CN1640034B

CN1640034B - Distributed radio system with multiple transceivers for simulcasting and selective processing of received signals

Info

Publication number: CN1640034B
Application number: CN038045532A
Authority: CN
Inventors: 玛丽·杰茜; 贾里·海诺宁; 埃利奥特·胡尔; 布莱恩·施拉戴特斯基; 格雷格·文蒂米拉
Original assignee: RadioFrame Networks Inc
Current assignee: RadioFrame Networks Inc
Priority date: 2002-02-25
Filing date: 2003-02-24
Publication date: 2010-09-22
Anticipated expiration: 2023-02-24
Also published as: EP1479182A4; HK1080233B; CN1640034A; AU2003217751A1; HK1080233A1; KR100935678B1; KR20040102015A; EP1479182A1

Abstract

A distributed radio system with multiple transceivers for simulcasting and selective processing of received signals is provided. The distributed radio system includes a plurality of processing elements and radio frequency transmitter elements interconnected by an Ethernet network. The system designates a number of radio frequency transmitter elements to be elements of a radio frequency simulcast set. The system utilizes relatively few frequency channels, and transmits at very low power levels, and thus causes minimal interference with an existing macrocellular environment. When signals transmitted from a mobile unit are detected by multiple radio receivers, the system is able to select a desired radio receiver signal for processing. When selecting a received signal for processing, the system utilizes a distributed processing technique that performs the selection process throughout several levels. The distributed processing utilized by the selection process may be synchronized. A selection time window may be utilized for making the selection process for the upstream-received data traffic. The system utilizes a centralized data link layer. Transmissions are implemented by a two-level multicast technique. The data transmission traffic is bundled. Selective management of the Ethernet switches is performed by the system.

Description

Has the distributed radio system that is used to broadcast with the multi-section transceiver of selective processing of received signals

Technical field

The present invention relates generally to distributed radio system, relate in particular to and have the indoor heating system that is used to broadcast with the multi-section transceiver of selective processing of received signals.

Background technology

Wireless (for example honeycomb) operator is being faced with lasting challenge aspect the overlay area of their system of expansion, is restricted at aspects such as cost, power and frequencies simultaneously.The zone that expansion covers is in indoor environment.Yet,, wish that still such system does not disturb existing macrocellular environment even on the similar frequency bands that is used for a given cellular standards, work.

The present invention aims to provide a kind of system and method, and they can be with low-power level work, and existing macrocellular environment is produced minimum interference.More particularly, the present invention aims to provide a kind of system and method, wherein, utilizes the multi-section transceiver in an indoor heating system, and by utilizing the radio hookup technology can be with low-power level to a mobile unit effective signal radiated.

Summary of the invention

The invention discloses a kind of have be used for by radio hookup and mobile unit communication the distributed radio system of the multi-section transceiver of selective processing of received signals simultaneously.According to an aspect of the present invention, native system is with low-down power level.Because use low-power level, native system can not disturb existing macrocellular environment.

According to another aspect of the present invention, the relatively less channel of native system utilization.Native system can provide large-area geographical the covering with less relatively channel, simultaneously by using a limited number of channel, from being distributed in several the different identical radiofrequency signals of radio transmitter radio hookup of a building building or zone similarity, launch with low-power level.This makes native system can cover relatively large inside plant.And as mentioned above, even be used for the similar frequency bands of a given cellular standards, native system is still very little to the interference of existing macrocellular environment, and native system can be with low-down power level work by utilizing the radio hookup technology.

According to another aspect of the present invention, in one embodiment, distributed radio system comprises a plurality of processing unit and a plurality of radio frequency sending set parts by Ethernet (for example, IEEE 802) interconnection.This distributed radio system uses at least two radio frequency sending set parts, on a common radio-frequency carrier wave, broadcasts a common modulation signal.One group of radio frequency sending set parts of a public modulated radio signal of radio hookup are designated as a radio frequency radio hookup group.At work, each parts of a radio frequency radio hookup group are configured to receive the Ethernet sampled signal packets and send the information that contains by the radio-frequency carrier of modulating them in each grouping.

