiple retransmission attempts may cause exponential backoff of data throughput. This may result in long recovery times that last significantly longer than the originating fades or handoffs, and may cause TCP connections to stall for extended periods of time.
- TCP congestion control mechanisms tend to exhibit sub-optimal performance
- the congestion control mechanism aggressively increases the size ofthe congestion window until it senses a data packet loss. This process may adversely impact other connections that share the same reduced- bandwidth channel as the connection being initialized attempts to maximize its data throughput without regard ofthe other pre-existing connections.
- many applications such as web browsers, often open multiple TCP connections between a given sender and a given receiver. Under conventional TCP architectures, these connections operate independently and may compete with one another for the same bandwidth, even though they serve the same application. This may lead to inefficient use of resources with decreased overall throughput as connections initialized at later times disrupt connections already in progress.
- Embodiments ofthe present invention provide systems and method for data transport acceleration and management within a network communication system.
- a transmit timer incorporated within the sender is utilized to provide timer-based data flow control, instead of the acknowledgement-based flow control utilized in conventional TCP architectures.
- the transmit time may be used by the sender to determine the rate at which data packets are transmitted to the receiver.
- the period ofthe transmit timer may be based on smoothed round trip time and smoothed congestion window measurements to provide a better estimate of available bandwidth.
- the period of the timer may be determined by the ratio of the smoothed round trip time and the smoothed congestion window and may be periodically updated to reflect the current state of the communication channel.
- a short timer may be used to enable faster error recovery if the duplicate acknowledgment or selective acknowledgment mechanisms for loss detection provided by typical TCP implementations fail.
- the short timer may be used to provide a dial tone that maintains a connection active for a predetermined period of time during traffic congestion while the network is probed to detect whether transmission conditions improve.
- the exponential backoff of the timeout timer that causes stalls in conventional TCP implementations may also be disabled, and a long timer may be used in conjunction with the short timer to provide a more robust response to random packet loss.
- the short timer and the long timer may be set in response to detection of a lost packet. Expiration ofthe short timer may be configured to cause the sender to retransmit the lost packet. If the long timer expires before an acknowledgement for the retransmitted packet is received, the sender may then reduce the congestion window (and the corresponding transmission rate) and transition to a slow start state.
- the exponential backoff ofthe timeout timer may also be delayed by, for example, disabling the exponential backoff for a predetermined number of expirations ofthe long timer, thereby reducing stalls due to transient phenomenon and protecting against the possibility of network congestion collapse.
- These approaches may significantly reduce the vulnerability to random packet loss by causing the sender to reduce the congestion window in response to a second packet loss within a predetermined time period, instead of the first packet loss as used in conventional TCP implementations.
- Another embodiment ofthe present invention provides for early detection of network traffic congestion by comparing a current smoothed estimate ofthe round trip time with a previous value ofthe smoothed estimate ofthe round trip time, such as the value ofthe smoothed estimate ofthe round trip time substantially one round-trip time earlier.
- the size ofthe congestion window may reduced, which reduces the transmission rate ofthe transmit timer. If the change in the next smoothed round trip time estimates again exceeds the predetermined threshold, the congestion window may be reduced by a greater level than the initial reduction.
- another embodiment manages access to the shared channel by each connection based on a set of weights corresponding to the plurality of connections and on host-level statistics. For example, a comparison between a then-current smoothed estimate of a round trip time for a particular connection and a minimum value ofthe smoothed estimate ofthe round trip time for the plurality of other connections may be used to adjust the rate of change of the congestion window for that particular connection in a slow start state.
- the congestion window for a particular connection is updated during a congestion avoidance state based on a four-state model that employs host-level statistics.
- the statistics may include a congestion window for that particular connection, a smoothed estimate ofthe round trip time for that particular connection, or a change in the sum ofthe congestion windows for a plurality of connections.
- the available bandwidth among the active connections may then be allocated based on a set of host-level variables, including a sum ofthe congestion windows ofthe individual connections and a minimum value ofthe smoothed estimate ofthe round trip time for all connections.
- a TCP communication session transitions from a slow start state to a congestion avoidance state based on a comparison between a then-current smoothed estimate ofthe round trip time and a minimum value ofthe smoothed estimate ofthe round trip time recorded over a prior period of time.
- the minimum value ofthe smoothed estimate ofthe round trip time is determined over all the connections sharing the channel.
- the then-current smoothed estimate ofthe round trip time may then be compared to the minimum value of the smoothed estimate ofthe round trip time recorded over a prior period of time to determine whether to switch from increasing the congestion window at a relatively-fast rate in a slow start state to increasing the congestion window at a more gradual rate in congestion avoidance.
- Figure 1 illustrates an exemplary network communication system in which certain aspects ofthe present invention may be advantageously practiced
- Figure 2 illustrates another exemplary communication system in which certain aspects ofthe present invention may be advantageously practiced
- Figure 3 illustrates an exemplary platform that may be used in accordance with the present invention
- Figure 4 illustrates an exemplary method in flowchart form providing for traffic management of data communications in accordance with an embodiment ofthe present invention
- Figure 5 illustrates a flowchart showing exemplary processes that may take place during an initialization state in accordance with an embodiment of the present invention
- Figure 6A illustrates a flowchart showing exemplary processes that may take place during data transmission in accordance with an embodiment ofthe present invention
- Figure 6B illustrates a flowchart showing exemplary processes that may take place while processing incoming acknowledgment signals in accordance with an embodiment ofthe present invention
- Figure 6C illustrates a process branch that may be executed while processing acknowledgement signals in conjunction with the embodiment of
- Figure 7 illustrates a flowchart showing exemplary processes that may take place during a congestion avoidance state in accordance with an embodiment ofthe present invention
- Figure 8 A illustrates an exemplary state model that may be used to adjust the congestion window or data transmission rate in accordance with an embodiment ofthe present invention
- Figure 8B illustrates a flow diagram showing exemplary state transitions corresponding to the exemplary state model from Figure 8A;
- Figure 8C illustrates an exemplary process that may be employed to detect network traffic congestion in accordance with an embodiment ofthe present invention
- Figure 9 illustrates a flowchart showing exemplary processes that may take place during a fast retransmit state in accordance with an embodiment of the present invention.
- Figure 10 illustrates a flowchart showing exemplary processes that may take place during a fast recovery state in accordance with an embodiment ofthe present invention.
- aspects ofthe present invention provide systems and methods for data transport acceleration and management within a communications network. Either alone or in combination, these systems and methods provide improved data transfer between devices connected via communication networks, including wireless and wireline networks.
- the following description is presented to enable a person skilled in the art to make and use the invention. Descriptions of specific embodiments or applications are provided only as examples. Various modifications, substitutions and variations of embodiments will be apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope ofthe invention. The present invention should therefore not be limited to the described or illustrated embodiments, and should be accorded the widest scope consistent with the principles and features disclosed herein.
- Figure 1 illustrates an exemplary network communication system in which certain aspects ofthe present invention may be advantageously practiced.
- the exemplary system includes a server 102 that communicates with a client 116 via a network 112, such as the Internet, General Packet Radio Service (GPRS) or other wireline or wireless network.
- the server 102 and the client 116 may each comprise a computing device, such as a computer, mobile computer, laptop computer, wireless device, personal digital assistant, or cellular telephone, that are configured to transmit and receive data in accordance with a data communication protocol.
- the terms “server” and “sender” generally represent a computing device that transmits data
- client and “receiver” generally denote a computing device that receives data.
- the functions ofthe server 102 and the client 116 may be interchangeable depending on the direction ofthe data stream.
- the server 102 could function as a client and the client 116 could function as a server in the event the server 102 receives a data stream from the client 116.
- the server 102 and the client 116 may both act as a server and a client with respect to each other during the same communication session in the event the server 102 and the client 116 are performing duplex independent transmissions, such as when the server 102 is streaming multimedia data to the client 116 while the client 116 is transmitting an email message to the server 102.
- additional servers and clients may exist and be interconnected via the network 112, and the server 102 may communicate with one or more additional clients.
- the server 102 includes an IP layer 104, a TCP layer 106 and a server application 107.
- the IP layer 104 provides data-level support for communications between the server 102 and the client 116, including managing the format ofthe data packets and the addressing scheme.
- the TCP layer 106 provides connection-level support for communications between the server 102 and the client 116, including managing the establishment of connections and delivery of data from and to the server application 107.
- the IP layer 104 and the TCP layer 106 cooperate to provide virtual data connections between the server 102 and the client 116.
- the server application 107 may comprise one or more applications, such as a web application, video conferencing application or email application, that rely on the TCP layer 106 and the IP layer 104 to transmit and receive data.
- the client 116 also includes a client IP layer 118, a client TCP layer 120 and a client application 122 that perform functions similar to the ones provided by the IP layer 104, the TCP layer 106 and the server application 107.
- the server 102 and the client 116 communicate via the network 112 using data transmissions configured in accordance with the TCP protocol.
- the IP layer 104 and the TCP layer 106 ofthe server 102 may communicate with the client IP layer 118 and the client TCP layer 120 to establish one or more virtual connections between the server 102 and the client 116 via the network 112. These virtual connections may serve the server application 107 running on the server 102 and the client application 122 running on the client 116.
- the server 102 may transmit data requested by a web browser application running within the client application 122.
- Data transmitted by the server 102 to the client 116 is formatted into data packets by the IP layer 104 and the TCP layer 106 and addressed to the client 116 according to the Internet Protocol ("IP") scheme.
- IP Internet Protocol
- the data packets formatted by the IP layer 104 and the TCP layer 106 are transmitted via the server channel 110 to the network 112.
- the network 112 then routes the data packets through the network 112 to the client 16 via the client channel 114.
- the client IP layer 118 and the client TCP layer 120 disassemble the incoming data packets, extract the appropriate information and transmit appropriate acknowledgement signals back to the server 102.
- TCP implementations incorporated within network communications systems may experience sub-optimal performance depending on the application or networking environment. For example, because the flow control mechanism incorporated within conventional TCP implementations transmits new packets in response to each received acknowledgement (and the size ofthe congestion window), these flow control mechanisms may produce bursty data transmission in the event the networking environment utilizes asymmetric uplink and downlink channels that prevent the client from transmitting acknowledgement signals in a timely manner.
- Conventional TCP implementations may also erroneously assume that random packet loss that is common within wireless networking environments corresponds to congestion loss, rather than to a mere temporary degradation in the signal quality ofthe wireless channel.
- This erroneous detection of congestion loss may cause the server 102 to substantially reduce the size ofthe congestion window and the corresponding number of unacknowledged packets the server may transmit to the client.
- the error recovery mechanisms may also be susceptible to signal handoffs and deep fades in that the unacknowledged packets may cause an exponential increase in the retransmit timer and a corresponding decrease in the data throughput, which in many cases may represent overly aggressive responses to the actual problems.
- the aggressive initialization typically employed within TCP implementation during slow start may cause new connections to substantially interfere in the existing connections.
- Embodiments ofthe present invention alleviate many ofthe foregoing problems by enhancing the operation of TCP without severing end-to-end TCP syntax.
- a transmit timer 108 that may be used to provide timer-based flow control of data transmitted by the server 102.
- embodiments of the present invention transmit a packet in response to each expiration ofthe transmit timer 108.
- the period of the timer may be determined by the ratio ofthe smoothed round trip time and the smoothed congestion window and may be periodically updated as a predetermined number of new measurements are taken.
- This timer-based flow control and the smoothing used to compute the period ofthe transmit timer 108 may reduce bursty transmissions and provide a better measurement of network congestion.
- the transmission rate determined in accordance with this aspect of the invention may be applied to both new and retransmitted packets.
- Another embodiment ofthe present invention reduces the adverse effects associated with random packet loss by resetting the congestion window in response to the second packet loss (rather than the first) to provide improved robustness to random packet loss.
- the initialization criterion used to determine whether to transition from aggressive growth ofthe congestion window in slow start to less aggressive growth in congestion avoidance may be based on the smoothed round trip time (instead ofthe size ofthe congestion window) in order to provide a better estimation of network congestion.
- the congestion windows and initialization criteria for all active connections may be linked together and allocated among the active connections so as to avoid destructive competition among existing connections. As a result, these aspects ofthe invention may provide significant improvements in performance over conventional TCP implementations.
- FIG. 2 illustrates another exemplary communication system in which certain aspects ofthe present invention may be advantageously practiced.
- a service module 228 is deployed between a server 202 and a client 218 to manage TCP data transmissions.
- the embodiment of Figure 2 utilizes the service module 228 to intercept and regulate data transmission between the server 208 and the client 218.
- the service module 228 may include a classifier 234 that classifies packets communicated between the server 202 and the client 218 to determine whether the connection corresponds to a service application 242 supported by the service module 228.
