WO1993026109A1 - Apparatus for providing cryptographic support in a network - Google Patents
Apparatus for providing cryptographic support in a network Download PDFInfo
- Publication number
- WO1993026109A1 WO1993026109A1 PCT/US1993/005462 US9305462W WO9326109A1 WO 1993026109 A1 WO1993026109 A1 WO 1993026109A1 US 9305462 W US9305462 W US 9305462W WO 9326109 A1 WO9326109 A1 WO 9326109A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- cell
- providing
- interface
- reassembler
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/04—Network architectures or network communication protocols for network security for providing a confidential data exchange among entities communicating through data packet networks
-
- 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
- H04L9/06—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols the encryption apparatus using shift registers or memories for block-wise or stream coding, e.g. DES systems or RC4; Hash functions; Pseudorandom sequence generators
- H04L9/0618—Block ciphers, i.e. encrypting groups of characters of a plain text message using fixed encryption transformation
- H04L9/0637—Modes of operation, e.g. cipher block chaining [CBC], electronic codebook [ECB] or Galois/counter mode [GCM]
-
- 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
- H04L9/08—Key distribution or management, e.g. generation, sharing or updating, of cryptographic keys or passwords
- H04L9/0816—Key establishment, i.e. cryptographic processes or cryptographic protocols whereby a shared secret becomes available to two or more parties, for subsequent use
- H04L9/0819—Key transport or distribution, i.e. key establishment techniques where one party creates or otherwise obtains a secret value, and securely transfers it to the other(s)
- H04L9/083—Key transport or distribution, i.e. key establishment techniques where one party creates or otherwise obtains a secret value, and securely transfers it to the other(s) involving central third party, e.g. key distribution center [KDC] or trusted third party [TTP]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L2209/00—Additional information or applications relating to cryptographic mechanisms or cryptographic arrangements for secret or secure communication H04L9/00
- H04L2209/12—Details relating to cryptographic hardware or logic circuitry
- H04L2209/125—Parallelization or pipelining, e.g. for accelerating processing of cryptographic operations
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L2209/00—Additional information or applications relating to cryptographic mechanisms or cryptographic arrangements for secret or secure communication H04L9/00
- H04L2209/24—Key scheduling, i.e. generating round keys or sub-keys for block encryption
-
- 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
- H04L69/30—Definitions, standards or architectural aspects of layered protocol stacks
- H04L69/32—Architecture of open systems interconnection [OSI] 7-layer type protocol stacks, e.g. the interfaces between the data link level and the physical level
- H04L69/322—Intralayer communication protocols among peer entities or protocol data unit [PDU] definitions
- H04L69/324—Intralayer communication protocols among peer entities or protocol data unit [PDU] definitions in the data link layer [OSI layer 2], e.g. HDLC
Definitions
- the present invention generally relates to high ⁇ speed networks and more particularly relates to a high- performance host interface with cryptographic support.
- One preferred application of the present invention is in the host interface described in U.S. patent application serial no. 708,775, titled "High-Performance Host Interface For ATM Networks," filed May 28, 1991, which is hereby incorporated by reference into this specification as if fully set forth herein.
- Network usage and capabilities are both increasing at high rates. For example, the traffic load on the National Science Foundation's NSFNET backbones in the United States doubles every few months. Much of this increased traffic is due to the increased connectivity provided by interconnection with other networks, large numbers of personal workstations connected to LANs, and large numbers of computers connected via 9600 bit per second (bps) dial-ins. Much of this traffic is traditional Internet (inter-network) traffic, e.g., electronic mail, netnews, and file transfer, although a significant percentage of traffic is due to remote terminal sessions via login and telnet.
- inter-network Internet
- Bandwidth is the most important area in which improvement is taking place, as bandwidth can be used to achieve a variety of other quality-of-service goals.
- network technologies with bandwidths near 1 Gbps will be deployed in several areas of the United States. See “Gigabit Testbed Initiative Summary,” Corporation for National Research Initiatives, 1985 Preston White Drive, Suite 100, Reston, VA 22091 USA (January, 1992) info ⁇ nri.reston,va,us.
- AURORA for example, is an experimental wide area network testbed whose main objective is the exploration and evaluation of network technologies.
- the Gbps network will link four sites:
- AURORA gigabit per second networks
- the major axes of choice are: choice of encryption algorithm, hardware or software implementation, and placement of cryptographic support in a networking architecture.
- DES Data Encryption Standard
- FIPS Publication 46 Data Encryption Standard
- RSA Rivest-Shamir-Adleman
- R.L. Rivest, A. Shamir, and L. Adleman "A method of obtaining digital signatures and public-key cryptosystems, " Communications of the ACM 21(2), pp. 120-126 (February 1978)
- a primary object of the present invention is to provide a high speed (e.g., gigabit per second) network with cryptographic support.
- cryptographic support apparatus is combined with the host/network interface.
- an interface for coupling a host computer to a high-speed telecommunications network comprises reassembler means for reassembling and decrypting data that has been encrypted in accordance with a pre-defined key and segmented into a plurality of asynchronous transfer mode (ATM) cells, where each cell preferably comprises a virtual channel identifier (VCI) , a multiplexing identifier (MID) (if the data is transmitted using the CCITT specified Class 4 connectionless transfer ATM adaptation layer (AAL) ) , and a cell body.
- VCI virtual channel identifier
- MID multiplexing identifier
- the reassembler means comprises cell manager means for separating each cell body from that cell body's corresponding VCI (and MID if present) , linked list manager means for storing linked list data indicative of addresses at which said cell bodies are to be stored, content addressable memory (CAM) means for storing said VCIs (and MIDs) and providing pointers into said linked list data, lookup controller means for writing said VCIs (and MIDs) to said CAM means, reassembly buffer means for storing said cell bodies, and decryption means for decrypting said data in accordance with said key and writing the decrypted data to said reassembly buffer means.
- cell manager means for separating each cell body from that cell body's corresponding VCI (and MID if present)
- linked list manager means for storing linked list data indicative of addresses at which said cell bodies are to be stored
- CAM content addressable memory
- lookup controller means for writing said VCIs (and MIDs) to said CAM means
- One preferred embodiment of the invention further comprises segmenter means for encrypting data received from a host computer and segmenting the encrypted data into a plurality of ATM cells.
- the present invention also encompasses host interfaces comprising means for providing private keys for each of a plurality of virtual circuits, means for providing real-time agile keying across said virtual circuits, means for providing concurrent cleartext and ciphertext streams, means for providing application- specific control of cryptographic keying, and means for providing cleartext headers and ciphertext payloads. Other features of the present invention are described below.
- Figure 1 is a block diagram of the AURORA network topology.
- Figure 2 depicts a single unit, sixteen round DES block.
- Figure 3 illustrates the expansion of the single- unit DES block of Figure 2 into four DES units, with four rounds per unit.
- Figure 4 depicts a network architecture employing bus-connected cryptographic support.
- Figure 5 is a block diagram of a Segmenter of a host interface.
- Figure 6 is a block diagram of a Reassembler of a host interface.
- Figure 7 is a block diagram of a Segmenter modified in accordance with the present invention to include support for encryption.
- Figure 8 is a block diagram of a Reassembler modified in accordance with the present invention to include support for decryption.
- DES Data Encryption Standard
- Computer data communications rates have also increased, and today's high- performance computer networks further increase the encryption bandwidth needed to avoid performance bottlenecks in systems requiring security.