According to another aspect of the present invention, in one embodiment, this simulcast method is appointed as a plurality of specific radio frequency sending set parts the parts of radio frequency radio hookup group.First group of Ethernet grouping that transmission will be received by each parts of radio frequency radio hookup group, this first group of Ethernet grouping is used to a multicast address each parts of radio frequency radio hookup group that go to programme, then, each unit response of radio frequency radio hookup group is in the Ethernet grouping that contains this multicast address, as medium access control (MAC) address of appointment.Second group of Ethernet grouping that transmission will be received by each parts of radio frequency radio hookup group, this second group of Ethernet grouping be used to programme each parts of radio frequency radio hookup group make it use a specific radio-frequency carrier to carry out work.In addition, periodically send the Ethernet sampled signal packets.The Ethernet sampled signal packets has an assigned address, and it is equivalent to the multicast address of programming.Each parts of radio frequency radio hookup group are configured to receive this Ethernet sampled signal packets.Each parts of radio frequency radio hookup group are modulated the common radio-frequency carrier wave according to the sample data that is comprised in the Ethernet sampled signal packets.

According to another aspect of the present invention, the processing unit of system comprises a CPU and a plurality of airlink processing unit at least.CPU is responsible for setting up being connected between this system and the external environment condition (for example macrocellular environment), or being connected between this system and the public switch telephone network, also is responsible for the network management of whole system.In one embodiment, CPU can be a network chassis unit.CPU is connected to several airlink processing units by Ethernet.Each airlink processing unit also all is connected to several radio sets by Ethernet.In one embodiment, the airlink processing unit can be each airlink chassis units.

According to another aspect of the present invention, data link layer can put together.For transfer of data is arrived each mobile unit, importantly identical by the data of several transmitters radio hookups.If each simulcasting radio frequency transmitter components do not launch identical signal, then may cause common-channel interference, the degradation of signals that makes mobile unit thus and received.For the data that guarantee to launch are identical, the second layer (data link layer) is concentrated in CPU.The grouping of second layer Ethernet is sent to several airlink processing units by CPU, and the latter becomes waveform with second layer data processing, and these waveforms are at last by each radio hookup radio transmitter emission.

According to another aspect of the present invention, realize emission by a kind of double-deck multicasting technology, in utilizing an embodiment of this technology, the emission data are as each Ethernet grouping of multicast, send to a plurality of airlink processing units by CPU with second layer form, thereafter, the emission data also as the Ethernet grouping of multicast, send to relevant multi-section radio set by each airlink processing unit with ground floor form (sample waveform).

According to another aspect of the present invention, when the signal from a mobile unit emission was detected by the multi-section radio receiver, native system can select the radio receiver of an expection for processing.The multi-section radio receiver provides the data that detect to system simultaneously.The data that detect comprise identical information, but are in different quality level.Native system can select the signal of an expection for processing.

According to another aspect of the present invention, when selecting a received signal to be used to handle, utilize distributed proccessing to carry out a kind of gradual selection course.A given mobile device can be placed with quite near several possible radio receivers.Wish to select and only handle the signal that provides by receiver with the strongest signal from this mobile unit.Carry out trade-off decision by a process, this process can be seen all the relevant receivers in the system.

According to another aspect of the present invention, in a preferred embodiment, system interconnects via having band-limited ethernet link.In the trade-off decision process, this configuration can make it be difficult to transmit data from each receiver usually, will exceed network capacity soon because do like this.Utilize a kind of distributed proccessing to realize gradual selection exactly by solution provided by the invention.More particularly, selection course can be carried out on some different levels, is included in CPU and at each airlink processing unit.Last selection course is positioned at CPU, and it is in the top layer of system hierarchy.Yet each airlink processing unit only will be selected the signal that receives from a radio set, and only transmit this signal for final decision to CPU.So, simulcast receivers is handled and is distributed in whole system, and, along with received signal is soaring step by step in hierarchical structure, little by little select the signal of optimum reception.