- the service module 228 breaks the connection between the server 202 and the client 218 to form a server side connection 208 between the service application 242 and the sever 202 and a client side connection 210 between the service application 242 and the client 218. Because this process forms two separate connections that pass through the TCP layer 236 and the IP layer 232 ofthe service module 228, embodiments ofthe present invention may be configured to provide transport acceleration and data management over both the server side connection 208 and the client side connection 210, thereby enhancing end-to-end performance.
- an exemplary platform that may be used in accordance with the present invention is illustrates generally at 300.
- the exemplary platform includes a network interface card 302, a microprocessor 304, a memory system 308 and an interrupt controller 303, which are all operationally coupled to a system bus 306.
- the network interface card 302 interfaces via one or more data ports with other systems or nodes of a communication network, such as a network router, gateway or base transceiver station.
- the interrupt controller 303 is coupled to the processor 304 via one or more communication lines.
- the memory system 308 which may comprise a random access memory, hard drive, floppy disk, compact disk, or another type of computer readable medium, stores various modules that control the functionality ofthe exemplary platform, such as applications 310 and an operating system and networking stack 312.
- the operating system and networking stack 312 comprises a TCP layer 314 and an IP layer 318.
- the TCP layer 314 comprises a number of features that enhance TCP performance according to various aspects ofthe present invention, including a transmit timer 316 which controls the rate at which data packets are transmitted by the platform 300.
- the network interface card 302 when the network interface card 302 receives a data packet, the network interface card 302 generates an interrupt command and communicates it to the interrupt controller 303. The interrupt controller 303 then passes the interrupt command to the processor 304 in accordance with the priority assigned to that particular interrupt.
- the processor 304 executes one or more interrupt handlers incorporated within the operating system and networking stack 312 to process the incoming data packet.
- Embodiments ofthe present invention may also incorporate other functionalities within the operating system and networking stack 312, such as functionalities for classifying a connection, breaking a connection between a wireless client and a server and swapping source addresses for outgoing packets.
- the operating system and networking stack 312 may also interact with the applications 310 to provide various services, such as web browser applications, video conferencing or email applications.
- FIG. 4 illustrates an exemplary method in flowchart form for providing traffic management of data communications in accordance with an embodiment ofthe present invention.
- the exemplary traffic management process illustrated in Figure 4 may be deployed in various configurations or devices that participate in, or regulate data communications.
- the traffic management process 400 may be deployed within the TCP layer 106 of the server 102 in the embodiment of Figure 1, or within the TCP layer 236 of the service module 228 in the embodiment of Figure 2.
- the connection goes through an initial setup process where the client and server exchange handshake control messages that include synchronization and acknowledgement signals.
- the client and server can transmit data packets to each other.
- data transport control is managed through a series of states that guide the traffic management process depending on various trigger events that occur within the system, such as arrival of acknowledgement (ACK) or selective acknowledgement (SACK) signals, or the expiration of a transmit timer, short timer or long timer.
- trigger events such as arrival of acknowledgement (ACK) or selective acknowledgement (SACK) signals, or the expiration of a transmit timer, short timer or long timer.
- specific actions or transitions between the states ofthe traffic management process take place in response to such trigger events.
- the exemplary traffic management process includes an idle state 402, an initialization state 404, a slow start state 406, a congestion avoidance state 410, a fast retransmit state 414 and a fast recovery state 418.
- Various aspects ofthe present invention are deployed within these states to provide enhanced data traffic management, including, for example, a transmit timer 408 included within the slow start state 406 that controls the transmission rate of data packets and enables more efficient management of data transmission.
- the traffic management process 400 leaves the idle state 402 and enters the initialization state 404.
- the initialization state 404 a number of variables relevant to traffic management are initialized, including, for example, a congestion window
- the traffic management process enters the slow start state 406.
- the congestion window is increased relatively fast while attempting to anticipate traffic congestion and transmission rate limitations.
- Availability of data transmission resources is inferred from various conditions, including based on measurements of various characteristics ofthe network, such as a round trip time.
- the traffic management process 400 enters the congestion avoidance state 410.
- the congestion window is altered more gradually based on a four-state process in an attempt to efficiently utilize the communication channel while minimizing or avoiding losses.
- the congestion window is correspondingly increased, which results in a decrease ofthe period of a transmit timer according to an aspect ofthe present invention.
- a decrease in the period ofthe transmit timer may translate in an increase ofthe transmission rate. Consequently, upon detection of an increase in the available bandwidth, the transmission rate may be increased to ensure efficient utilization of network resources.
- the congestion window may be decreased, thereby increasing the period of the transmit timer and curtailing the transmission rate to avoid overloading the transmission channel.
- a long timeout trigger may lead the traffic management process 400 to return to the slow start state 406, where the congestion window is reset to its initial value. If only limited packet loss is detected via a predetermined number of duplicate acknowledgement signals, selective acknowledgement signals or short timeouts, however, the traffic management process enters the fast retransmit state 414. In the fast retransmit state 414, data packets presumed lost are retransmitted. Once data packets are retransmitted in the fast retransmit state 414, the traffic management process 400 enters the fast recovery state 418.
- FIG. 5 illustrates a flowchart showing exemplary processes that may take place during an initialization state in accordance with an embodiment of the present invention.
- An initialization state 504 illustrated in Figure 5 may correspond to the initialization state 404 in the embodiment of Figure 4.
- a number of variables and constants relevant to traffic management are initialized as further described below.
- an initialization state 504 is located between the idle state 502 and the slow start state 506.
- a number of processes are executed during the initialization state 504 to configure various variables that may be used subsequently to manage the data traffic. These processes include initialization of a congestion window (snd_cwnd) (508), of a congestion window threshold (snd_ssthresh) (510), and of other variables (520) and constants (530) as further described below.
- snd_cwnd a congestion window threshold
- 520 constants
- Each of these initialization processes may be performed independently, either before, simultaneously with, or subsequent to other initialization processes. For example, initialization ofthe congestion window
- the congestion window threshold (510) may take place before, at the same time with, or after initialization ofthe congestion window threshold (510).
- the corresponding variables or constants are stored in memory for subsequent use during the traffic management process that controls data communications in connection with various embodiments ofthe present invention.
- the congestion window (snd_cwnd) is used to initially determine the period of a transmit timer and implicitly the data packet transmission rate.
- This congestion window is initialized in step 508 to a particular value determined by a configurable parameter.
- the congestion window may be initialized to a size of 2 or 3 data packets.
- the congestion window for a new connection between a server and a client established over a channel shared by pre-existing connections between the same server and client may be initialized based on the minimum smoothed estimate of round trip time determined over all pre-existing connections.
- the congestion window is initialized based on historical information regarding previous data communications over a particular data channel.
- the congestion window may be initialized based on the identity of a subscriber that originates or receives a data transmission.
- the identity of particular subscribers may be determined based on Internet Protocol (IP) addresses or based on identity information provided as part ofthe data received from the subscriber.
- IP Internet Protocol
- subscribers may be allocated priority levels, and corresponding congestion windows may be initialized to provide transmission rates and quality of service proportional to their respective priority levels. For example, a subscriber with a higher priority may be allocated a larger initial congestion window corresponding to a higher initial data rate.
- the congestion window may be initialized based on the type of data being transmitted. For example, a connection a for multimedia data streaming may receive a larger initial congestion window as compared to a connection established to transmit an email message to account for the larger data rate requirement ofthe multimedia connection.
- a congestion window threshold variable (snd_ssthresh) that controls the behavior of an embodiment ofthe present invention during the slow start state is initialized in step 510.
- the congestion window threshold is initialized to zero to disable an initial comparison performed in step 624 ofthe exemplary process illustrated in Figure 6B.
- Additional variables associated with various aspects ofthe present invention are initialized in step 520, including a minimum smoothed estimate of round trip time (min_srtt) value.
- min_srtt minimum smoothed estimate of round trip time
- the minimum smoothed estimate round trip time for a new connection being established over that communication channel may be set equal to the minimum value ofthe smoothed estimate round trip time measured for existing connections. If there are no pre-existing connections, the minimum smoothed estimate round trip time may be initialized to zero.
- a number of constants that may be employed at different stages ofthe traffic data management process according to various embodiments ofthe invention are initialized in step 530.
- These constants include, for example, a threshold value that may be used during a slow start state to decide whether to enter a fast retransmit state ("sysctl_dupack"), a constant “al” used by an embodiment ofthe invention operating in a slow start state to decide whether to remain in that state under various circumstances, a constant "a2” that may be employed to detect traffic congestion, a constant "a3” that may be used to determine how aggressively to probe during a congestion control state based on the number of packages acknowledged by incoming messages, a constant "a4" that may be used to determine a smoothed congestion window according to a recurrent formula and a constant "a5" that may help determine the extent to which the transmission rate is decreased during the fast retransmit state.
- a threshold value that may be used during a slow start state to decide whether to enter a fast retransmit state
- Figure 6A illustrates a flowchart showing exemplary processes that may take place during data transmission in accordance with an embodiment ofthe present invention.
- data packet transmission or retransmission takes place at intervals determined by a transmit timer. Transmission or retransmission of data packets may occur while a traffic management process progresses through various states as illustrated by various embodiments ofthe invention, including, for example, the slow start, congestion avoidance, fast retransmit or fast recovery states illustrated in the embodiment of Figure 4. Certain conditions defined within some of these states may supersede the transmit timer under specific circumstances and may prevent transmission or retransmission of particular data packets. For example, if an embodiment ofthe invention determines that the network transmission resources are overloaded, transmission of additional data packets may be suspended until the congestion in the network subsides.
- a transmit timer provides a mechanism for controlling the data transmission rate by adjusting the period ofthe transmit timer based on modification ofthe congestion window and smoothed round trip time measurements.
- a short timer that may be used to detect loss of individual data packets.
- a long timer that may be used to detect more severe losses of multiple data packets.
- the data transmission rate at which data packets are sent is controlled by the transmit timer with a period determined based on the ratio between a then-current value of a smoothed estimate of a round trip time and a then-current value of a smoothed congestion window.
- the smoothed estimate ofthe round trip time between a server and a client represents a measure ofthe time that messages or control signals take to propagate between the two devices.
- the smoothed estimate ofthe round trip time (srtt) at a particular time "t" is determined according to the following formula:
- measured_rtt corresponds to the round trip time ("rtt") measured between the server and the client.
- the round trip time may be measured based on the time that passes between transmission of a particular packet and receipt of an acknowledgment from the client corresponding to that packet.
- the smoothed estimate of the round trip time (srtt) may be determined based on the average and mean deviation of round-trip time samples previously measured by the system and may be periodically updated to include recent measurements. In one case, the round trip time is determined based only on a predetermined number of more recent measurements. In another case, the round trip time may be determined based on all previous available measurements.
- the smoothed estimate ofthe round trip time provides a more accurate indication ofthe actual propagation delay of data travelling between the server and the client as compared to the round trip time.
- Ki and K2 are configurable parameters that may be set to different values depending on the characteristics ofthe network, server and client. In one embodiment, K ⁇ is set to 7/8 and K2 is set to 1/8.
- the smoothed congestion window represents a weighted measure ofthe number of data packets that the server is prepared to transmit to the client and is determined based on the then-current congestion window. The congestion window estimates the delay-bandwidth product ofthe connection to the receiver.
- the congestion window is determined based on acknowledgement signals provided by the client, and/or on information regarding the condition ofthe network inferred by the server through other methods (e.g., based on a timer, or based on measurements of network characteristics such as round trip time).
- the smoothed congestion window is periodically updated using the following formula:
- smoothed_cwnd [t] a 4 * smoothed_cwnd[t-l] + (l-a4) * snd_cwnd,
- smoothed_cwnd [t] represents the smoothed congestion window at a particular moment in time generically denoted as t
- snd_cwnd represents the then current congestion window
- a 4 represents a constant that may have been initialized in step 530 of initialization state 504 from the embodiment of Figure 5.
- a server transmits data packets to a client at intervals determined in accordance with equation (3).
- data packets are transmitted at a rate determined based on the ratio between a smoothed estimate ofthe round trip time and a smoothed congestion window.
- this transmission rate is employed for transmission of all data packets, including retransmissions of data packets lost or delayed in the communication channel or network.
- this transmission rate is employed only for transmission of selected data packets.
- equations (2) or (3) may be modified to include additional terms or scaling factors.
- the transmission rate may be determined using different formulas for different data packets, depending on conditions existing in the network or along a communication channel.
- An advantage of employing a data transmission rate determined in accordance with this aspect ofthe present invention is that bursts in data transmissions may be decreased or eliminated.