- Architectural means to increase DES throughput have been studied by the present inventors, with the intent of producing a DES implementation to satisfy these bandwidth demands.
- a key idea underlying the present invention is the use of the physical parallelism possible in a hardware realization to improve performance without resorting to exotic technologies. As described below, a useful system can be implemented when such parallelism can be cheaply exploited using architectural innovations.
- a cryptographic subsystem was analyzed as a series of levels, ranging from the basic combinational logic of a single level to the use of overlapped operation of the host processor and the cryptographic subsystem.
- the DES is implemented as sixteen "rounds" of key-directed encipherment; each round implements permutations and substitutions on the 64-bit input using the 56 bits of key for selection of the particular transformation.
- the 64 bits of output of a round are fed back as input for the next round until sixteen rounds are achieved, at which point the 64 bits are output.
- the algorithm can be split into several parallel computations for increased speed.
- the time required can be effectively overlapped with the use of the subkey in the rest of the round operation.
- An additional overlap can be made of the two stages of exclusive-or (XOR) gates at the expense of increased complexity and gatecount.
- DMA Direct Memory Access
- Computation elements may be configured to implement a fraction of the rounds in a pipeline-like approach.
- the main idea is the conversion of the iterative "rounds" of DES into a pipeline-like stream processing structure, analogous to hardware "loop unrolling.” This can be accomplished by tagging each 64-bit input with the 56-bit key with which it undergoes the single-round transformation; a round can effectively be thought of as a 120-bit wide piece of combinational logic.
- a single-board implementation of the DES requires that the sixteen prescribed rounds of DES be performed on the 64-bit data input and 56-bit key input before it is presented as output. The sixteen rounds mean that sixteen passes through the internal data path of any implementation be performed. Such a situation is illustrated in Figure 2 .
- each data block is accompanied by its key throughout the computation.
- Switching keys between successive data blocks without flushing the pipeline is possible because the key and data streams are synchronized.
- the tradeoff in keying interconnection may not be worthwhile as compared to maintaining separate key registers for each stage of the pipeline.
- Each round would then maintain its own key load (shadow) register in this approach.
- the standard key shifting sequence would be partially executed on each stage. Since a partial execution of the key schedule would not result in a complete cycling of the keytext, the key would be reloaded from the shadow register when it has completed its share of the computation on a data block.
- a limitation of the pipeline approach is the bandwidth ceiling imposed by the number of rounds in the algorithm (sixteen) . This means that a single pipeline of processors cannot provide more than sixteen times the encryption bandwidth of a single processor. Additionally, the pipeline suffers in scalability since the number of stages must be a factor of sixteen. This disadvantage is more notable when considering an upgrade of a pipeline: to gain any increase in bandwidth requires a doubling in computation resources. DMA support decouples the host processor from the encryption function to allow CPU processing of other tasks to proceed in parallel with the encryption request. See David P. Anderson and P. Venkat Rangan, "High-Performance Interface Architectures for Cryptographic Hardware," EURO-CRYPT ' 87 Proceedings , Springer-Verlag, Amsterdam (1987) .
- VLSI Technology's VM007 encryption chips were chosen as the basis for a Micro Channel Architecture bus card in one embodiment of the present invention.
- the chip has a number of important features, most notably the fact that it can be clocked at 30 Mhz, yielding Electronic Code Book (ECB) and
- Cipher Block Chaining speeds of 192 Mpbs. This is entirely adequate for the current generation host interfaces, which typically operate at OC-3c speeds (155 Mbps) .
- a key can be loaded in 8 clock cycles, or about 0.3 microseconds. Several registers for keying are provided.
- a Micro Channel Architecture board has been implemented and tested, and software support has been written. It operates at rated speed, which exceeds the maximum transfer rate (130 Mbps) possible between the host and Micro Channel Architecture devices with current generation Micro Channel Architecture I/O Channel Controllers.
- a prototype ATM host interface for the IBM RISC system/6000 has been designed and implemented. See C. Brendan S. Traw and Jonathan M. Smith, "A High-Performance Host Interface for ATM Networks," in Proceedings, SIGCOMM 1991 , Zurich, SWITZERLAND (September 4-6, 1991), pp. 317- 325.
- the host interface is implemented as a pair of Micro Channel Architecture boards. See IBM Corporation, IBM RISC System/6000 POWERstation and POWERserver: Hardware Technical Reference, Micro Channel Architecture, IBM Order Number SA23-2647-00, 1990.
- the Segmenter converts data objects, such as IP packets, to ATM cells.
- ATM cells include 48 bytes of data and five bytes of header, so that a reasonably sized data object (e.g., 8192 bytes) on the machine requires many ATM cells to be put on the network medium.
- the Reassembler takes ATM cells from the network and re-aggregates them into data objects. Among the functions the Segmenter and Reassembler perform are the computation of CRC checksums, multiplexing and demultiplexing of virtual circuits, and a small amount (128 Kbytes) of buffering.
- the model for the host interface's role in a system is this: the host software makes any control decisions, while the host interface performs cellification, decellification, and any data movement to and from the host. This significantly reduces the burden on the workstation, as it is organized for data processing, not memory manipulation.
- a block diagram of a Segmenter 10 is presented in Figure 5.
- the block diagram illustrates the Segmenter 10, comprising a SONET (synchronous optical network) framer 12, ATM header generator 14, ATM adaptation layer (AAL) generator 16, FIFO data buffer 18 and segmentation controller 20, coupled to a Micro Channel interface/bus 22.
- VCI virtual circuit identifier
- a multiplexing identifier (MID) is loaded into the AAL generator 16 if the data is to be transmitted via the Class 4 ATM adaption layer (AL4) . See CCITT Recommendation 1.363, B-ISDN ATM Adaptation Layer (AAL) Specification, 1990.
- the host then sets up a streaming mode (an optimized bus transfer mode for contiguous data) transfer to move the data which is to be transmitted from a pinned buffer in host memory (not shown) via the Micro Channel interface/bus 22 to the FIFO buffer 18.
- a streaming mode an optimized bus transfer mode for contiguous data
- the location and size of the host's buffer are specified during stream set-up. While this transfer is being made, the Segmenter 10 produces the header check CRC and formats the control information into the appropriate ATM and AAL4 header formats.
- the segmentation controller 20 removes the data for the first cell from the FIFO buffer 18 and adds an ATM header, AAL4 header and AAL4 trailer. If the cell is carrying AAL4 data, the payload CRC is calculated as the data is moved to the SONET framer 12 and placed in the appropriate field at the end of the cell. See Thomas J. Robe and Kenneth A. Walsh, "A SONET STS-3C User-Network Interface IC," in Proceedings, Custom Integrated Circuits Conference , San Diego, CA (May, 1991) . This process is repeated until the FIFO buffer is drained.
- a Reassembler 30 is illustrated in Figure 6.
- the Reassembler 30 includes a cell manager 32, content addressable memory (CAM) lookup controller 34, linker list manager (LLM) 36 and dual port reassembly buffer (DPRB) 38. These four major subsections operate concurrently.
- the cell manager 32, CAM lookup controller 34, dual port reassembly buffer 38 and linked list manager 36 exploit this concurrency to form an ATM cell-processing "pipeline.”
- the cell manager 34 verifies the integrity of the header and payload (if the cell is carrying AAL4 data) of cells received from the network by the SONET framer. It extracts the VCI from the ATM header and the MID and length from the AAL header.
- the body of the cell is placed in a FIFO buffer (not shown) for later movement into the dual port reassembly buffer 38.