According to another aspect of the present invention, selection course can be distributed more or less, and this depends on the available data rate of system.For example, one use global system for mobile communications (GSM) technology, have a system of quite high data transfer rate, can allow CPU to have time enough to remove to receive signal, as the part of selection course from the some radio sets in the given airlink processing unit.A kind of additional technology that is used to screen signal may relate to the minimum signal strength signal intensity of requirement, and this signal strength signal intensity is detected in radio set itself, so that upwards be forwarded to corresponding airlink processing unit.

According to another aspect of the present invention, the employed distributed treatment of selection course can realize synchronously.In one embodiment, the signal of wishing to receive from each radio set is basically all corresponding to the identical moment, so that correctly assess the difference between the received signal.Wish that CPU can be synchronized with airlink processing unit and radio set, make CPU know and look to when arriving, and in selection course, should when make decisions from the received signal of airlink processing unit.Preferably, realize synchronously at the end points (that is, CPU and radio set) of system at least.Yet, if the processing procedure in each airlink processing unit can realize that also system delay can also minimize further synchronously.

According to another aspect of the present invention, in order to set up the selection course that the upstream receives data service, a select time window can be set.So, no matter whether the received signal data of all expections have arrived CPU, in the regular hour window, by the decision-making that makes one's options of the selection course in the CPU.In general, as the part of this method, base when CPU has, it is synchronized with each airlink processing unit and each radio set.

According to another aspect of the present invention, data service can be tied.Tied from the immigration business that each radio set receives, so that do not flood CPU, the latter serves the task of sending the Ethernet grouping.Concerning the situation of a plurality of radio-frequency channels, can periodically send a plurality of Ethernet groupings (each Ethernet grouping contains one group of received signal of having selected respectively) to CPU from each airlink processing unit, rather than be that each bar radio-frequency channel sends other Ethernet grouping, data are bundled into one or more Ethernet groupings, therefore, reduced to the grouping of a plurality of Ethernets of transmission and served required expense.Departure business by each radio set emission also can be tied, and feasible information from many radio-frequency channels is set as the individual ethernet grouping.

According to another aspect of the present invention, native system can be carried out optionally management to Ethernet switch.In certain zone of native system, the ethernet link between each radio set and each airlink processing unit for example, the data transmission service amount may be very high, because the original sample of many such each waveforms of business agent.Many such Ethernet groupings also are multicastapackets, and they can be flooded in whole network usually.Yet, in certain embodiments,, for example, may be tied down by these business or overwhelm, so this may be undesirable between being connected between CPU and each airlink processing unit because other networks in the system connect.In big system, people wish Ethernet switch is configured to and can filters or rejection the transmission of the designated packet on the port of appointment.

To understand, the advantage of disclosed system and method is, they can be used to provide large-area geographical (for example the covering, in an indoor heating system), and shared channel is less relatively, and relatively low to the power requirement of transmission, simultaneously, even work on the similar frequency bands that is used for a kind of given cellular standards, they also can make the minimum interference to existing macrocellular environment.

Description of drawings

By with reference to following detailed description and in conjunction with all accompanying drawings, will make above-mentioned various aspects and many bonus of the present invention become easier to understand, in all accompanying drawings:

Fig. 1 is a kind of block diagram of system configuration;

Fig. 2 is the data flow diagram of expression by a non-simulcasting system;

Fig. 3 is the block diagram of a pseudo-simulcasting system;

Fig. 4 be expression be used for a single sub-district, three sector systems use pattern again;

Fig. 5 is the block diagram of a simulcasting system, wherein, carries out the second layer and handle on network chassis unit, and carry out the ground floor processing on the link rack unit aloft;

Fig. 6 is the data flow diagram of expression by the system of Fig. 5;

Fig. 7 is the block diagram of a simulcasting system, wherein, carry out the second layer and handle on network chassis unit, and the ground floor processing is cut apart between network chassis unit and airlink chassis units;

Fig. 8 is the data flow diagram of expression by the system of Fig. 7;

Fig. 9 is the block diagram of a simulcasting system, wherein, carries out the ground floor and the second layer and handle on network chassis unit;

Figure 10 is the initial message figure that expression is used for different scripts;

Figure 11 is the Event Sequence Diagram that expression occurs as a part of inserting the process of carrying out for each radio blade (radio blade);

Figure 12 is the schematic diagram that channel of expression is provided with circular buffer;

Figure 13 is the flow chart of a message router task;

Figure 14 is the flow chart of a departure (outbound) information handling task; And

Figure 15 is the flow chart of an immigration (inbound) information handling task;

Embodiment

Fig. 1 represents system configuration.A network rack (chassis) unit NCU is used as a CPU, and is connected to airlink chassis units ACU-X, ACU-Y and ACU-Z by ethernet link.Network chassis unit NCU is responsible for setting up the interface that this system leads to external environment condition (for example macrocell system or public switch telephone network), also is responsible for the network management of whole system.