- this aspect ofthe present invention may help avoid situations where bursts of data are transmitted within short periods of time with an erroneous assumption that the communication channel or receiving device may accommodate the corresponding data surge. This situation may occur under typical TCP implementations, for example, when signals acknowledging receipt of data packets are sent by a client connected to a high-speed network through a slow channel. In that case, acknowledgement signals that are initially spaced out in time by the slower channel may be compressed in time while propagating through the faster network, and may reach the server within a relatively short period of time.
- acknowledgement signals that are initially spaced out in time become queued up along the transmission path due to traffic congestion and are eventually compressed in time.
- a server utilizing typical TCP may react by sending a burst of data without regard of the actual transmission resources in the network, thereby inducing or exacerbating traffic congestion or data loss.
- data packets are transmitted at a particular transmission rate based on a transmit timer, thereby avoiding sudden variations in the data rate. This may help alleviate or redress performance limitations that may otherwise be encountered by a server and client communicating via TCP under circumstances similar with the ones described above.
- Another embodiment ofthe present invention provides a short timer that may be used to detect loss of data packets.
- a short timer may be set when a packet is transmitted or retransmitted. Expiration ofthe short timer leads to a short timeout state. This provides an additional mechanism (in addition to typical TCP duplicate acknowledgments, selective acknowledgment and timeout) for detecting packet loss.
- the short timer may be important when selective acknowledgments (SACKs) that can explicitly indicate packet loss are not available, and may be disabled if selective acknowledgments are available.
- SACKs selective acknowledgments
- the short timer may also be used to maintain connections active during extended disturbances occurring along the communication channel, such as deep fades, by preventing the traffic management process from reacting prematurely to an apparent termination of the connection.
- an embodiment ofthe invention may maintain a connection with a client over a communication channel that experiences an extended breakdown (e.g., a prolonged state of congestion or a deep fade) without terminating the connection to the client.
- an extended breakdown e.g., a prolonged state of congestion or a deep fade
- this may be achieved because certain actions and transitions between states (e.g., termination of a connection) occur in response to certain events, such as timeouts or receipt of acknowledgment signals.
- an embodiment ofthe invention may be configured to respond to expiration of a short timer by re-transmitting a particular data packet at a particular data rate rather than decreasing the data rate or waiting for receipt of an acknowledgment signal.
- the timeout interval ofthe short timer may be defined based on the then-current smoothed estimate ofthe round trip time and the maximum value ofthe smoothed estimate ofthe round trip time recorded since the previous timeout.
- the period (srto) ofthe short timer may be determined based on the following formula:
- srtt is the then-current smoothed estimate ofthe round trip time.
- mdev corresponds to a running average ofthe deviation ofthe measured value of the round trip time from the average value of the round trip time, and may be computed as follows:
- mdev[t] 3 A * mdev[t-l] + l A *
- An aspect ofthe present invention may also provide a long timer that may be used to detect extended periods of silence from the client or more severe data loss, including, for example, loss of multiple data packets.
- the long timer may be reset upon transmission or retransmission of a data packet or upon receipt of an acknowledgment signal. Expiration ofthe long timer leads to a long timeout state and induces the traffic management process to transition into a slow start state from a congestion avoidance state, a fast retransmit state, or a fast recovery state.
- the timeout interval ofthe long timer may be defined based on the then-current smoothed estimate ofthe round trip time and the maximum variation ofthe round trip time. For example, the expiry period ofthe long timer may be determined using the following formula:
- rto max(4 * srtt, 2 * srtt + 4 * max_deviation_in_rtt),
- max_deviation_in_rtt corresponds to the largest value of mdev determined over an arbitrary period of time. In one case, max_deviation_in_rtt is the largest value of mdev recorded since the initialization of that particular connection.
- initialization state 602 data transmission may begin.
- transmission of data packets takes place at a transmission rate determined based on a transmit timer. Transmission of data may occur simultaneously with progression ofthe traffic management process through various states like slow start, congestion avoidance, fast retransmit and fast recovery as illustrated in Figure 4. While in each of these states, the traffic management process may modify various parameters (e.g., the congestion window) in response to different conditions present in the network (e.g., traffic congestion).
- Modification of such parameters may result in adjustment of the period ofthe transmit timer, and consequently ofthe transmission rate.
- a data transmission process 604 commences upon expiration of a transmit timer.
- the expiry period ofthe transmit timer is determined using equation (3) from above.
- the traffic management process advances to step 606.
- a decision is made at step 606 whether any packets need to be retransmitted. Packets may need to be retransmitted, for example, if they were previously declared lost by the system, as determined by receipt a predetermined number of duplicate acknowledgements, expiration ofthe short timer, etc. If packets do need to be retransmitted, the first packet in the retransmit queue is transmitted at step 608.
- step 606 If the decision at step 606 indicates that no retransmission of packets is necessary, the traffic management process progresses to step 610 where it transmits a new packet if permitted by the sender advertised window (snd_adv_wnd), which is included in each acknowledgement and indicates the size ofthe clients buffer.
- Sender advertised window (snd_adv_wnd)
- An aspect ofthe present invention provides a sender advertised window that helps determine the amount of data that the server may send and assists with recovery from data errors.
- the sender advertised window is further described below in connection with the embodiment of Figure 9.
- the transmit timer controls both retransmission of lost data packets in step 608 and transmission of new data packets in step 610 enables synchronization ofthe transmit and retransmit processes to effectively manage data traffic. Since, as previously discussed, the period ofthe transmit timer is adjusted dynamically to ensure efficient utilization of available network transmission capacity, and therefore data packets are transmitted only when bandwidth is anticipated to be available, the embodiment of Figure 6A effectively multiplexes retransmitted and new data packets through the transmission channel. As a result, this embodiment ofthe invention achieves a fair queuing mechanism that efficiently allocates bandwidth for both transmission and retransmission of data to help optimize overall throughput.
- the traffic management process proceeds to step 612 where it sets both a short timer and a long timer as previously discussed.
- the short timer detects loss of individual data packets, while the long timer identifies more severe losses involving multiple data packets.
- the traffic management process then enters step 613 where it determines the next expiration period for the transmit timer, which indicates when the next data packet will be transmitted.
- the period ofthe transmit timer depends on the ratio ofthe smoothed round trip time to the smoothed congestion window, as was described above in connection with equation (3).
- the system increments the then-current time recorded by the timer by the value T obtained from equation (3).
- Step 614 is a default state that recurs throughout various phases ofthe traffic management process and provides a phase in which the system awaits one ofthe aforementioned triggering events to determine the subsequent evolution ofthe system. If the transmit timer set in step 613 expires before the short timer or the long timer, then the traffic management process returns to step 606 to attempt to transmit or retransmit a data packet.
- the system If the short timer set in step 612 expires first, however, the system assumes that a data packet has been lost and progresses to step 618 to retransmit the packet.
- the system retransmits the first data packet in the transmission sequence that has not been acknowledged as received by the client.
- the sequence number for this data packet is denoted in step 618 as snd_una.
- the client has acknowledged all data packets in transmission sequence through and including snd_una - 1.
- the system may retransmit a different data packet than snd_una, possibly one ofthe data packets that were sent to the client after the snd_una packet.
- the system Upon retransmission of a data packet in step 618, the system resets both the short and the long timers in step 619.
- the traffic management process enters an optional step 615, where it disables or delays a TCP exponential backoff mechanism that typically exists in TCP.
- Typical TCP utilizes a single timer to identify prolonged periods of non- responsiveness from the client, possibly indicating network congestion or disturbances along the communication channel.
- expiration ofthe timer induces a timeout state that progressively increases subsequent periods ofthe timer exponentially (typically as a power of 2). Consequently, in typical TCP implementations, sustained lack of response from the client may result in successive timeouts with exponentially-increasing delays.
- an aspect ofthe present invention may avoid such exponential increase in the period of successive timeouts and associated disadvantages. More specifically, the period ofthe long timer may be controlled based on a smoothed estimate of the round trip time as described above in connection with the embodiment of Figure 6A, and may therefore be maintained within a more desirable range.
- the long timer may be made longer than the retransmit timeout ("RTO") value that exists in typical TCP implementations, thus reducing the occurrence of spurious timeouts.
- this may be achieved by making the period ofthe short timer smaller than the period ofthe long timer such that the short timer tends to expire before the long timer, and by resetting the long timer whenever the short timer is reset (e.g., upon transmission or retransmission of data packets). Consequently, in this embodiment, as long as transmission or retransmission of packets is permitted by the sender advertised window, the short timer essentially disables the operation ofthe long timer.
- the long timer is only relevant in this embodiment if a corresponding sub-process is triggered by arrival of an acknowledgment signal since the short timer is not reset by such an event. In that case, if transmission or retransmission of data packets is suspended for a sufficiently-long period of time, the long timer is operational and may expire before the short timer. From optional step 615, the traffic management process advances to step 616 where it transmits the next packet expected by the client, i.e., the packet with sequence number snd_una. The long timer is then reset in step 617.
- step 614 If an acknowledgment signal is received while the traffic management process is expecting triggering events in step 614, the traffic management process advances to branch point "A," from where it enters step 620 ofthe flow chart illustrated in Figure 6B.
- Figure 6B illustrates a flowchart showing exemplary processes that may take place while processing incoming acknowledgment signals in accordance with an embodiment ofthe present invention.
- a decision is made whether the acknowledgment signal received in step 614 ofthe embodiment of Figure 6 A is a duplicate acknowledgment.
- Duplicate acknowledgments are acknowledgments that correspond to the same data packet and may indicate that a preceding data packet was not received, possibly because it is lost or because data packages were reordered by the network.
- Duplicate acknowledgments may be generated when the client expects a data packet with a specific sequence number but receives a data packet with a higher sequence number. In that case, for each out-of-sequence packet received, the client generates an acknowledgment signal that identifies the data packet originally expected. In an alternative implementation, the client generates an acknowledgment signal only upon receipt of multiple out-of-sequence packets.
- step 620 If the decision at step 620 is that the incoming acknowledgment signal is not a duplicate acknowledgment, the system progresses to step 622 where it records the number of data packets that were successfully delivered to the receiver by increasing a corresponding variable ack_free_cnt.
- the variable ack_free_cnt maintains a count ofthe number of buffers freed by incoming acknowledgment signals. Determination ofthe number of data packets successfully delivered to the receiver may be made based on various criteria. In one embodiment, if successive incoming acknowledgments correspond to non- consecutive data packets in the data stream, the number of packets successfully delivered to the receiver is inferred in step 622 by assuming that all data packages numbered between those corresponding packages will be eventually received.
- acknowledgement signals may be interpreted to indicate that all preceding unacknowledged data packets will be successfully delivered, and the number of such unacknowledged packets is used to increment the variable ack_free_cnt in step 622. For example, if the server receives acknowledgments for data packets with sequence numbers up to and including packet 152, and if the next acknowledgment corresponds to packet 155, the server assumes that packets 153 and 154 will be eventually received and increments ack_free_cnt by two.
- each acknowledgment signal is interpreted to indicate successful delivery of a single corresponding data packet in step 622, and the variable ack_free_cnt is incremented by one for each acknowledgement signal received.
- the traffic management process then progresses to step 624, where a decision regarding the congestion state ofthe network is made.
- the decision process in step 624 comprises two conditions.
- the first condition assesses whether the congestion window snd_cwnd is smaller or equal to the congestion window threshold snd_ssthresh:
- This condition may employ past history regarding the data channel, the identity of a subscriber or the nature ofthe data being transmitted to determine how far the congestion window should be relatively-rapidly increased.
- the congestion window threshold may have been previously initialized to zero in step 510 of initialization state 504 from Figure 5. Consequently, in that case, this condition in step 624 would be initially inoperative, and the second condition would control the original decision in step 624.
- the congestion window threshold may be updated to a non-zero value subsequently, and the first condition would then become operative.
- the second condition in step 624 compares the then-current smoothed estimate ofthe round trip time (srtt) with a weighted value ofthe minimum recorded smoothed estimate ofthe round trip time min_srtt:
- This condition is particularly important when the first condition is inoperative.
- the first condition in step 624 may also be inoperative if past history relevant to the condition is missing (e.g., relevant information regarding past history was not stored or was improperly retrieved) or is unreliable (e.g., after a timeout).
- the minimum recorded value of the smoothed estimate ofthe round trip time (min_srtt) represents the minimum smoothed estimate ofthe round trip time that was recorded during previous measurements and is used as a measure ofthe propagation delay. This value is shared among multiple connections to a single host that share a communication channel, and may be employed to avoid destructive competition for bandwidth among these connections.
- weight ai assigned to the minimum smoothed estimate ofthe round trip time in step 624 is the constant used to decide whether to remain in the slow start state that was initialized in step 530 of initialization state 504 from Figure 5.
- the second condition in step 624 may be broadly interpreted to permit rapid growth ofthe congestion window until the queuing delay is at least (ai -1) times the propagation delay. This interpretation is more accurate in typical networks that have buffering ofthe order ofthe bandwidth- delay product.