- the CAM lookup controller 34 manages two CAMs that provide lookup support for a total of 256 simultaneous virtual connections and the reassembly of 246 datagrams (a datagram is a finite length packet with sufficient information to be routed from source to destination independently of previous transmissions) .
- the host (not shown) is able to flush undesired virtual circuits and datagrams from the reassembler through the CAM lookup controller.
- a reference resulting from the CAM lookup operation is passed to the LLM 36.
- the LLM establishes and maintains a linked list data structure for each of the virtual circuit and datagrams that is being received. Data received from the network is placed at the end of the appropriate list while data destined for host memory is read from the front of the list.
- the LLM 36 allocates space in the DPRB for data coming into the Reassembler 30 from the network and passes the location to the DPRB controller 38.
- the cell body placed into the FIFO (not shown) by the cell manager 32 is removed and written into the appropriate reassembly buffer inside the DPRB controller 38.
- the host (not shown) is able to read data from a particular virtual circuit or datagram by specifying a list reference to the LLM 36 that determines the location in the DPRB where data is stored.
- the location is passed to the DPRB controller 38 and the DPRB controller removes the data from the buffer for transfer into host memory (not shown) over the Micro Channel bus.
- Performance of the above-described architecture is quite good. Measurements indicate that the system can send and receive at about 130 Mbps. Detailed performance data is available. See C. Brendan S. Traw and Jonathan M. Smith, "Implementation and Performance of an ATM Host Interface for Workstations," in Proceedings, IEEE Workshop on the Architecture and Implementation of High-Performance Communications Subsystems (HPCS '92), Tuscon, AZ (February 17-19, 1992).
- link encryption itself has several drawbacks. First, it provides little control for applications, e.g., for customized key choices. Second, it makes key management for arbitrary host-to-host encryption in a switched network difficult; packet-switching exacerbates this problem since headers are encrypted and must be readable if used for routing. A solution that combines some of the flexibility of the separate card architectures with the high-performance characteristics of a link-encryption solution is described below.
- FIG 7 is a block diagram of a Segmenter 10' modified in accordance with the present invention to include support for encryption.
- This support is provided by an encryption unit 40, which may comprise a VM007 data encryption processor.
- Figure 8 is a block diagram of a Reassembler 30' modified to include support for decryption.
- Decryption support is provided by a decryption unit 42, which may also employ the VM007 processor.
- Blocks 40 and 42 contain logic circuitry for controlling key latching and loading. Logic circuitry of this type will be known to those skilled in the art after reading this specification, therefore a detailed description is unnecessary.
- the addition of the encryption/decryption chip 40, 42 in each case serves to add another functional block that can operate concurrently with the other blocks.
- Key lookup is done by adding an extra field to the "head-of- list" data structure in the pointer memory.
- These "head- of-list" structures are located using the CAM with the Virtual-Circuit Identifier as a tag.
- the cryptographic keys are loaded into the VM007 in parallel with data movement from the cell manager's data buffer into the dual port reassembly buffer.
- Virtual Circuits are currently supported in the host interface) .
- cryptographic hardware can advantageously be incorporated into the host/network interface. This adds a marginal amount of complexity to the interface and comfortably meshes with functions the interface already performs, such as cellification, decellification, and computing checksums.
- the advantages to this approach stem from the fact that neither host memory access nor bus transactions are required beyond those transactions necessary to move data from the application data areas to the host/network interface. Key management is closely correlated with ATM header manipulation, and can be efficiently performed with the addition of a small amount of logic in the host interface.
Abstract
A host interface comprising a reassembler (30) for reassembling and decrypting data that has been encrypted in accordance with a pre-defined key and segmented into a plurality of asynchronous transfer mode cells. Each cell comprises a virtual channel identifier, and a cell body. The reassembler (30) comprises a cell manager (32) for separating each cell body, a linked list manager (36) for managing and storing linked list data indicative of addresses at which the cells bodies are to be stored, a lookup controller (34) for writing the virtual channel identifiers for the cells to a content addressable memory to be managed and stored, a reassembly buffer (38) for storing the cells bodies, and a decryption device (42) for decrypting the data and writing the decrypted data to the reassembly buffer (38).
Description
APPARATUS FOR PROVIDING CRYPTOGRAPHIC SUPPORT IN A NETWORK FIELD OF THE INVENTION
The present invention generally relates to high¬ speed networks and more particularly relates to a high- performance host interface with cryptographic support. One preferred application of the present invention is in the host interface described in U.S. patent application serial no. 708,775, titled "High-Performance Host Interface For ATM Networks," filed May 28, 1991, which is hereby incorporated by reference into this specification as if fully set forth herein.
BACKGROUND OF THE INVENTION
Network usage and capabilities are both increasing at high rates. For example, the traffic load on the National Science Foundation's NSFNET backbones in the United States doubles every few months. Much of this increased traffic is due to the increased connectivity provided by interconnection with other networks, large numbers of personal workstations connected to LANs, and large numbers of computers connected via 9600 bit per second (bps) dial-ins. Much of this traffic is traditional Internet (inter-network) traffic, e.g., electronic mail, netnews, and file transfer, although a significant percentage of traffic is due to remote terminal sessions via login and telnet.
Networks are improving in a number of ways. Bandwidth is the most important area in which improvement is taking place, as bandwidth can be used to achieve a variety of other quality-of-service goals. In the very
near future, network technologies with bandwidths near 1 Gbps will be deployed in several areas of the United States. See "Gigabit Testbed Initiative Summary," Corporation for National Research Initiatives, 1985 Preston White Drive, Suite 100, Reston, VA 22091 USA (January, 1992) infoønri.reston,va,us.
AURORA, for example, is an experimental wide area network testbed whose main objective is the exploration and evaluation of network technologies. The Gbps network will link four sites:
(1) Bellcore's Morristown Research and Engineering Laboratory in Morristown, NJ;
(2) IBM Research's Computer Science Laboratory in Hawthorne, NY; (3) MIT's Laboratory for Computer Science in
Cambridge, MA; (4) University of Pennsylvania's Distributed Systems Laboratory in Philadelphia, PA. The topology of AURORA is illustrated in Figure 1. Among the applications envisioned for gigabit per second networks (e.g., AURORA) are interactive teleconferencing, advanced multimedia systems with support for sensory data, and advanced displays with extremely high-quality imaging capabilities. Many of these applications would not be possible without large available bandwidths. Many more applications can be imagined, and it is expected that many may be commercialized.
Many proposed applications, however, will not be feasible without significant attention paid to the issue of security. Traditionally, security schemes relied on either administrative means (e.g., systems with extremely restricted access, careful monitoring and audits) or cryptographic support, or a combination of these techniques. Since networks, by their nature, require distributed control, cryptographic means have somewhat more utility than solely administrative means to achieve security. A detailed survey of cryptographic techniques is
available in D.R. Denning, Cryptography and Data Security, Addison-Wesley (1982) .