- the value of ai is equal to 1.2.
- step 624 If both conditions in step 624 evaluate as true (indicating that growth of the congestion window may continue at a relatively fast rate), the system branches into step 625 ofthe slow start state, where it increases the size ofthe congestion window based on the value ofthe variable ack_free_cnt that was previously updated in step 622.
- the congestion window is increased based on the number of data packets or data segments that are acknowledged as received by a client.
- the congestion window, and indirectly the data transmission rate are adjusted based on the number of data packets acknowledged by acknowledgement signals sent by a client.
- Various methods for acknowledging data packets using acknowledgement signals were previously discussed in connection with updating the variable ack_free_cnt in step 622.
- the number of data segments or data packets effectively acknowledged by an acknowledgement signal sent by a client device is weighted and is used thereafter to increase the congestion window. For example, if an incoming acknowledgement signal communicates that the client device has received a total of five data packets, the congestion window may be increased by five data packets. Alternatively, the congestion window may be increased by a different number of packets if a weighing function is used to weigh the number of packets acknowledged. In the latter example, for a particular weighing function, if an incoming acknowledgement signal communicates that the client device has received five data packets, the congestion window may be increased by three or by nine data packets, depending on the characteristics and form of the weighing function.
- An advantage of this aspect ofthe present invention is that the data transmission rate does not decrease unnecessarily under certain circumstances, including, for example, when channel communication resources exist but acknowledgement signals are lost.
- the server may increase the congestion window by a smaller amount in response to the fewer incoming acknowledgement signals. This may happen under conventional TCP, for example, where the congestion window may be increased by one data packet or data segment for each acknowledgment signal received. This may result in a correspondingly-smaller increase in the data transmission rate, which is undesirable if the data channel has available bandwidth that could accommodate a comparatively-higher transmission rate.
- An aspect ofthe present invention may address and potentially resolve this situation by increasing the congestion window, and implicitly the transmission rate, by a relatively-larger amount in response to fewer incoming acknowledgement signals.
- An embodiment ofthe invention employs the process for updating the variable ack_free_cnt that was discussed above in conjunction with step 622 to increase the congestion window in step 625 of Figure 6B according to the following formula:
- a 3 is a constant that may be used to determine how aggressively to probe for bandwidth availability while performing congestion control.
- the variable a 3 may have been initialized at step 530 ofthe initialization state 502 of Figure 5. From step 625, the traffic management process advances to step 626 where it updates variable ack_free_cnt using the following formula:
- ack_free_cnt (ack_free_cnt) mod (a 3 ),
- step 624 If both ofthe conditions in step 624 evaluate as false, the system then branches to step 628 of a congestion avoidance state where it updates the variable ack_free_cnt substantially similarly as previously described in connection with step 626.
- the traffic management process then invokes a four- state model for congestion avoidance. This state model is further described in connection with Figures 8A, 8B and 8C. If the decision in step 620 ofthe embodiment of Figure 6B indicates that the acknowledgment signal that prompted the traffic management process to leave step 614 ofthe embodiment from Figure 6 A was a duplicate acknowledgment signal, this embodiment assumes that a data packet has been lost. Consequently, the system enters a fast retransmit state by progressing to step 662 ofthe flow process illustrated in Figure 6C via branch point "B."
- Figure 6C illustrates a process branch that may be executed while processing acknowledgement signals in conjunction with the embodiment of Figure 6B.
- Branch point "B" from Figure 6B leads the traffic management process to step 662 in Figure 6C.
- the system estimates the number of data packets that have been delivered to the receiver out of order and stores this number into a variable out_of_order_cnt. In one embodiment ofthe present invention, this number is determined based on information provided by a client via selective acknowledgment (SACK) messages. If SACK information is not available, the number of data packets that have been delivered to the receiver out of order may be estimated based on the number of duplicate acknowledgment signals received by the sender.
- SACK selective acknowledgment
- step 664 the number of data packets that have been delivered to the receiver out of order estimated in step 662 is compared to a threshold variable systctl_dupack that indicates when the system should assume that data packages that have not been delivered within a certain period of time are lost.
- the threshold systctl_dupack is initialized in step 520 ofthe initialization state 504 from Figure 5.
- the threshold systctl_dupack has a value of 3. If the comparison in step 664 indicates that the number of data packets estimated in step 662 to have been delivered to the receiver out of order exceeds the threshold value, the system enters a fast retransmit state in step 666.
- FIG. 7 illustrates a flowchart showing exemplary processes that may take place during a congestion avoidance state in accordance with an embodiment ofthe present invention.
- the system may enter congestion avoidance from step 624 in Figure 6B.
- the congestion avoidance state 704 the transmission rate is altered gradually to utilize efficiently available bandwidth resources while minimizing or avoiding losses.
- Data packets are transmitted while the transmission rates for individual connections sharing the communication channel are independently and dynamically adjusted to improve global data throughput.
- the congestion window is updated in response to variations in network capacity using a state diagram and taking into account the weights of various connections sharing the communication channel. Probing for bandwidth availability also takes into account weights of individual connections to avoid destructive competition for channel capacity.
- the traffic management process enters the congestion avoidance state 704 from the slow start state 702.
- the system performs traffic management for various independent connections sharing the communication channel.
- these independent connection are suggestively denoted as Connection #1, . . ., Connection # N, where the variable N generically represents an arbitrary number of connections.
- a number of other independent connections are not shown in Figure 7, but may comprise steps substantially similar with the steps illustrated for Connection # 1 and Connection # N.
- traffic management for the various individual connections may take place selectively and in any relative order.
- connection-specific congestion window threshold variables for each individual connection in a manner that helps achieve a fair queuing mechanism providing weight-based transmission rate allocation for the individual connections.
- the connection-specific congestion window threshold variables determined in steps 710 and 720 provide Quality of Service support for the traffic management process by making window dynamics during congestion avoidance a function of individual connection weights.
- each connection-level congestion window threshold variable is determined based on the following formula:
- CW THRESH snd__cwnd * average_weight / connection_weight ,
- CW'S ⁇ RESH represents the connection-level congestion window threshold for a particular individual connection generically identified as K
- snd_cwnd ⁇ is the congestion window for that particular connection
- average_weight represents a measure ofthe average ofthe weights of active connections
- connection_weight ⁇ is the actual weight of that particular individual connection.
- the system then enters step 712 where a decision is made regarding the capacity ofthe network at that particular time.
- a variable ack_free_cnt is maintained for each individual connection to indicate how many data packets have been successfully delivered to the receiver, and therefore by how much the network capacity may have increased. In one embodiment, this variable is the variable ack_free_cnt that was described in conjunction with step 622 et seq.
- connection # 1 is allowed to transmit a certain number of data packets, which depends on its weight, and the condition in step 712 evaluates whether Connection # 1 is permitted to use network resources that may have become available, or if another connection with a higher weight should receive priority.
- the expression for the connection-level congestion window threshold provided by equation (4) permits individual connections with higher weights to change their congestion windows more frequently because their thresholds are smaller. This mechanism provides priority access to the communication channel to connections with higher weights.
- step 714 the congestion window is updated based on a state model that will be further described in conjunction with the embodiments of Figures 8A, 8B and 8C.
- the state model evaluates network resources in step 714 and adjusts the congestion window for Connection # 1.
- the system then enters step 716 where it resets the variable ack_free_cnt ⁇ that tracks the number of data packets have been successfully delivered to the receiver.
- step 712 If the condition in step 712 evaluates as true, therefore indicating that Connection # 1 has sufficient priority to transmit data, the system permits Connection # 1 to transmit a data packet and progresses to step 718 where it increases the variable ack_free_cnt ! to account for additional data packets that may have successfully reached the receiver since ack_free_cnt! was last updated.
- Connection # 1 also applies in a substantially similar manner to other individual connections, including Connection #N illustrated in the embodiment of Figure 7.
- Connection #N illustrated in the embodiment of Figure 7.
- appropriate reference modifications would have to be made to refer to corresponding steps and variables associated with such other individual connections.
- references to the variable ack_free_cnt ⁇ would be instead to variable ack_free_cntN-
- Figure 8A illustrates an exemplary state model that may be used to adjust the congestion window or data transmission rate in accordance with an embodiment ofthe present invention.
- the state model illustrated in Figure 8 A may be employed to update the congestion window while the traffic management process is operating in a congestion avoidance state, including, for example, in steps 714 or 724 ofthe congestion avoidance state 704 from the embodiment of Figure 7.
- a traffic management process may assume that certain measurements of network conditions are accurately and promptly reflected in variables maintained within the corresponding system. Under certain circumstances, however, this assumption may be incorrect. For example, in step 624 of slow start state 604 from Figure 6B, a decision is made based on the size ofthe congestion window and measurements ofthe round trip time.
- the state model illustrated in Figure 8 A may be used to update the congestion window while allowing sufficient settling time for accurate measurements of round trip time.
- Figure 8 A are accompanied by increases or decreases in the size ofthe congestion window maintained by the traffic management process, suggestively indicated as "increase” or "decrease.”
- Figure 8B illustrates a flow diagram showing exemplary state transitions corresponding to the exemplary state model from Figure 8A.
- the progression of the traffic management process through the states of Figure 8 A may also be described in reference to the state flow diagram of Figure 8B.
- the four states illustrated in Figure 8B may correspond to the four states from Figure 8 A as follows: state CONGAVOID_GOOD (810) in Figure 8B may correspond to state 1 (802) in Figure 8A; state CONGAVOID_BAD (814) in Figure 8B may correspond to state 2 (804) in Figure 8 A; state
- CONGAVOID_REDUCED (824) in Figure 8B may correspond to state -1 (806) in Figure 8A; and state CONGAVOID_POST_REDUCE (828) in Figure 8B may correspond to state 0 (808) in Figure 8 A.
- the traffic management process may begin initially in state CONGAVOID_GOOD (810). The management process then progresses to step
- Determination of traffic conditions in step 812 is made by running an external process that employs a number of network or channel capacity measurements to assess availability of transmission resources. In the embodiment of Figure 8 A, this external process identifies network traffic congestion. The functionality and operation ofthe external process used to determine traffic congestion in step 812 is further described below in connection with the embodiment of Figure 8C.
- step 812 If no congestion is detected in step 812, the traffic management process increases the congestion window in step 816 and returns to state
- CONGAVOID_GOOD 810.
- the congestion window is increased or decreased by one data packet at a time. In another embodiment, the congestion window may be increased by more than one packet.
- the traffic management process enters state CONGAVOID_BAD (814). The traffic management process then proceeds to state 820 where another decision regarding the existence of congestion in the network is made using a process substantially similar with the process previously described in conjunction with step 812. If no congestion is detected in step 820, then the traffic management process returns to state CONGAVOID_GOOD (810). Otherwise, if congestion is detected in step 820, the traffic management process decreases the congestion window in step 822 and enters state CONGAVOID_REDUCED (824).
- State CONGAVOID_REDUCED (824) is a transient state that immediately follows a reduction in the size ofthe congestion window. Once in this state, the traffic management process attempts to increase the data transmission rate by increasing the congestion window in step 826, and therefore decreasing the period ofthe transmit timer. This may be done to maximize throughput and efficiently utilize network resources. From step 826, the traffic management process enters state CONGAVOID_POST_REDUCE (828), which is an intermediate state followed by another probing for network traffic conditions in step 830. The decision made in step 830 regarding the existence of network traffic congestion relies on a process substantially similar with the processes used for similar purposes in steps 812 and 820.
- step 830 the traffic management process returns to state CONGA VOID B AD (814) in anticipation of a possible need to decrease the congestion window. If no congestion is determined in step 830, the traffic management process assumes that additional network transmission capacity exists, and increases the congestion window in step 832 to take advantage of this additional capacity. The traffic management process then returns to state CONGAVOID_GOOD (810).
- Figure 8C illustrates an exemplary process that may be employed to detect network traffic congestion in accordance with an embodiment ofthe present invention.
- the process illustrated in Figure 8C may represent the external process employed in steps 812, 820 or 830 ofthe embodiment of Figure 8B to determine network traffic congestion.
- determination of network traffic congestion is made using host-level statistics that take into account various individual connections that share a communication channel, therefore preventing destructive competition for channel resources among these individual connections.
- An advantage of this embodiment is that the global data throughput across individual connections is increased.
- determination of network traffic congestion is made based on a number of traffic-related measurements, including the round trip time (which provides a measure ofthe propagation delay within the network) and the congestion windows for various active connections.
- determination of network congestion is made in two stages. In a first stage, a decision whether variable sum_cwnd_change is less than zero is made in step 840.
- the variable sum_cwnd_change represents the change in the sum ofthe congestion windows for all active connections and provides a measure ofthe total amount of data sent by all connections and ofthe transmission rate over the channel shared by these active connections.