The major axes of choice are: choice of encryption algorithm, hardware or software implementation, and placement of cryptographic support in a networking architecture. Of the available cryptographic methods that have withstood significant attack, the Data Encryption Standard (DES) (described in NBS, Data Encryption Standard (FIPS Publication 46) , National Bureau of Standards, U.S. Department of Commerce, Washington, DC (January, 1977)) and the Rivest-Shamir-Adleman (RSA) scheme (described in R.L. Rivest, A. Shamir, and L. Adleman, "A method of obtaining digital signatures and public-key cryptosystems, " Communications of the ACM 21(2), pp. 120-126 (February 1978)) seem most attractive. Although the use of public- key technology offers many advantages, the poor performance of RSA implementations are a major problem. A recent survey of hardware implementations of the RSA algorithm indicates that the fastest available implementations are no faster than 1 Mbps. Significantly faster implementations of the private-key DES algorithm are available in hardware and architectural techniques for further improving its performance have been reported. See 1989 IC MASTER, Fact Sheet, Western Digital WDD20C03A, 1989; VLSI Technology, Inc. VM007 Data Encryption Processor: Advance Information Sheet, 8375 South River Parkway, October, 1991; and Albert G. Broscius, Hardware Analysis and Implementation of the NBS Data Encryption Standard, University of Pennsylvania, School of Engineering and Applied Sciences (April, 1991) , MSE Thesis (CIS) .
Many applications envisioned for ultra-high-speed networks require cryptographic transformations for data in transit. The need for security and privacy will drive transformations in workstation architectures, especially high-performance communications subsystems. There is a need for apparatus and methods for providing cost-effective
and high-performance cryptographic support in high speed networks while sacrificing little in flexibility.
SUMMARY OF THE INVENTION
Accordingly, a primary object of the present invention is to provide a high speed (e.g., gigabit per second) network with cryptographic support. As described below, according to a presently preferred embodiment of the invention, cryptographic support apparatus is combined with the host/network interface. According to the present invention, an interface for coupling a host computer to a high-speed telecommunications network comprises reassembler means for reassembling and decrypting data that has been encrypted in accordance with a pre-defined key and segmented into a plurality of asynchronous transfer mode (ATM) cells, where each cell preferably comprises a virtual channel identifier (VCI) , a multiplexing identifier (MID) (if the data is transmitted using the CCITT specified Class 4 connectionless transfer ATM adaptation layer (AAL) ) , and a cell body. In preferred embodiments of the invention, the reassembler means comprises cell manager means for separating each cell body from that cell body's corresponding VCI (and MID if present) , linked list manager means for storing linked list data indicative of addresses at which said cell bodies are to be stored, content addressable memory (CAM) means for storing said VCIs (and MIDs) and providing pointers into said linked list data, lookup controller means for writing said VCIs (and MIDs) to said CAM means, reassembly buffer means for storing said cell bodies, and decryption means for decrypting said data in accordance with said key and writing the decrypted data to said reassembly buffer means. One preferred embodiment of the invention further comprises segmenter means for encrypting data received from a host computer and segmenting the encrypted data into a plurality of ATM cells.
The present invention also encompasses host interfaces comprising means for providing private keys for each of a plurality of virtual circuits, means for providing real-time agile keying across said virtual circuits, means for providing concurrent cleartext and ciphertext streams, means for providing application- specific control of cryptographic keying, and means for providing cleartext headers and ciphertext payloads. Other features of the present invention are described below.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is a block diagram of the AURORA network topology.
Figure 2 depicts a single unit, sixteen round DES block. Figure 3 illustrates the expansion of the single- unit DES block of Figure 2 into four DES units, with four rounds per unit.
Figure 4 depicts a network architecture employing bus-connected cryptographic support. Figure 5 is a block diagram of a Segmenter of a host interface.
Figure 6 is a block diagram of a Reassembler of a host interface.
Figure 7 is a block diagram of a Segmenter modified in accordance with the present invention to include support for encryption.
Figure 8 is a block diagram of a Reassembler modified in accordance with the present invention to include support for decryption.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
The Data Encryption Standard (DES) is the most widely used publicly available secret-key algorithm. Since its introduction, DES implementations have improved in achieving high encryption rates. Computer data communications rates have also increased, and today's high-
performance computer networks further increase the encryption bandwidth needed to avoid performance bottlenecks in systems requiring security. Architectural means to increase DES throughput have been studied by the present inventors, with the intent of producing a DES implementation to satisfy these bandwidth demands.
A key idea underlying the present invention is the use of the physical parallelism possible in a hardware realization to improve performance without resorting to exotic technologies. As described below, a useful system can be implemented when such parallelism can be cheaply exploited using architectural innovations. A cryptographic subsystem was analyzed as a series of levels, ranging from the basic combinational logic of a single level to the use of overlapped operation of the host processor and the cryptographic subsystem. The DES is implemented as sixteen "rounds" of key-directed encipherment; each round implements permutations and substitutions on the 64-bit input using the 56 bits of key for selection of the particular transformation. The 64 bits of output of a round are fed back as input for the next round until sixteen rounds are achieved, at which point the 64 bits are output.
At the combinational logic level, where a single round is implemented, the algorithm can be split into several parallel computations for increased speed. By generating subkeys one cycle in advance, the time required can be effectively overlapped with the use of the subkey in the rest of the round operation. An additional overlap can be made of the two stages of exclusive-or (XOR) gates at the expense of increased complexity and gatecount.
Use of multiple round implementations can increase computation bandwidth if the DES mode of operation chosen does not require feedback of ciphertext (see National Bureau of Standards, Federal Information
Processing Standard #81: Operational Modes of the DES), ruling out all but the Electronic Code Book (ECB) method.
Unfortunately, the ECB method is known to be susceptible to plaintext frequency-analysis based attacks, since identical input blocks result in identical output blocks. A proposed operating method resists this attack yet does not require feedback of ciphertext. See Anthony McAuley and David C. Feldmeier, Minimizing Protocol Ordering Constraints to Improve Performance, available via anonymous ftp from Internet host thumper.bellcore.com, 1991.
At the board level, concurrent I/O processing and DES computation provides for steady-state operation of the encryption unit. This can be accomplished using double- buffering and a small amount of extra logic. In addition to this buffering, the use of Direct Memory Access (DMA) for the encryption board allows the host processor to continue other work concurrently with whatever the encryption unit is achieving.
Computation elements may be configured to implement a fraction of the rounds in a pipeline-like approach. The main idea is the conversion of the iterative "rounds" of DES into a pipeline-like stream processing structure, analogous to hardware "loop unrolling." This can be accomplished by tagging each 64-bit input with the 56-bit key with which it undergoes the single-round transformation; a round can effectively be thought of as a 120-bit wide piece of combinational logic. A single-board implementation of the DES requires that the sixteen prescribed rounds of DES be performed on the 64-bit data input and 56-bit key input before it is presented as output. The sixteen rounds mean that sixteen passes through the internal data path of any implementation be performed. Such a situation is illustrated in Figure 2 .
However, if one employs multiple boards, for example four, as illustrated in Figure 3, one can reduce the number of rounds performed by each board. There is a small increase in the logic for managing the keying, but this is a second-order effect as far as performance is concerned. The primary relationship is this: with up to a
maximum of sixteen DES units (one per round) , there is a linear cost/performance tradeoff achievable by (1) doubling the number of DES units, and (2) halving the number of rounds each unit performs. For example, Figure 3 shows four units each performing four rounds. The upper limit is sixteen units, each performing one round.
Ignoring second order effects, the major delay is induced by a round. Thus, as each "expansion" takes place the unit count is doubled and the round count is halved, meaning that the delay for the first bit through the sixteen unit case is almost exactly the same as the delay for the first bit through the single unit case.
When keying needs to be frequently updated, the pipeline style allows a matching of the datapath flow with a parallel keypath. In this way, each data block is accompanied by its key throughout the computation. Switching keys between successive data blocks without flushing the pipeline is possible because the key and data streams are synchronized. For infrequent key changes, the tradeoff in keying interconnection may not be worthwhile as compared to maintaining separate key registers for each stage of the pipeline. Each round would then maintain its own key load (shadow) register in this approach. The standard key shifting sequence would be partially executed on each stage. Since a partial execution of the key schedule would not result in a complete cycling of the keytext, the key would be reloaded from the shadow register when it has completed its share of the computation on a data block. A limitation of the pipeline approach is the bandwidth ceiling imposed by the number of rounds in the algorithm (sixteen) . This means that a single pipeline of processors cannot provide more than sixteen times the encryption bandwidth of a single processor. Additionally, the pipeline suffers in scalability since the number of stages must be a factor of sixteen. This disadvantage is more notable when considering an upgrade of a pipeline: to
gain any increase in bandwidth requires a doubling in computation resources. DMA support decouples the host processor from the encryption function to allow CPU processing of other tasks to proceed in parallel with the encryption request. See David P. Anderson and P. Venkat Rangan, "High-Performance Interface Architectures for Cryptographic Hardware," EURO-CRYPT ' 87 Proceedings , Springer-Verlag, Amsterdam (1987) .
_ A DES board using SSI TTL and MSI PALs employing the MUX key register approach has been developed. See Albert G. Broscius and Jonathan M. Smith, "Exploiting Parallelism in Hardware Implementation of the DES," Proceedings, CRYPTO 1991 Conference , Santa Barbara, CA (August, 1991) . Testing of a wirewrapped prototype with a single algorithm kernel indicated an encryption rate of 93 Mbps. An interface to the Micro Channel Architecture (MCA) bus of the IBM RISC System/6000 was also implemented. This implementation reveals a great deal about the internal structure of the DES hardware itself. Unfortunately, the parts cost of the implementation, potential performance limitations due to packaging, and the large amount of handwork required to replicate the work inhibited its use. In addition, concurrently with the completion of the work, very high-speed DES chips were made commercially available. See VLSI Technology, Inc. VM007 Data Encryption Processor: Advance Information Sheet.
Experience with the wire-wrapped board indicates that high encryption speeds are possible, and that architectural solutions can provide considerable leverage, independent of technology choices for the electronics.
VLSI Technology's VM007 encryption chips were chosen as the basis for a Micro Channel Architecture bus card in one embodiment of the present invention. The chip has a number of important features, most notably the fact that it can be clocked at 30 Mhz, yielding Electronic Code Book (ECB) and
Cipher Block Chaining (CBC) speeds of 192 Mpbs. This is entirely adequate for the current generation host
interfaces, which typically operate at OC-3c speeds (155 Mbps) . A key can be loaded in 8 clock cycles, or about 0.3 microseconds. Several registers for keying are provided. A Micro Channel Architecture board has been implemented and tested, and software support has been written. It operates at rated speed, which exceeds the maximum transfer rate (130 Mbps) possible between the host and Micro Channel Architecture devices with current generation Micro Channel Architecture I/O Channel Controllers.
The major difficulty with this approach is the number of bus transactions necessary to move data to and from the network interface. This represents a serious performance limitation due more to the latency involved in roundtrips than in the bus throughput. Alternatives to this approach are described below. The nature of the traffic is illustrated in Figure 4.
The prototypes discussed above indicate that a truly high-performance encryption subsystem should preferably be incorporated into the host/network interface. A prototype ATM host interface for the IBM RISC system/6000 has been designed and implemented. See C. Brendan S. Traw and Jonathan M. Smith, "A High-Performance Host Interface for ATM Networks," in Proceedings, SIGCOMM 1991 , Zurich, SWITZERLAND (September 4-6, 1991), pp. 317- 325. The host interface is implemented as a pair of Micro Channel Architecture boards. See IBM Corporation, IBM RISC System/6000 POWERstation and POWERserver: Hardware Technical Reference, Micro Channel Architecture, IBM Order Number SA23-2647-00, 1990. The Segmenter converts data objects, such as IP packets, to ATM cells. These cells are then sent over the network. ATM cells include 48 bytes of data and five bytes of header, so that a reasonably sized data object (e.g., 8192 bytes) on the machine requires many ATM cells to be put on the network medium. The Reassembler takes ATM cells from the network and re-aggregates them into data objects.
Among the functions the Segmenter and Reassembler perform are the computation of CRC checksums, multiplexing and demultiplexing of virtual circuits, and a small amount (128 Kbytes) of buffering. The model for the host interface's role in a system is this: the host software makes any control decisions, while the host interface performs cellification, decellification, and any data movement to and from the host. This significantly reduces the burden on the workstation, as it is organized for data processing, not memory manipulation.
A block diagram of a Segmenter 10 is presented in Figure 5. The block diagram illustrates the Segmenter 10, comprising a SONET (synchronous optical network) framer 12, ATM header generator 14, ATM adaptation layer (AAL) generator 16, FIFO data buffer 18 and segmentation controller 20, coupled to a Micro Channel interface/bus 22. When data is to be transmitted into the network, the virtual circuit identifier (VCI) to be used is loaded into the ATM header generator 14. A multiplexing identifier (MID) is loaded into the AAL generator 16 if the data is to be transmitted via the Class 4 ATM adaption layer (AL4) . See CCITT Recommendation 1.363, B-ISDN ATM Adaptation Layer (AAL) Specification, 1990. The host then sets up a streaming mode (an optimized bus transfer mode for contiguous data) transfer to move the data which is to be transmitted from a pinned buffer in host memory (not shown) via the Micro Channel interface/bus 22 to the FIFO buffer 18. See IBM Corporation, IBM RISC system/6000 POWERstation and POWERserver: Hardware Technical Reference, General Information Manual, IBM Order Number SA23-2643-00, 1990. The location and size of the host's buffer are specified during stream set-up. While this transfer is being made, the Segmenter 10 produces the header check CRC and formats the control information into the appropriate ATM and AAL4 header formats. As soon as sufficient data has been placed into the FIFO buffer, the segmentation controller 20 removes the data for the first cell from the FIFO buffer 18
and adds an ATM header, AAL4 header and AAL4 trailer. If the cell is carrying AAL4 data, the payload CRC is calculated as the data is moved to the SONET framer 12 and placed in the appropriate field at the end of the cell. See Thomas J. Robe and Kenneth A. Walsh, "A SONET STS-3C User-Network Interface IC," in Proceedings, Custom Integrated Circuits Conference , San Diego, CA (May, 1991) . This process is repeated until the FIFO buffer is drained. A Reassembler 30 is illustrated in Figure 6. The Reassembler 30 includes a cell manager 32, content addressable memory (CAM) lookup controller 34, linker list manager (LLM) 36 and dual port reassembly buffer (DPRB) 38. These four major subsections operate concurrently. The cell manager 32, CAM lookup controller 34, dual port reassembly buffer 38 and linked list manager 36 exploit this concurrency to form an ATM cell-processing "pipeline." The cell manager 34 verifies the integrity of the header and payload (if the cell is carrying AAL4 data) of cells received from the network by the SONET framer. It extracts the VCI from the ATM header and the MID and length from the AAL header. The body of the cell is placed in a FIFO buffer (not shown) for later movement into the dual port reassembly buffer 38.
The CAM lookup controller 34 manages two CAMs that provide lookup support for a total of 256 simultaneous virtual connections and the reassembly of 246 datagrams (a datagram is a finite length packet with sufficient information to be routed from source to destination independently of previous transmissions) . The host (not shown) is able to flush undesired virtual circuits and datagrams from the reassembler through the CAM lookup controller.