- step 840 If the condition in step 840 is true, therefore indicating that the sum of all congestion windows has not increased, an embodiment ofthe invention declares in step 842 that no congestion has been detected.
- One reason for declaring that no congestion exists when the sum of the congestion windows decreases is that such a decrease suggests that additional traffic capacity may have become available in the network, and therefore probing for network congestion may not be necessary.
- the traffic management process proceeds to a second decision stage in step 850, where another evaluation regarding the existence of congestion in the network is made based on the following equation:
- srtt represents the smoothed estimate ofthe round trip time.
- the smoothed estimate ofthe round trip time may be determined substantially as previously described in connection with equation (1) and the embodiment of Figure 6 A.
- Variable srtt_change represents the change in the smoothed estimate ofthe round trip time since the last update. In another embodiment, the change in the smoothed estimate ofthe round trip time may be determined over a longer period of time.
- Scaling constant a 2 represents a constant used in congestion detection that may have been initialized in step 530 of initialization state 504 with a value between 0 and 1. In a particular embodiment, when the network exhibits a significant amount of traffic imbalance, variable a 2 is assigned a numerical value of 1.
- Variable snd_cwnd represents the congestion window for a particular connection under consideration.
- FIG 9 illustrates a flowchart showing exemplary processes that may take place during a fast retransmit state in accordance with an embodiment of the present invention.
- Fast retransmit state 910 illustrated in Figure 9 may correspond to the fast retransmit state 414 in the embodiment of Figure 4.
- the traffic management process may enter the fast retransmit state 910 in step 666 ofthe process flow illustrated in Figure 6C.
- the fast retransmit state 910 data packets presumed to be lost along the transmission channel are retransmitted and the data transmission rate is adjusted gradually depending on channel resource availability. Further, a "sender advertised window"
- the traffic management process When first entering fast retransmit, the highest data packet sequence number sent thus far is stored in a variable high_seq. The system remains in fast retransmit state 910 and continues to retransmit lost packets until the transmitted packet sequence index number snd_una exceeds the value of high_seq, which indicates that the client has received all the packets that were resent.
- the traffic management process Upon entering the fast retransmit state 910, the traffic management process encounters decision step 920 that determines whether the fast retransmit state 910 is being entered for the first time in the current cycle. This condition essentially estimates the duration for which a corresponding connection between a server and a client has been operating in the fast retransmit state.
- step 922 the number of data packets estimated to have been delivered to the receiver out of order, which is stored in a variable out_of_order_cnt, is compared to a threshold variable sysctl dupack that indicates when the system should assume that data packages that have not been delivered within a certain period of time are lost.
- the decision in step 922 may be substantially similar with the decision made in step 664 from Figure 6C, which was previously described in conjunction with the embodiment of Figure 6C.
- step 924 the traffic management process enters step 924 where the highest sequence new packet number snd_nxt is stored in variable high_seq and the then-current size ofthe congestion window snd_cwnd is stored in the congestion window threshold snd_ssthresh.
- the sender advertised window is then increased in step 930.
- the sender advertised window (“snd_adv_wnd") indicates the amount of data that the sender may transmit.
- the sender advertised window may be used to match the amount of data transmitted by the server to the network transmission capacity to maximize throughput and efficiently recover from data errors.
- the sender advertised window is used to limit the highest data packet sequence number that may be sent by the server to a value of
- snd_una represents the sequence number ofthe next expected packet as indicated by the acknowledgment signals received from the client (i.e., all packets up to snd_una-l are known to have been received).
- packets_in_flight represents an estimate ofthe number of data packets that have been sent by the server but have not yet reached the receiver
- sysctl_dup_ac represents a threshold variable which may be configured to determine when the system should assume that data packages that have not been delivered within a certain period of time are lost
- no active conn is the number of data connections to a single server or client sharing the communication channel
- snd_cwnd represents the congestion window for a particular connection.
- the traffic management process then advances to step 928 where it retransmits a data packet that is assumed to be lost. Retransmission of data packets in step 928 is performed using a transmit timer substantially as described in conjunction with the embodiment of Figure 6A.
- the traffic management then sets a short timer and a long timer at step 930. Short and long timers have been previously described in connection with the embodiment of Figure 6A. If the decision in step 922 evaluates as false, the system proceeds to step 932 where it resets a long timer. The system then enters state 934 where it waits for arrival of an acknowledgment signal or for expiration of the transmit timer, short timer, or long timer.
- step 910 The preceding discussion relative to fast retransmit state 910 has described a series of steps that are executed when the decision in step 920 determines that fast retransmit state 920 was entered for the first time in the present cycle. If fast retransmit state 920 was not entered for the first time, however, step 940 is executed. In step 940, a decision is made whether snd_una is larger than high_seq. If so, then all packets were resent have been successfully received at the client, and the system leaves fast retransmit state 910 via branch point "C.” Otherwise, if the decision in step 940 indicates that not all data packets have been acknowledged as received, the system proceeds to step 942.
- step 942 a decision is made whether more than two packets have been lost and need to be retransmitted. If so, then the congestion window snd_cwnd is reduced in step 944 from the value it had upon entering the fast retransmit state 910 as follows:
- a 5 is a constant that determines the extent to which the transmission rate is decreased during the fast retransmit state and which may have been initialized in step 530 ofthe initialization phase 504 from Figure 5.
- the system then proceeds to step 946. If the decision in step 942 evaluates as false, then the system skips step
- step 946 the system updates relevant variables and transmits a data packet if appropriate.
- transmission of a data packet at step 946 is not based on a transmit timer as previously described in connection with various aspects ofthe present invention, but instead relies on substantially-typical TCP logic. If a data packet was retransmitted in step 948, the system sets a short timer and a long timer in step 950. Short timers and long timers have been previously described in connection with the embodiment of Figure 6A.
- early congestion detection permits use of a less drastic cutback in the data transmission rate upon detection of data loss than in typical TCP implementations.
- Such early congestion detection may be achieved through use of various aspects of the present invention disclosed herein. Consequently, when a data loss is detected, it is likely that the data loss is not due to traffic congestion, but it is instead a result of other phenomena affecting the network or the communication channel, like, for example, random signal fades in wireless transmissions.
- One embodiment ofthe invention takes advantage of this fact by reducing the congestion window, and therefore indirectly the transmission rate, only when a second data packet loss is detected. By not reacting upon detection of a first data loss, this embodiment avoids premature reaction to apparent, but possibly non-existent, network congestion. This is in contrast to TCP, which generally reacts upon detection of losses by immediately reducing the congestion window and transmission rate. This mechanism of delayed reaction to data losses may be implemented with the assistance ofthe embodiment of Figure 9. If the system determines that the packet that was presumed lost and was previously retransmitted was not the first lost packet, then the congestion window is decreased. This process ensures that the congestion window is only decreased if a second data loss is detected, but not before.
- the decrease ofthe congestion window may occur only upon detection of a higher number of data losses than two.
- the number of data losses that induce a decrease in the congestion window, and correspondingly transmission rate may be defined to be higher.
- the data transmission rate may be adjusted independently ofthe size ofthe corresponding sender advertised window to separate congestion control from flow control within the traffic management process.
- flow control may be performed by adjusting a congestion window to permit more aggressive probing of communication channel resources, or by using a timer to control the transmission rate to provide robustness to loss of control signals, as previously discussed. Since adjustment ofthe transmission rate may be performed based on the size ofthe congestion window and on the round trip time between the client and the server, as described in conjunction with the embodiment of Figure 6A, congestion control depends primarily on congestion window and round trip time in related implementations.
- error recovery which is a component of flow control, depends on the size ofthe advertised window.
- the sender advertised window is at least as large as the congestion window.
- the size ofthe sender advertised window may be increased further while the congestion window may be decreased.
- flow congestion may be decoupled from flow control. Consequently, a congestion window or a timer may be used during error recovery to adjust the transmission rate while controlling congestion, while the sender advertised window may be independently adjusted to ensure efficient error recovery.
- the congestion window may be reduced in response to congestion and data loss while the sender advertised window may be increased to permit transmission of more data packets, therefore accelerating error recovery.
- FIG. 10 illustrates a flowchart showing exemplary processes that may take place during a fast recovery state in accordance with an embodiment ofthe present invention.
- Fast recovery state 1004 illustrated in Figure 10 may correspond to fast recovery state 418 in the embodiment of Figure 4.
- the data transmission rate is gradually adjusted depending on channel resource availability.
- Fast recovery state 1004 tends to be a transitory state.
- the traffic management process enters fast recovery state 1004 from fast retransmit state 910 in Figure 9 via branch point "D".
- decision step 1006 a determination is made whether the congestion window was decreased while the traffic management process operated in fast retransmit state 910 of Figure 9.
- step 1010 configures a variable corresponding to a particular state in a state management model that updates the congestion window.
- step 1010 configures a state variable corresponding to the state model previously described in connection with Figures 8A and 8B.
- the traffic management process drives the state model into state CONGAVOID_REDUCED, which corresponds to state 806 in Figure 8A and state 824 in Figure 8B.
- the traffic management process first reduces the congestion window in step 1008 before updating the state variable in step 1010. From step 1010, the traffic management process then proceeds to step 1012, where it decreases the value ofthe congestion window threshold snd_ssthresh. In one embodiment, the congestion window threshold is decreased by a factor of %. The traffic management process then sets a long timer in step 1014. The traffic management process then enters step 1016 where it waits for arrival of an acknowledgment signal or for the expiration of a timer.
Abstract
Description
Claims
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP02733982A EP1393508B1 (en) | 2001-04-12 | 2002-04-11 | Data transport acceleration and management within a network communication system |
DE60214369T DE60214369D1 (en) | 2001-04-12 | 2002-04-11 | DATA TRANSPORT ACCELERATION AND MANAGEMENT IN A NETWORK COMMUNICATION SYSTEM |
Applications Claiming Priority (6)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US28354201P | 2001-04-12 | 2001-04-12 | |
US60/283,542 | 2001-04-12 | ||
US30921201P | 2001-07-31 | 2001-07-31 | |
US60/309,212 | 2001-07-31 | ||
US10/061,574 US7099273B2 (en) | 2001-04-12 | 2002-01-29 | Data transport acceleration and management within a network communication system |
US10/061,574 | 2002-01-29 |
Publications (3)
Publication Number | Publication Date |
---|---|
WO2002084960A2 WO2002084960A2 (en) | 2002-10-24 |
WO2002084960A9 true WO2002084960A9 (en) | 2002-12-12 |
WO2002084960A3 WO2002084960A3 (en) | 2003-12-11 |
Family
ID=27370077
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/US2002/011779 WO2002084960A2 (en) | 2001-04-12 | 2002-04-11 | Data transport acceleration and management within a network communication system |
Country Status (5)
Country | Link |
---|---|
US (1) | US7099273B2 (en) |
EP (1) | EP1393508B1 (en) |
AT (1) | ATE338407T1 (en) |
DE (1) | DE60214369D1 (en) |
WO (1) | WO2002084960A2 (en) |
Families Citing this family (225)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7454500B1 (en) | 2000-09-26 | 2008-11-18 | Foundry Networks, Inc. | Global server load balancing |
US7657629B1 (en) | 2000-09-26 | 2010-02-02 | Foundry Networks, Inc. | Global server load balancing |
US9130954B2 (en) | 2000-09-26 | 2015-09-08 | Brocade Communications Systems, Inc. | Distributed health check for global server load balancing |
US6975591B1 (en) * | 2000-11-22 | 2005-12-13 | International Business Machines Corporation | Methodology for improving TCP throughput over lossy communication links |
DE10102323A1 (en) * | 2001-01-19 | 2002-07-25 | Philips Corp Intellectual Pty | Method and device for the reliable transmission of data packets |
US7016304B2 (en) * | 2001-05-18 | 2006-03-21 | Intel Corporation | Link level retry scheme |
WO2002096021A2 (en) * | 2001-05-18 | 2002-11-28 | Bytemobile, Inc. | Quality of service management for multiple connections within a network communication system |
US20030128672A1 (en) * | 2001-06-19 | 2003-07-10 | Sridhar Komandur | Transmission and flow control |
KR100459557B1 (en) * | 2001-08-23 | 2004-12-03 | 삼성전자주식회사 | Method for allocating hybrid automatic retransmission request channel number for indicating state information of data in high speed downlink packet access communication system |
JP3814505B2 (en) * | 2001-09-19 | 2006-08-30 | 富士通株式会社 | IP network system having provided service control function |
US7299280B2 (en) * | 2001-10-17 | 2007-11-20 | The Regents Of University Of California | Method and apparatus for TCP with faster recovery |
IL146802A0 (en) * | 2001-11-28 | 2003-01-12 | Wave Group Ltd | A self-contained panoramic or spherical imaging device |
US6744730B2 (en) * | 2001-11-30 | 2004-06-01 | Nokia Corporation | Throughput enhancement after interruption |
DE60236138D1 (en) * | 2002-01-03 | 2010-06-10 | Innovative Sonic Ltd | Window based blockage avoidance for a high speed wireless communication system |
US8051197B2 (en) * | 2002-03-29 | 2011-11-01 | Brocade Communications Systems, Inc. | Network congestion management systems and methods |
US8484370B1 (en) * | 2002-04-16 | 2013-07-09 | Trimble Navigation Limited | Method and system for efficient extended data communications using GPRS |
FR2841086B1 (en) * | 2002-06-14 | 2004-08-27 | Nortel Networks Ltd | METHOD FOR CONTROLLING DATA TRANSMISSION AND CONTROL UNIT FOR IMPLEMENTING THE METHOD |
EP1383281A1 (en) * | 2002-07-19 | 2004-01-21 | TELEFONAKTIEBOLAGET LM ERICSSON (publ) | Method for calculating a transmission window size |
US7418494B2 (en) * | 2002-07-25 | 2008-08-26 | Intellectual Ventures Holding 40 Llc | Method and system for background replication of data objects |
US7086061B1 (en) | 2002-08-01 | 2006-08-01 | Foundry Networks, Inc. | Statistical tracking of global server load balancing for selecting the best network address from ordered list of network addresses based on a set of performance metrics |
US7676576B1 (en) | 2002-08-01 | 2010-03-09 | Foundry Networks, Inc. | Method and system to clear counters used for statistical tracking for global server load balancing |
US7574508B1 (en) | 2002-08-07 | 2009-08-11 | Foundry Networks, Inc. | Canonical name (CNAME) handling for global server load balancing |
US7802008B2 (en) * | 2002-08-12 | 2010-09-21 | Matsushita Electric Industrial Co., Ltd. | Quality of service management in network gateways |
DE60232278D1 (en) * | 2002-09-06 | 2009-06-18 | Ericsson Telefon Ab L M | METHOD AND DEVICES FOR REGULATING THE REASON |
US7567509B2 (en) * | 2002-09-13 | 2009-07-28 | Dialogic Corporation | Methods and systems for jitter minimization in streaming media |
KR100912178B1 (en) * | 2002-09-27 | 2009-08-14 | 주식회사 케이티 | Congestion control method over wireless link |
US7069326B1 (en) * | 2002-09-27 | 2006-06-27 | Danger, Inc. | System and method for efficiently managing data transports |
US7428595B2 (en) * | 2002-09-30 | 2008-09-23 | Sharp Laboratories Of America, Inc. | System and method for streaming TCP messages in an enterprise network |
US7961617B2 (en) * | 2002-10-29 | 2011-06-14 | Telefonaktiebolaget Lm Ericsson (Publ) | System and method for wireless network congestion control |
US7616638B2 (en) | 2003-07-29 | 2009-11-10 | Orbital Data Corporation | Wavefront detection and disambiguation of acknowledgments |
US7630305B2 (en) | 2003-07-29 | 2009-12-08 | Orbital Data Corporation | TCP selective acknowledgements for communicating delivered and missed data packets |
US8233392B2 (en) | 2003-07-29 | 2012-07-31 | Citrix Systems, Inc. | Transaction boundary detection for reduction in timeout penalties |
US8270423B2 (en) | 2003-07-29 | 2012-09-18 | Citrix Systems, Inc. | Systems and methods of using packet boundaries for reduction in timeout prevention |
US7168022B2 (en) * | 2002-12-27 | 2007-01-23 | Ntt Docomo, Inc. | Transmission control method and system |
KR100493883B1 (en) * | 2003-01-02 | 2005-06-10 | 삼성전자주식회사 | System and method for managing application |
US7460473B1 (en) * | 2003-02-14 | 2008-12-02 | Istor Networks, Inc. | Network receive interface for high bandwidth hardware-accelerated packet processing |
JP4283589B2 (en) * | 2003-03-25 | 2009-06-24 | 株式会社エヌ・ティ・ティ・ドコモ | COMMUNICATION DEVICE, COMMUNICATION CONTROL METHOD, AND PROGRAM |
US20070115814A1 (en) * | 2003-03-29 | 2007-05-24 | Regents Of The University Of California, The | Method and apparatus for improved data transmission |
US7272672B1 (en) * | 2003-04-01 | 2007-09-18 | Extreme Networks, Inc. | High speed bus with flow control and extended burst enhancements between sender and receiver wherein counter is maintained at sender for free buffer space available |
US7593338B2 (en) * | 2003-06-27 | 2009-09-22 | Samsung Electronics Co., Ltd. | Congestion control method and system for reducing a retransmission timeout count in a transmission control protocol |
US20040266451A1 (en) * | 2003-06-30 | 2004-12-30 | Aleksandr Stolyar | Method and apparatus for improving scheduler performance in wireless packet data systems |
US8238241B2 (en) | 2003-07-29 | 2012-08-07 | Citrix Systems, Inc. | Automatic detection and window virtualization for flow control |
US7656799B2 (en) | 2003-07-29 | 2010-02-02 | Citrix Systems, Inc. | Flow control system architecture |
US8437284B2 (en) | 2003-07-29 | 2013-05-07 | Citrix Systems, Inc. | Systems and methods for additional retransmissions of dropped packets |
US8432800B2 (en) | 2003-07-29 | 2013-04-30 | Citrix Systems, Inc. | Systems and methods for stochastic-based quality of service |
KR100913897B1 (en) | 2003-07-29 | 2009-08-26 | 삼성전자주식회사 | Method for controlling congestion of TCP for reducing the number of retransmission timeout |
US8185703B2 (en) * | 2003-07-31 | 2012-05-22 | Silicon Graphics International Corp. | Detection and control of resource congestion by a number of processors |
US7385923B2 (en) * | 2003-08-14 | 2008-06-10 | International Business Machines Corporation | Method, system and article for improved TCP performance during packet reordering |
ITBA20030039A1 (en) * | 2003-08-29 | 2005-02-28 | Grieco Luigi Alfredo | RATE-BASED CONGESTION CONTROL OF THE ENTERING TRAFFIC |
GB0321425D0 (en) | 2003-09-12 | 2003-10-15 | Ericsson Telefon Ab L M | Radio resource usage optimisation in a packet network |
US9584360B2 (en) | 2003-09-29 | 2017-02-28 | Foundry Networks, Llc | Global server load balancing support for private VIP addresses |
US7321567B2 (en) * | 2003-09-30 | 2008-01-22 | Motorola, Inc. | Method and apparatus for preventing a spurious retransmission after a planned interruption of communications |
US7502322B2 (en) * | 2003-09-30 | 2009-03-10 | Nokia Corporation | System, method and computer program product for increasing throughput in bi-directional communications |
US20070008884A1 (en) * | 2003-10-08 | 2007-01-11 | Bob Tang | Immediate ready implementation of virtually congestion free guarantedd service capable network |
JP2005167353A (en) * | 2003-11-28 | 2005-06-23 | Ntt Docomo Inc | Transmitter device and program |
US7650379B2 (en) * | 2003-12-09 | 2010-01-19 | Viasat, Inc. | Method for channel congestion management |
US7254609B2 (en) * | 2003-12-09 | 2007-08-07 | Viasat, Inc. | Method for channel congestion management |
JP4349114B2 (en) * | 2003-12-10 | 2009-10-21 | ソニー株式会社 | Transmission device and method, reception device and method, recording medium, and program |
KR100999118B1 (en) * | 2003-12-12 | 2010-12-07 | 삼성전자주식회사 | Method for transmitting data in ad-hoc network |
US7609640B2 (en) * | 2003-12-19 | 2009-10-27 | Nokia Corporation | Methods and applications for avoiding slow-start restart in transmission control protocol network communications |
US7310682B2 (en) * | 2004-01-08 | 2007-12-18 | Lsi Corporation | Systems and methods for improving network performance |
JP4594110B2 (en) | 2004-01-23 | 2010-12-08 | 株式会社エヌ・ティ・ティ・ドコモ | Transmitting apparatus and program |
US7397759B2 (en) * | 2004-03-15 | 2008-07-08 | Microsoft Corporation | Response for spurious timeout |
US20070275760A1 (en) * | 2004-05-05 | 2007-11-29 | Telefonaktiebolaget Lm Ericsson (Publ) | Hdspa Flow Control, Control Frames Rtt Measurement |
US8930569B2 (en) * | 2004-05-05 | 2015-01-06 | Qualcomm Incorporated | Methods and apparatus for optimum file transfers in a time-varying network emvironment |
US7496651B1 (en) | 2004-05-06 | 2009-02-24 | Foundry Networks, Inc. | Configurable geographic prefixes for global server load balancing |
US7584301B1 (en) | 2004-05-06 | 2009-09-01 | Foundry Networks, Inc. | Host-level policies for global server load balancing |
US20050266795A1 (en) * | 2004-05-26 | 2005-12-01 | Chengshing Lai | [method of communication using audio/video data] |
JP4068592B2 (en) * | 2004-05-28 | 2008-03-26 | 株式会社東芝 | Wireless communication device |
US8125910B2 (en) * | 2004-06-25 | 2012-02-28 | Nec Corporation | Communication system |
US7796517B2 (en) * | 2004-06-28 | 2010-09-14 | Minghua Chen | Optimization of streaming data throughput in unreliable networks |
US20060015639A1 (en) * | 2004-07-13 | 2006-01-19 | Taylor Benjamin F | Method for managing inter-zone bandwidth in a two-way messaging network |
US20070280107A1 (en) * | 2004-07-23 | 2007-12-06 | Reiner Ludwig | Data Unit Sender Control Method |
US7656800B2 (en) * | 2004-07-30 | 2010-02-02 | Cisco Technology, Inc. | Transmission control protocol (TCP) |
WO2006023604A2 (en) * | 2004-08-17 | 2006-03-02 | California Institute Of Technology | Method and apparatus for network congestion control using queue control and one-way delay measurements |
US7423977B1 (en) | 2004-08-23 | 2008-09-09 | Foundry Networks Inc. | Smoothing algorithm for round trip time (RTT) measurements |
US7975064B2 (en) * | 2004-09-16 | 2011-07-05 | International Business Machines Corporation | Envelope packet architecture for broadband engine |
KR100631757B1 (en) * | 2004-10-02 | 2006-10-09 | 삼성전자주식회사 | Method and apparatus for transmitting wireless data by dynamically changing data rate |
KR100644637B1 (en) * | 2004-10-11 | 2006-11-10 | 삼성전자주식회사 | Apparatus and method for printing data using server message block protocol |
US7760638B2 (en) * | 2004-11-29 | 2010-07-20 | Nec Corporation | High-throughput communication system, communication terminal, session relay, and communication protocol |
US20060140193A1 (en) * | 2004-12-29 | 2006-06-29 | Nokia Corporation | Optimization of a TCP connection |