A reference resulting from the CAM lookup operation is passed to the LLM 36. The LLM, as its name suggests, establishes and maintains a linked list data structure for each of the virtual circuit and datagrams that is being received. Data received from the network is
placed at the end of the appropriate list while data destined for host memory is read from the front of the list.
The LLM 36 allocates space in the DPRB for data coming into the Reassembler 30 from the network and passes the location to the DPRB controller 38. The cell body placed into the FIFO (not shown) by the cell manager 32 is removed and written into the appropriate reassembly buffer inside the DPRB controller 38. The host (not shown) is able to read data from a particular virtual circuit or datagram by specifying a list reference to the LLM 36 that determines the location in the DPRB where data is stored. The location is passed to the DPRB controller 38 and the DPRB controller removes the data from the buffer for transfer into host memory (not shown) over the Micro Channel bus.
Performance of the above-described architecture is quite good. Measurements indicate that the system can send and receive at about 130 Mbps. Detailed performance data is available. See C. Brendan S. Traw and Jonathan M. Smith, "Implementation and Performance of an ATM Host Interface for Workstations," in Proceedings, IEEE Workshop on the Architecture and Implementation of High-Performance Communications Subsystems (HPCS '92), Tuscon, AZ (February 17-19, 1992).
There are essentially two ways in which encryption hardware can be employed. One method is to make the encryption card generally available for access on the Micro Channel bus, as described above. This has the advantage that any application can employ the encryption hardware, for example, to encrypt and decrypt data written to disk, or to provide application-to-application cryptographic support. However, this approach, as illustrated in Figure 5, suffers from the need to access the bus multiple times for each network transaction; e.g., data must cross the bus to be encrypted, cross the bus after encryption and, if the post-encryption transfer is to
system memory, a third transfer into the host interface is required (although this is not strictly necessary on the Micro Channel Architecture as it supports card-to-card transfers) . Since the considerable bus activity will severely inhibit network performance, this method is somewhat undesirable.
The other alternative is to incorporate the cryptographic hardware into the host/network interface. This is often used when link encryption is desired, but has the drawback that the cryptographic hardware is private to the interface. In addition, link encryption itself has several drawbacks. First, it provides little control for applications, e.g., for customized key choices. Second, it makes key management for arbitrary host-to-host encryption in a switched network difficult; packet-switching exacerbates this problem since headers are encrypted and must be readable if used for routing. A solution that combines some of the flexibility of the separate card architectures with the high-performance characteristics of a link-encryption solution is described below.
As shown in Figures 5 and 6, both the Segmenter and Reassembler portion of the host interface are broken into several functional blocks, which operate concurrently. This hardware concurrency permits high-performance solutions using inexpensive technology. The Reassembler, in particular, makes great use of concurrency by setting up a pipeline that has four logical stages. Preferred architectures are illustrated in Figures 7 and 8.
Figure 7 is a block diagram of a Segmenter 10' modified in accordance with the present invention to include support for encryption. This support is provided by an encryption unit 40, which may comprise a VM007 data encryption processor. Figure 8 is a block diagram of a Reassembler 30' modified to include support for decryption. Decryption support is provided by a decryption unit 42, which may also employ the VM007 processor. Blocks 40 and 42 contain logic circuitry for controlling key latching and
loading. Logic circuitry of this type will be known to those skilled in the art after reading this specification, therefore a detailed description is unnecessary.
The addition of the encryption/decryption chip 40, 42 in each case serves to add another functional block that can operate concurrently with the other blocks. Key lookup is done by adding an extra field to the "head-of- list" data structure in the pointer memory. These "head- of-list" structures are located using the CAM with the Virtual-Circuit Identifier as a tag. The cryptographic keys are loaded into the VM007 in parallel with data movement from the cell manager's data buffer into the dual port reassembly buffer. This solution provides the following significant advantages: 1. Private keys for each virtual circuit (256
Virtual Circuits are currently supported in the host interface) . 2. Real-time "agile" keying across Virtual Circuits. 3. Concurrent cleartext and ciphertext streams.
4. Application-specific control of cryptographic keying.
5. Cleartext headers and ciphertext payloads (which ease transport through packet- switched fabrics) .
Unless workstation architectures change greatly, cryptographic hardware can advantageously be incorporated into the host/network interface. This adds a marginal amount of complexity to the interface and comfortably meshes with functions the interface already performs, such as cellification, decellification, and computing checksums. The advantages to this approach stem from the fact that neither host memory access nor bus transactions are required beyond those transactions necessary to move data from the application data areas to the host/network interface. Key management is closely correlated with ATM header manipulation, and can be efficiently performed with
the addition of a small amount of logic in the host interface.
While the DES algorithm is used in preferred embodiments to prototype experimental solutions, the architectural proposals described herein are not dependent on its use. These architectural solutions relate to the placement of cryptographic hardware in a workstation connected to a high-bandwidth network. These solutions do not stand in the way of application-to-application ("end- to-end") privacy, under the assumption that the workstation operating system is a "trusted" subsystem. Distinct keys can be loaded for each ATM virtual circuit (i.e., each application) .
Claims
1. An interface for coupling a host computer to a high-speed telecommunications network, comprising reassembler means for reassembling and decrypting data that has been encrypted in accordance with a pre-defined key and segmented into a plurality of asynchronous transfer mode (ATM) cells, each cell comprising a virtual channel identifier (VCI) and a cell body, said reassembler means comprising: (a) cell manager means for separating each cell body from that cell body's corresponding VCI;
(b) linked list manager means for storing linked list data indicative of addresses at which said cell bodies are to be stored; (c) content addressable memory (CAM) means for storing said VCIs and providing pointers into said linked list data;
(d) lookup controller means for writing said VCIs to said CAM means; (e) reassembly buffer means for storing said cell bodies; and
(f) decryption means for decrypting said data in accordance with said key and writing the decrypted data to said reassembly buffer means.
2. An interface as recited in claim 1, further comprising segmenter means for encrypting data received from a host computer and segmenting the encrypted data into a plurality of ATM cells.