US7366132B2 (en) * | 2005-03-30 | 2008-04-29 | At&T Corp. | Loss tolerant transmission control protocol |
US20060250949A1 (en) * | 2005-03-30 | 2006-11-09 | At&T Corp. | Loss tolerant transmission control protocol |
US7889654B2 (en) * | 2005-03-30 | 2011-02-15 | At&T Intellectual Property Ii, L.P. | Loss tolerant transmission control protocol |
US7630394B2 (en) * | 2005-04-14 | 2009-12-08 | Ramot At Tel-Aviv University Ltd. | Method, apparatus and computer-readable code for data compression of network packets |
US7782758B2 (en) * | 2005-05-06 | 2010-08-24 | California Institute Of Technology | Efficient loss recovery architecture for loss-decoupled TCP |
US20070097976A1 (en) * | 2005-05-20 | 2007-05-03 | Wood George D | Suspect traffic redirection |
US8423728B1 (en) * | 2005-06-16 | 2013-04-16 | Emc Corporation | Physical based scheduler using run count and skip count values |
JP4859419B2 (en) * | 2005-09-14 | 2012-01-25 | 株式会社エヌ・ティ・ティ・ドコモ | Radio base station and user common data transmission method |
KR100664947B1 (en) * | 2005-09-23 | 2007-01-04 | 삼성전자주식회사 | Method for controlling throughput and communication apparatus using the same |
US20070070894A1 (en) * | 2005-09-26 | 2007-03-29 | Fan Wang | Method to determine a scheduling priority value for a user data connection based on a quality of service requirement |
US7719967B2 (en) * | 2005-09-28 | 2010-05-18 | Netapp, Inc. | Cumulative TCP congestion control |
US20070223395A1 (en) * | 2005-11-23 | 2007-09-27 | Ist International, Inc. | Methods and apparatus for optimizing a TCP session for a wireless network |
US20070223492A1 (en) * | 2005-11-23 | 2007-09-27 | Ist International, Inc. | Methods and apparatus for optimizing a TCP session for a wireless network |
US7760633B2 (en) * | 2005-11-30 | 2010-07-20 | Cisco Technology, Inc. | Transmission control protocol (TCP) congestion control using transmission delay components |
US7664067B2 (en) * | 2005-12-15 | 2010-02-16 | Microsoft Corporation | Preserving socket connections over a wireless network |
GB0600417D0 (en) | 2006-01-10 | 2006-02-15 | Level 5 Networks Inc | Virtualisation support |
AU2007210948A1 (en) * | 2006-02-01 | 2007-08-09 | Bob Tang | Immediate ready implementation of virtually congestion free guaranteed service capable network: external internet NextGenTCP NextGenFTP NextGenUDPs |
US8151323B2 (en) | 2006-04-12 | 2012-04-03 | Citrix Systems, Inc. | Systems and methods for providing levels of access and action control via an SSL VPN appliance |
GB0611249D0 (en) * | 2006-06-07 | 2006-07-19 | Nokia Corp | Communication system |
US7684319B2 (en) * | 2006-06-29 | 2010-03-23 | Intel Corporation | Transmission control protocol congestion window |
US7684332B2 (en) * | 2006-08-22 | 2010-03-23 | Embarq Holdings Company, Llc | System and method for adjusting the window size of a TCP packet through network elements |
US20080056125A1 (en) * | 2006-09-06 | 2008-03-06 | Nokia Corporation | Congestion control in a wireless network |
EP1898546A3 (en) | 2006-09-08 | 2011-04-27 | NTT DoCoMo, Inc. | Communication terminal, comunication control method, and communication control program |
FR2906425B1 (en) * | 2006-09-21 | 2009-10-23 | Ipanema Technologies Sa | METHOD OF OPTIMIZING TRAFFIC CONTROL IN A PACKET TELECOMMUNICATION NETWORK |
WO2008066428A1 (en) * | 2006-11-28 | 2008-06-05 | Telefonaktiebolaget Lm Ericsson (Publ) | A method and a system for down link control in a cellular telephony system |
KR100846344B1 (en) * | 2007-01-05 | 2008-07-15 | 삼성전자주식회사 | Apparatus and method for reducing transmission control protocol latency in portable communication system |
JP4778453B2 (en) * | 2007-01-24 | 2011-09-21 | 株式会社エヌ・ティ・ティ・ドコモ | Communication terminal, congestion control method, and congestion control program |
TWI328377B (en) * | 2007-01-24 | 2010-08-01 | Quanta Comp Inc | Wireless communication system for calculating time of flight |
GB0701668D0 (en) * | 2007-01-29 | 2007-03-07 | Tang Bob | Immediate ready implementation of virtually congestion free guaranteed service capable network: external internet nextgenTCP nextgenFTP nextgenUDPs |
US8631147B2 (en) * | 2007-03-12 | 2014-01-14 | Citrix Systems, Inc. | Systems and methods for configuring policy bank invocations |
US8490148B2 (en) | 2007-03-12 | 2013-07-16 | Citrix Systems, Inc | Systems and methods for managing application security profiles |
US7853679B2 (en) | 2007-03-12 | 2010-12-14 | Citrix Systems, Inc. | Systems and methods for configuring handling of undefined policy events |
US7760642B2 (en) * | 2007-03-12 | 2010-07-20 | Citrix Systems, Inc. | Systems and methods for providing quality of service precedence in TCP congestion control |
US7865589B2 (en) | 2007-03-12 | 2011-01-04 | Citrix Systems, Inc. | Systems and methods for providing structured policy expressions to represent unstructured data in a network appliance |
US7796510B2 (en) | 2007-03-12 | 2010-09-14 | Citrix Systems, Inc. | Systems and methods for providing virtual fair queueing of network traffic |
US7853678B2 (en) | 2007-03-12 | 2010-12-14 | Citrix Systems, Inc. | Systems and methods for configuring flow control of policy expressions |
US8615008B2 (en) | 2007-07-11 | 2013-12-24 | Foundry Networks Llc | Duplicating network traffic through transparent VLAN flooding |
JP2009027720A (en) * | 2007-07-23 | 2009-02-05 | Polycom Inc | System and method executing lost packet recovery with congestion avoidance |
US7912911B2 (en) * | 2007-08-23 | 2011-03-22 | Broadcom Corporation | Method and system for increasing throughput rate by dynamically modifying connection parameters |
US8248928B1 (en) | 2007-10-09 | 2012-08-21 | Foundry Networks, Llc | Monitoring server load balancing |
FR2922391B1 (en) * | 2007-10-15 | 2009-12-04 | Canon Kk | METHOD AND DEVICE FOR DATA TRANSMISSION |
EP2077647A1 (en) | 2008-01-04 | 2009-07-08 | Deutsche Thomson OHG | Method for managing a data connection and network component |
US8819512B1 (en) * | 2008-01-19 | 2014-08-26 | Appex Networks Holding Limited | Method for detecting TCP packet losses and expediting packet retransmission |
WO2009095628A1 (en) * | 2008-02-01 | 2009-08-06 | France Telecom | Method and device for regulating sending in a wireless telecommunication network |
US8837279B2 (en) * | 2008-06-06 | 2014-09-16 | Optis Wireless Technology, Llc | Technique for improving congestion control |
JP5028339B2 (en) * | 2008-06-09 | 2012-09-19 | キヤノン株式会社 | Communication apparatus and control method |
US8422370B2 (en) * | 2008-06-27 | 2013-04-16 | International Business Machines Corporation | Reducing retransmission of out of order packets |
US8149706B2 (en) * | 2008-07-07 | 2012-04-03 | Verizon Patent And Licensing Inc. | Method and system for adjusting bandwidth using multiple timers |
US8094557B2 (en) * | 2008-07-09 | 2012-01-10 | International Business Machines Corporation | Adaptive fast retransmit threshold to make TCP robust to non-congestion events |
WO2010042580A1 (en) * | 2008-10-08 | 2010-04-15 | Citrix Systems, Inc. | Systems and methods for allocating bandwidth by an intermediary for flow control |
US8565242B2 (en) * | 2008-11-04 | 2013-10-22 | Blackberry Limited | Transport protocol performance using network bit rate information |
US8300620B1 (en) * | 2008-12-29 | 2012-10-30 | Sprint Communications Company L.P. | Dynamically tuning a timer mechanism according to radio frequency conditions |
FR2941830A1 (en) * | 2009-01-30 | 2010-08-06 | Ixel | Data sequence's sequential transfer method for global system for mobile communication data network, involves performing transfer of data and following data based on reception of acknowledgement of transferred data |
US7961613B2 (en) * | 2009-02-05 | 2011-06-14 | Silver Spring Networks, Inc. | System and method of monitoring packets in flight for optimizing packet traffic in a network |
US8406133B2 (en) * | 2009-02-24 | 2013-03-26 | Silver Spring Networks, Inc. | System and method of regulating a packet rate to optimize traffic in a network |
WO2010111261A1 (en) | 2009-03-23 | 2010-09-30 | Azuki Systems, Inc. | Method and system for efficient streaming video dynamic rate adaptation |
US8374091B2 (en) * | 2009-03-26 | 2013-02-12 | Empire Technology Development Llc | TCP extension and variants for handling heterogeneous applications |
US8144720B2 (en) * | 2009-04-24 | 2012-03-27 | Telefonaktiebolaget L M Ericsson (Publ) | Uplink radio resource allocation in the presence of power limited users |
US8159939B1 (en) * | 2009-05-08 | 2012-04-17 | Adobe Systems Incorporated | Dynamic network congestion control |
EP2446590B1 (en) | 2009-06-22 | 2015-11-25 | Citrix Systems, Inc. | Systems and methods for platform rate limiting |
JP5573053B2 (en) * | 2009-09-04 | 2014-08-20 | ソニー株式会社 | Wireless communication apparatus and wireless communication method |
US20110090795A1 (en) * | 2009-09-11 | 2011-04-21 | Victor On Kwok Li | Differentiation among occurrences of packet reordering, congestive packet loss, or random packet loss in communication networks |
US8599690B2 (en) * | 2009-12-09 | 2013-12-03 | Electronics And Telecommunications Research Institute | Method and apparatus for fairly allocating resource to network users |
TWI492574B (en) * | 2010-02-05 | 2015-07-11 | Realtek Semiconductor Corp | A method of detecting status of remote device and/or transmission control in communication system |
US8248944B2 (en) * | 2010-03-04 | 2012-08-21 | Microsoft Corporation | Selectively disabling reliability mechanisms on a network connection |
JP5352852B2 (en) * | 2010-03-19 | 2013-11-27 | 株式会社日立製作所 | Mobile communication system and communication method |
US8908524B2 (en) * | 2010-03-30 | 2014-12-09 | Telefonaktiebolaget L M Ericsson (Publ) | Method of congestion detection in a cellular radio system |
US9019854B2 (en) * | 2010-04-26 | 2015-04-28 | Telefonaktiebolaget L M Ericsson (Publ) | Method for setting and adjusting a parameter dependent on a round trip time |
US8427958B2 (en) | 2010-04-30 | 2013-04-23 | Brocade Communications Systems, Inc. | Dynamic latency-based rerouting |
US8458327B1 (en) * | 2010-07-15 | 2013-06-04 | Google Inc. | System and method of reducing network latency |
AU2010361098B2 (en) * | 2010-09-24 | 2014-08-14 | Intel Corporation | Method and system for access point congestion detection and reduction |
US9154394B2 (en) | 2010-09-28 | 2015-10-06 | Brocade Communications Systems, Inc. | Dynamic latency-based rerouting |
US8549148B2 (en) | 2010-10-15 | 2013-10-01 | Brocade Communications Systems, Inc. | Domain name system security extensions (DNSSEC) for global server load balancing |
EP2715978B1 (en) * | 2011-05-24 | 2020-02-19 | Tata Consultancy Services Ltd. | A system and method for reducing the data packet loss employing adaptive transmit queue length |
US9191862B2 (en) * | 2011-09-06 | 2015-11-17 | Qualcomm Incorporated | Method and apparatus for adjusting TCP RTO when transiting zones of high wireless connectivity |
US9591069B2 (en) | 2011-10-31 | 2017-03-07 | Adobe Systems Incorporated | Peer-to-peer assist for live media streaming |
US9515942B2 (en) | 2012-03-15 | 2016-12-06 | Intel Corporation | Method and system for access point congestion detection and reduction |
US9065634B2 (en) * | 2012-04-04 | 2015-06-23 | Qualcomm Incorporated | Method and apparatus for selective acknowledgement of packets from an access probe |
JP5923376B2 (en) * | 2012-04-27 | 2016-05-24 | 株式会社日立製作所 | TCP relay device |
US9705964B2 (en) * | 2012-05-31 | 2017-07-11 | Intel Corporation | Rendering multiple remote graphics applications |
US9264365B2 (en) * | 2012-07-31 | 2016-02-16 | International Business Machines Corporation | Split transport control protocol (TCP) flow control management in a cellular broadband network |
US9246843B2 (en) * | 2012-09-04 | 2016-01-26 | Apple Inc. | Detecting and recovering from a transmission channel change during a streaming media session |
US20140281019A1 (en) * | 2013-03-15 | 2014-09-18 | Microsoft Corporation | Network Transmission Adjustment Based On Application-Provided Transmission Metadata |
CN105191225A (en) * | 2013-03-28 | 2015-12-23 | 株式会社东芝 | Communication device, communication method, and communication program |
BR112015028451A2 (en) * | 2013-05-13 | 2017-07-25 | Huawei Tech Co Ltd | overload process and control device |
US10084834B2 (en) * | 2013-05-27 | 2018-09-25 | Electronics And Telecommunications Research Institute | Randomization of packet size |
US9485186B2 (en) * | 2013-07-23 | 2016-11-01 | Cisco Technology, Inc. | Network congestion control with awareness of random packet losses |
US20150055482A1 (en) * | 2013-08-20 | 2015-02-26 | Brocade Communications Systems, Inc. | TCP Extended Fast Recovery and Segment Timing |