3. An interface for coupling a host computer to a high-speed telecommunications network, comprising:
(a) means for providing private keys for each of a plurality of virtual circuits;
(b) means for providing real-time agile keying across said virtual circuits;
(c) means for providing concurrent cleartext and ciphertext streams;
(d) means for providing application-specific control of cryptographic keying; and (e) means for providing cleartext headers and ciphertext payloads.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US90016392A | 1992-06-17 | 1992-06-17 | |
US07/900,163 | 1992-06-17 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO1993026109A1 true WO1993026109A1 (en) | 1993-12-23 |
Family
ID=25412059
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/US1993/005462 WO1993026109A1 (en) | 1992-06-17 | 1993-06-02 | Apparatus for providing cryptographic support in a network |
Country Status (2)
Country | Link |
---|---|
US (1) | US5329623A (en) |
WO (1) | WO1993026109A1 (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0676879A2 (en) * | 1994-03-08 | 1995-10-11 | Advanced Micro Devices, Inc. | Packet management device for fast-packet network |
WO1995028782A1 (en) * | 1994-04-18 | 1995-10-26 | Apple Computer, Inc. | Method and apparatus for decoding non-sequential data packets |
AU749946B2 (en) * | 1997-10-28 | 2002-07-04 | University Of Melbourne, The | Stabilized particles and methods of preparation and use thereof |
Families Citing this family (59)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5495580A (en) * | 1992-10-20 | 1996-02-27 | Xlnt Designs, Inc. | Ring network security system with encoding of data entering a subnetwork and decoding of data leaving a subnetwork |
US6026443A (en) * | 1992-12-22 | 2000-02-15 | Sun Microsystems, Inc. | Multi-virtual DMA channels, multi-bandwidth groups, host based cellification and reassembly, and asynchronous transfer mode network interface |
US5802287A (en) * | 1993-10-20 | 1998-09-01 | Lsi Logic Corporation | Single chip universal protocol multi-function ATM network interface |
IL111154A0 (en) * | 1993-10-21 | 1994-12-29 | Martino Ii John A | Systems and methods for electronic messaging |
WO1995014269A1 (en) * | 1993-11-19 | 1995-05-26 | The Trustees Of The University Of Pennsylvania | A high-performance host interface for networks carrying connectionless traffic |
WO1997000471A2 (en) * | 1993-12-15 | 1997-01-03 | Check Point Software Technologies Ltd. | A system for securing the flow of and selectively modifying packets in a computer network |
US5835726A (en) * | 1993-12-15 | 1998-11-10 | Check Point Software Technologies Ltd. | System for securing the flow of and selectively modifying packets in a computer network |
US5553302A (en) * | 1993-12-30 | 1996-09-03 | Unisys Corporation | Serial I/O channel having independent and asynchronous facilities with sequence recognition, frame recognition, and frame receiving mechanism for receiving control and user defined data |
US6185619B1 (en) | 1996-12-09 | 2001-02-06 | Genuity Inc. | Method and apparatus for balancing the process load on network servers according to network and serve based policies |
US5548721A (en) * | 1994-04-28 | 1996-08-20 | Harris Corporation | Method of conducting secure operations on an uncontrolled network |
JP3172387B2 (en) * | 1994-06-01 | 2001-06-04 | インターナショナル・ビジネス・マシーンズ・コーポレ−ション | I / O communication subsystem and method |
US5864683A (en) * | 1994-10-12 | 1999-01-26 | Secure Computing Corporartion | System for providing secure internetwork by connecting type enforcing secure computers to external network for limiting access to data based on user and process access rights |
US5680401A (en) * | 1995-10-27 | 1997-10-21 | Sun Microsystems, Inc. | Method and apparatus for asynchronously segmenting packets of multiple channels into ATM cells |
US5793983A (en) * | 1996-01-22 | 1998-08-11 | International Business Machines Corp. | Input/output channel interface which automatically deallocates failed subchannel and re-segments data block for transmitting over a reassigned subchannel |
US5913024A (en) * | 1996-02-09 | 1999-06-15 | Secure Computing Corporation | Secure server utilizing separate protocol stacks |
US5918018A (en) * | 1996-02-09 | 1999-06-29 | Secure Computing Corporation | System and method for achieving network separation |
US5768526A (en) * | 1996-03-08 | 1998-06-16 | Glenayre Electronics, Inc. | Method and apparatus for validating data packets in a paging system |
US5854891A (en) * | 1996-08-09 | 1998-12-29 | Tritheim Technologies, Inc. | Smart card reader having multiple data enabling storage compartments |
US5831980A (en) * | 1996-09-13 | 1998-11-03 | Lsi Logic Corporation | Shared memory fabric architecture for very high speed ATM switches |
US6003084A (en) * | 1996-09-13 | 1999-12-14 | Secure Computing Corporation | Secure network proxy for connecting entities |
US5959993A (en) * | 1996-09-13 | 1999-09-28 | Lsi Logic Corporation | Scheduler design for ATM switches, and its implementation in a distributed shared memory architecture |
US5950195A (en) * | 1996-09-18 | 1999-09-07 | Secure Computing Corporation | Generalized security policy management system and method |
US5983350A (en) * | 1996-09-18 | 1999-11-09 | Secure Computing Corporation | Secure firewall supporting different levels of authentication based on address or encryption status |
US5915087A (en) * | 1996-12-12 | 1999-06-22 | Secure Computing Corporation | Transparent security proxy for unreliable message exchange protocols |
US6357010B1 (en) | 1998-02-17 | 2002-03-12 | Secure Computing Corporation | System and method for controlling access to documents stored on an internal network |
US6321336B1 (en) | 1998-03-13 | 2001-11-20 | Secure Computing Corporation | System and method for redirecting network traffic to provide secure communication |
US6453419B1 (en) | 1998-03-18 | 2002-09-17 | Secure Computing Corporation | System and method for implementing a security policy |
US6182226B1 (en) | 1998-03-18 | 2001-01-30 | Secure Computing Corporation | System and method for controlling interactions between networks |
JP4097773B2 (en) * | 1998-04-20 | 2008-06-11 | オリンパス株式会社 | Digital image editing system |
EA003744B1 (en) * | 1998-11-18 | 2003-08-28 | Сага Софтвеар, Инк. | Extensible distributed enterprise application integration system |
US6765911B1 (en) * | 1999-02-03 | 2004-07-20 | International Business Machines Corporation | Communications adapter for implementing communications in a network and providing multiple modes of communications |
US7996670B1 (en) * | 1999-07-08 | 2011-08-09 | Broadcom Corporation | Classification engine in a cryptography acceleration chip |
US6882642B1 (en) | 1999-10-14 | 2005-04-19 | Nokia, Inc. | Method and apparatus for input rate regulation associated with a packet processing pipeline |
US6757249B1 (en) * | 1999-10-14 | 2004-06-29 | Nokia Inc. | Method and apparatus for output rate regulation and control associated with a packet pipeline |
US6934250B1 (en) | 1999-10-14 | 2005-08-23 | Nokia, Inc. | Method and apparatus for an output packet organizer |
US7032240B1 (en) * | 1999-12-07 | 2006-04-18 | Pace Anti-Piracy, Inc. | Portable authorization device for authorizing use of protected information and associated method |
US6601170B1 (en) | 1999-12-30 | 2003-07-29 | Clyde Riley Wallace, Jr. | Secure internet user state creation method and system with user supplied key and seeding |
US6704794B1 (en) * | 2000-03-03 | 2004-03-09 | Nokia Intelligent Edge Routers Inc. | Cell reassembly for packet based networks |
US6691143B2 (en) | 2000-05-11 | 2004-02-10 | Cyberguard Corporation | Accelerated montgomery multiplication using plural multipliers |
US20040073617A1 (en) | 2000-06-19 | 2004-04-15 | Milliken Walter Clark | Hash-based systems and methods for detecting and preventing transmission of unwanted e-mail |
JP2002044135A (en) * | 2000-07-25 | 2002-02-08 | Mitsubishi Electric Corp | Encryption device and encryption communication system |
GB0023409D0 (en) * | 2000-09-22 | 2000-11-08 | Integrated Silicon Systems Ltd | Data encryption apparatus |
US7096503B1 (en) * | 2001-06-29 | 2006-08-22 | Mcafee, Inc. | Network-based risk-assessment tool for remotely detecting local computer vulnerabilities |
US20030142818A1 (en) * | 2001-09-28 | 2003-07-31 | Nec Usa, Inc. | Techniques for efficient security processing |
US20030081772A1 (en) * | 2001-10-30 | 2003-05-01 | Blaker David M. | Parallel random number determinations for a stream cipher utilizing a common S-box |
US8189591B2 (en) * | 2001-10-30 | 2012-05-29 | Exar Corporation | Methods, systems and computer program products for packet ordering for parallel packet transform processing |
US20030196081A1 (en) * | 2002-04-11 | 2003-10-16 | Raymond Savarda | Methods, systems, and computer program products for processing a packet-object using multiple pipelined processing modules |
KR100446533B1 (en) * | 2002-10-08 | 2004-09-01 | 삼성전자주식회사 | Apparatus and method for ciphering in mobile communication system |
US7434043B2 (en) | 2002-12-18 | 2008-10-07 | Broadcom Corporation | Cryptography accelerator data routing unit |
US20040123120A1 (en) * | 2002-12-18 | 2004-06-24 | Broadcom Corporation | Cryptography accelerator input interface data handling |
US20040123123A1 (en) * | 2002-12-18 | 2004-06-24 | Buer Mark L. | Methods and apparatus for accessing security association information in a cryptography accelerator |
US7568110B2 (en) | 2002-12-18 | 2009-07-28 | Broadcom Corporation | Cryptography accelerator interface decoupling from cryptography processing cores |
US20040151170A1 (en) * | 2003-01-31 | 2004-08-05 | Manu Gulati | Management of received data within host device using linked lists |
US7885405B1 (en) * | 2004-06-04 | 2011-02-08 | GlobalFoundries, Inc. | Multi-gigabit per second concurrent encryption in block cipher modes |
TWI286897B (en) * | 2004-11-09 | 2007-09-11 | Via Tech Inc | Real time decryption system and method |
US20060136717A1 (en) | 2004-12-20 | 2006-06-22 | Mark Buer | System and method for authentication via a proximate device |
US8295484B2 (en) | 2004-12-21 | 2012-10-23 | Broadcom Corporation | System and method for securing data from a remote input device |
US8094819B1 (en) | 2006-07-28 | 2012-01-10 | Rockwell Collins, Inc. | Method and apparatus for high agility cryptographic key manager |
US10133883B2 (en) * | 2009-02-09 | 2018-11-20 | International Business Machines Corporation | Rapid safeguarding of NVS data during power loss event |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5081678A (en) * | 1989-06-28 | 1992-01-14 | Digital Equipment Corporation | Method for utilizing an encrypted key as a key identifier in a data packet in a computer network |
US5153920A (en) * | 1990-08-10 | 1992-10-06 | Siemens Aktiengesellschaft | Method and circuit arrangement for the acceptance and forwarding of message cells transmitted according to an asynchronous transfer mode by an asynchronous transfer mode switching equipment |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5253294A (en) * | 1983-02-22 | 1993-10-12 | At&T Bell Laboratories | Secure transmission system |
US5012466A (en) * | 1987-10-13 | 1991-04-30 | At&T Bell Laboratories | Call service initialization arrangement |
US5144622A (en) * | 1988-02-15 | 1992-09-01 | Hitachi, Ltd. | Network system |
US4975906A (en) * | 1988-02-15 | 1990-12-04 | Hitachi, Ltd. | Network system |
JP2760799B2 (en) * | 1988-04-28 | 1998-06-04 | 株式会社日立製作所 | Encryption method |
US5101404A (en) * | 1988-08-26 | 1992-03-31 | Hitachi, Ltd. | Signalling apparatus for use in an ATM switching system |
US5136584A (en) * | 1990-07-11 | 1992-08-04 | At&T Bell Laboratories | Hardware interface to a high-speed multiplexed link |
US5130984A (en) * | 1990-12-18 | 1992-07-14 | Bell Communications Research, Inc. | Large fault tolerant packet switch particularly suited for asynchronous transfer mode (ATM) communication |
US5214698A (en) * | 1991-03-20 | 1993-05-25 | International Business Machines Corporation | Method and apparatus for validating entry of cryptographic keys |
US5220606A (en) * | 1992-02-10 | 1993-06-15 | Harold Greenberg | Cryptographic system and method |
-
1993
- 1993-06-02 WO PCT/US1993/005462 patent/WO1993026109A1/en active Application Filing
- 1993-11-12 US US08/152,085 patent/US5329623A/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5081678A (en) * | 1989-06-28 | 1992-01-14 | Digital Equipment Corporation | Method for utilizing an encrypted key as a key identifier in a data packet in a computer network |
US5153920A (en) * | 1990-08-10 | 1992-10-06 | Siemens Aktiengesellschaft | Method and circuit arrangement for the acceptance and forwarding of message cells transmitted according to an asynchronous transfer mode by an asynchronous transfer mode switching equipment |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0676879A2 (en) * | 1994-03-08 | 1995-10-11 | Advanced Micro Devices, Inc. | Packet management device for fast-packet network |
EP0676879A3 (en) * | 1994-03-08 | 2002-11-27 | Advanced Micro Devices, Inc. | Packet management device for fast-packet network |
WO1995028782A1 (en) * | 1994-04-18 | 1995-10-26 | Apple Computer, Inc. | Method and apparatus for decoding non-sequential data packets |
US5541995A (en) * | 1994-04-18 | 1996-07-30 | Apple Computer Inc. | Method and apparatus for decoding non-sequential data packets |
AU749946B2 (en) * | 1997-10-28 | 2002-07-04 | University Of Melbourne, The | Stabilized particles and methods of preparation and use thereof |
Also Published As
Publication number | Publication date |
---|---|
US5329623A (en) | 1994-07-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5329623A (en) | Apparatus for providing cryptographic support in a network | |
US5796836A (en) | Scalable key agile cryptography | |
Kim et al. | Design and implementation of a private and public key crypto processor and its application to a security system | |
KR100908765B1 (en) | Packet Encryption System and Method | |
US7362859B1 (en) | Enhancement of utilization of encryption engine | |
US7233970B2 (en) | Computational method, system, and apparatus | |
US7218734B2 (en) | Ring arithmetic method, system, and apparatus | |
US20030142818A1 (en) | Techniques for efficient security processing | |
Samir et al. | ASIC and FPGA comparative study for IoT lightweight hardware security algorithms | |
US7653196B2 (en) | Apparatus and method for performing RC4 ciphering | |
Curiger et al. | VINCI: VLSI implementation of the new secret-key block cipher IDEA | |
Nahum et al. | Towards high performance cryptographic software | |
Salomao et al. | Hipcrypto: A high-performance VLSI cryptographic chip | |
Landau | Technical opinion: designing cryptography for the new century | |
Smith et al. | Cryptographic Support in a Gigabit Network | |
CN114679252A (en) | Resource sharing method for MACsec AES algorithm | |
WO1999014979A1 (en) | Cryptographic system for public atm/sonet communication system with virtual circuit lookup and pipelined data encryption and decryption | |
Haskins | Securing asynchronous transfer mode networks. | |
US8024392B2 (en) | Computational method, system, and apparatus | |
Elbirt et al. | Efficient implementation of Galois field fixed field constant multiplication | |
Lai et al. | A novel VLSI architecture for a variable-length key, 64-bit Blowfish block cipher | |
Broscius | Exploiting Parallelism in Hardware Implementation of the DES Albert G. Broscius Distributed Systems Lab Dept. of CIS | |
Lee et al. | High‐Speed Hardware Architectures for ARIA with Composite Field Arithmetic and Area‐Throughput Trade‐Offs | |
Tarman et al. | Final report for the robustness-agile asynchronous transfer mode (ATM) encryption laboratory directed research and development project | |
Carlson | OPNET performance simulation of Network Security Services |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): CA JP |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
DFPE | Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101) | ||
122 | Ep: pct application non-entry in european phase | ||
NENP | Non-entry into the national phase |
Ref country code: CA |