US9565138B2 (en) | 2013-12-20 | 2017-02-07 | Brocade Communications Systems, Inc. | Rule-based network traffic interception and distribution scheme |
US10043137B1 (en) | 2014-01-06 | 2018-08-07 | Nuu:Bit, Inc. | Dynamically optimized transport system |
US9648542B2 (en) | 2014-01-28 | 2017-05-09 | Brocade Communications Systems, Inc. | Session-based packet routing for facilitating analytics |
CN104954279B (en) * | 2014-03-28 | 2018-04-10 | 华为技术有限公司 | A kind of transfer control method, apparatus and system |
CN106688218B (en) * | 2014-04-15 | 2020-07-28 | 飞利浦灯具控股公司 | Method and apparatus for controlling handshaking in a packet transport network |
ES2666426T3 (en) | 2014-04-23 | 2018-05-04 | Bequant S.L. | Method and apparatus for network congestion control based on transmission rate gradients |
US9930097B2 (en) | 2014-07-03 | 2018-03-27 | Qualcomm Incorporated | Transport accelerator systems and methods |
US10178431B2 (en) | 2014-07-28 | 2019-01-08 | Adobe Inc. | Hybrid stream delivery |
CN105471701B (en) * | 2014-08-15 | 2018-12-14 | 中国电信股份有限公司 | The method and agency plant of message time-out duration are set |
CN104168284B (en) * | 2014-08-25 | 2019-02-05 | 联想(北京)有限公司 | A kind of data transmission method and the first electronic equipment |
US20160094608A1 (en) * | 2014-09-30 | 2016-03-31 | Qualcomm Incorporated | Proactive TCP Connection Stall Recovery for HTTP Streaming Content Requests |
US9992126B1 (en) | 2014-11-07 | 2018-06-05 | Speedy Packets, Inc. | Packet coding based network communication |
US9992088B1 (en) | 2014-11-07 | 2018-06-05 | Speedy Packets, Inc. | Packet coding based network communication |
US10999012B2 (en) | 2014-11-07 | 2021-05-04 | Strong Force Iot Portfolio 2016, Llc | Packet coding based network communication |
US9825733B1 (en) | 2014-11-07 | 2017-11-21 | Speedy Packets, Inc. | Packet coding based network communication |
US10320526B1 (en) | 2014-11-07 | 2019-06-11 | Strong Force Iot Portfolio 2016, Llc | Packet coding based network communication |
US10530700B2 (en) * | 2015-07-07 | 2020-01-07 | Strong Force Iot Portfolio 2016, Llc | Message reordering timers |
KR101737786B1 (en) | 2015-01-27 | 2017-05-19 | 서울대학교산학협력단 | Proxy apparatus and method for wireless resource estimation and traffic management |
US9985898B2 (en) | 2015-02-26 | 2018-05-29 | Citrix Systems, Inc. | System for bandwidth optimization with traffic priority determination |
US9826066B2 (en) * | 2015-02-26 | 2017-11-21 | Citrix Systems, Inc. | System for dynamic selection and application of TCP congestion avoidance flavors |
US10911353B2 (en) | 2015-06-17 | 2021-02-02 | Extreme Networks, Inc. | Architecture for a network visibility system |
US9866478B2 (en) | 2015-03-23 | 2018-01-09 | Extreme Networks, Inc. | Techniques for user-defined tagging of traffic in a network visibility system |
US10771475B2 (en) | 2015-03-23 | 2020-09-08 | Extreme Networks, Inc. | Techniques for exchanging control and configuration information in a network visibility system |
US10129088B2 (en) | 2015-06-17 | 2018-11-13 | Extreme Networks, Inc. | Configuration of rules in a network visibility system |
US10834065B1 (en) | 2015-03-31 | 2020-11-10 | F5 Networks, Inc. | Methods for SSL protected NTLM re-authentication and devices thereof |
CN104980257B (en) * | 2015-05-29 | 2020-08-25 | 北京盛特思信息技术有限公司 | Internet of things communication method and device |
US10057126B2 (en) | 2015-06-17 | 2018-08-21 | Extreme Networks, Inc. | Configuration of a network visibility system |
US10530688B2 (en) | 2015-06-17 | 2020-01-07 | Extreme Networks, Inc. | Configuration of load-sharing components of a network visibility router in a network visibility system |
US10158575B2 (en) * | 2015-06-17 | 2018-12-18 | Citrix Systems, Inc. | System for bandwidth optimization with high priority traffic awareness and control |
US10284484B1 (en) * | 2015-07-15 | 2019-05-07 | Cisco Technology, Inc. | Accelerated ramp up for delay-based rate adaptation |
US9979653B2 (en) | 2015-10-21 | 2018-05-22 | Citrix Systems, Inc. | System and method of providing improved throughput control under delay-based congestion situation in a network |
US10601721B2 (en) * | 2015-10-29 | 2020-03-24 | Altiostar Networks, Inc. | Maintenance of downlink throughput |
US20170126845A1 (en) * | 2015-10-29 | 2017-05-04 | VTX Holdings (Singapore) Pte. Ltd. | Network communication system |
US10404698B1 (en) | 2016-01-15 | 2019-09-03 | F5 Networks, Inc. | Methods for adaptive organization of web application access points in webtops and devices thereof |
US10091075B2 (en) | 2016-02-12 | 2018-10-02 | Extreme Networks, Inc. | Traffic deduplication in a visibility network |
US10674313B2 (en) * | 2016-03-17 | 2020-06-02 | Viavi Solutions Uk Limited | Automatic optimization procedure termination using a smoothing-based technique |
US10999200B2 (en) | 2016-03-24 | 2021-05-04 | Extreme Networks, Inc. | Offline, intelligent load balancing of SCTP traffic |
US10097462B2 (en) * | 2016-04-02 | 2018-10-09 | Niciria, Inc. | Throughput resilience during link failover |
US20170295516A1 (en) * | 2016-04-08 | 2017-10-12 | Qualcomm Incorporated | Dynamic medium access control reception-reorder timeout in a crowded wireless local area network |
US11009865B2 (en) | 2016-05-09 | 2021-05-18 | Strong Force Iot Portfolio 2016, Llc | Methods and systems for a noise pattern data marketplace in an industrial internet of things environment |
CN105827537B (en) * | 2016-06-01 | 2018-12-07 | 四川大学 | A kind of congestion improved method based on QUIC agreement |
EP3267639B1 (en) * | 2016-07-06 | 2019-12-25 | Alcatel Lucent | Congestion control within a communication network |
US10567259B2 (en) | 2016-10-19 | 2020-02-18 | Extreme Networks, Inc. | Smart filter generator |
US10511533B2 (en) | 2016-10-28 | 2019-12-17 | Level 3 Communications, Llc | Systems and methods for adjusting a congestion window value of a content delivery network |
CN109151848B (en) * | 2017-06-16 | 2021-07-02 | 创新先进技术有限公司 | Method, device and server for determining network service quality |
EP3721570B1 (en) * | 2017-12-19 | 2022-08-31 | Huawei Technologies Co., Ltd. | Client device and methods thereof for transmitting control messages |
SE541441C2 (en) * | 2017-12-28 | 2019-10-01 | DeviceRadio AB | Method and a node for preventing latency in data packet transmission |
US11871265B2 (en) | 2018-01-18 | 2024-01-09 | Samsung Electronics Co., Ltd | Method and an electronic device for dynamically controlling TCP congestion window |
US10841231B2 (en) * | 2018-11-15 | 2020-11-17 | Mellanox Technologies, Ltd. | Transmission timeout system |
CN110113777B (en) * | 2019-04-29 | 2022-03-01 | 深圳广熵科技有限公司 | Data acquisition method |
US10999206B2 (en) | 2019-06-27 | 2021-05-04 | Google Llc | Congestion control for low latency datacenter networks |
RU2747174C1 (en) * | 2020-07-14 | 2021-04-28 | Юрий Иванович Стародубцев | Method for proactive reconfiguration of communication network structure providing information exchange in interests of corporate management system under conditions of destructive influence |
RU2764390C1 (en) * | 2020-07-15 | 2022-01-17 | Юрий Иванович Стародубцев | Method for assessing awareness of source of destructive impacts on structure of corporate management system |
US11403175B2 (en) * | 2020-07-31 | 2022-08-02 | EMC IP Holding Company LLC | Rebuilding data previously stored on a failed data storage drive |
CN114401224B (en) * | 2022-01-19 | 2023-07-11 | 平安科技(深圳)有限公司 | Data current limiting method and device, electronic equipment and storage medium |
CN116828017A (en) * | 2022-03-22 | 2023-09-29 | 腾讯科技(深圳)有限公司 | Data processing method and device in frame synchronization game and computer equipment |
Family Cites Families (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5193151A (en) | 1989-08-30 | 1993-03-09 | Digital Equipment Corporation | Delay-based congestion avoidance in computer networks |
NZ331915A (en) | 1996-03-18 | 1999-05-28 | Gen Instrument Corp | Dynamic bandwidth allocation for a communication network |
US6038216A (en) | 1996-11-01 | 2000-03-14 | Packeteer, Inc. | Method for explicit data rate control in a packet communication environment without data rate supervision |
US6076114A (en) | 1997-04-18 | 2000-06-13 | International Business Machines Corporation | Methods, systems and computer program products for reliable data transmission over communications networks |
US6105064A (en) | 1997-05-30 | 2000-08-15 | Novell, Inc. | System for placing packets on network for transmission from sending endnode to receiving endnode at times which are determined by window size and metering interval |
US6097697A (en) | 1998-07-17 | 2000-08-01 | Sitara Networks, Inc. | Congestion control |
EP1018821A1 (en) * | 1999-01-08 | 2000-07-12 | TELEFONAKTIEBOLAGET L M ERICSSON (publ) | Communication device and method |
US6611495B1 (en) | 1999-02-22 | 2003-08-26 | Telefonaktiebolaget Lm Ericsson (Publ) | System and method for improved data transfer in packet-switched communication networks |
US7058723B2 (en) * | 2000-03-14 | 2006-06-06 | Adaptec, Inc. | Congestion control for internet protocol storage |
WO2002019654A2 (en) * | 2000-08-31 | 2002-03-07 | The Regents Of The University Of California | Method for improving tcp performance over wireless links |
US7130268B2 (en) * | 2000-10-17 | 2006-10-31 | Saverio Mascolo | End-to-end bandwidth estimation for congestion control in packet switching networks |
US20020071388A1 (en) * | 2000-11-16 | 2002-06-13 | Einar Bergsson | Selectable network protocol |
-
2002
- 2002-01-29 US US10/061,574 patent/US7099273B2/en active Active
- 2002-04-11 EP EP02733982A patent/EP1393508B1/en not_active Expired - Lifetime
- 2002-04-11 AT AT02733982T patent/ATE338407T1/en not_active IP Right Cessation
- 2002-04-11 WO PCT/US2002/011779 patent/WO2002084960A2/en active IP Right Grant
- 2002-04-11 DE DE60214369T patent/DE60214369D1/en not_active Expired - Lifetime
Also Published As
Publication number | Publication date |
---|---|
EP1393508A2 (en) | 2004-03-03 |
DE60214369D1 (en) | 2006-10-12 |
WO2002084960A3 (en) | 2003-12-11 |
WO2002084960A2 (en) | 2002-10-24 |
EP1393508B1 (en) | 2006-08-30 |
US7099273B2 (en) | 2006-08-29 |
US20020150048A1 (en) | 2002-10-17 |
ATE338407T1 (en) | 2006-09-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7099273B2 (en) | Data transport acceleration and management within a network communication system | |
US11876714B2 (en) | Method and apparatus for network congestion control based on transmission rate gradients | |
Sinha et al. | WTCP: A reliable transport protocol for wireless wide-area networks | |
US7136353B2 (en) | Quality of service management for multiple connections within a network communication system | |
US7369498B1 (en) | Congestion control method for a packet-switched network | |
US7061856B2 (en) | Data throughput over lossy communication links | |
US7184401B2 (en) | Link-aware transmission control protocol | |
EP2387837B1 (en) | Maximizing bandwidth utilization in networks with high latencies and packet drops using transmission control protocol | |
US6625118B1 (en) | Receiver based congestion control | |
EP1142226B1 (en) | Communication device and method | |
US6535482B1 (en) | Congestion notification from router | |
WO2006065008A1 (en) | Apparatus for arq controlling in wireless portable internet system and method thereof | |
US7738395B2 (en) | Communication system for improving data transmission efficiency of TCP in a wireless network environment and a method thereof | |
CA2372023A1 (en) | Overload control method for a packet-switched network | |
Aaron et al. | Techniques to improve TCP over wireless links |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A2 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ OM PH PL PT RO RU SD SE SG SI SK SL TJ TM TN TR TT TZ UA UG US US US UZ VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A2 Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
AK | Designated states |
Kind code of ref document: C2 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ OM PH PL PT RO RU SD SE SG SI SK SL TJ TM TN TR TT TZ UA UG US US US UZ VN YU ZA ZM ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: C2 Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 2002733982 Country of ref document: EP |
|
REG | Reference to national code |
Ref country code: DE Ref legal event code: 8642 |
|
WWP | Wipo information: published in national office |
Ref document number: 2002733982 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: JP |
|
WWW | Wipo information: withdrawn in national office |
Country of ref document: JP |
|
WWG | Wipo information: grant in national office |
Ref document number: 2002733982 Country of ref document: EP |