Network chassis unit NCU comprises network processing card NPC-A, NPC-B and NPC-C.In fact network processing card NPC-A, NPC-B and NPC-C can be the airlink transaction cards, and this will be described in detail below.Network chassis unit NCU also comprises an Ethernet switch ES, and it is connected to a microprocessor PROC.In one embodiment, this Ethernet switch ES may be a wide area (broadcom) Ethernet switch, and this microprocessor PROC may be one 8240 microprocessor.This Ethernet switch ES is connected to an integrated site controller ISC by an ethernet link.When an integrated site controller is shown, should be appreciated that the exchange control unit that can replace other types.In one embodiment, the communication between this Ethernet switch ES and integrated site controller ISC is based on 15ms the 3rd layer of message/VSELP grouping regularly.Integrated site controller ISC comprises an access control gateway A CG, and the latter comprises again at sector A, the processing of B and C, and this will be described in detail below.

Each airlink chassis units ACU-X, ACU-Y and ACU-Z comprise one group of airlink transaction card APC-A, APC-B and APC-C.Each airlink chassis units also comprises an Ethernet switch ES, and the latter is connected to a microprocessor PROC and each airlink transaction card APC, and each APC comprises a microprocessor PROC again.In one embodiment, in the configuration of Fig. 1 and in other following configurations, the microprocessor PROC of each rack unit ACU or NCU can be 8240 microprocessors, and simultaneously, the microprocessor among each transaction card APC or the NPC can be 8260 microprocessors.

Communication between each airlink transaction card APC and Ethernet switch ES comprises second layer PDU/IQ grouping.Ethernet switch ES is connected to the Ethernet switch ES of network chassis unit NCU by ethernet link.The Ethernet switch ES of airlink chassis units ACU is connected to a series of radio frequency unit RFU-A, RFU-B and RFU-C further by ethernet link.Communication between Ethernet switch ES and each radio frequency unit RFU is IQ grouping (7.5ms).Each radio frequency unit RFU comprises the radio blade RB (not shown) in order to transmission signals.

As will describing in more detail below, the basic demand of simulcast method of the present invention is, is each sector A from each radio blade RB simultaneously, B and the identical data of C emission.In order to realize this requirement, people wish to allow and are in each radio frequency unit RFU and are synchronized to together with all interior radio blade RB, and are sent to (being in identical sector) all radio blade RB at the protocol Data Unit PDU of a particular time-slot in same time slot.

In order to guarantee the data of identical protocol Data Unit PDU, related control program ACP, each tasks such as RAP RAP and voice channel procedure VCP all are placed among the network chassis unit NCU, rather than are placed among the airlink chassis units ACU.This is shown in each network processing card NPC-A among Fig. 1 to NPC-C, and so, just the data that send to the protocol Data Unit PDU of all airlink transaction card APC for needs provide a common growing point.Be used to send outside each grouping of protocol Data Unit PDU by broadcasting the address, and guarantee the slot synchronization between each airlink transaction card APC, distribute when just can produce identical data.

In U.S. Provisional Patent Application series number the 60/359th, in No. 638, also narrated System and method for of the present invention, the application's book is advocated priority from above-mentioned temporary patent application, and above-mentioned temporary patent application form in full is by income this paper, as a reference.In U.S. Provisional Patent Application series number the 60/359th, No. 637 (this temporary patent application is commonly assigned, and its full content is at this as a reference) in, and RFNI-1-18802 number, the application that is entitled as " have distributed real-time handle radio system " (are transferred the possession of and co-applications with the application is common to act on behalf of files, and its full content is at this as a reference) in, a relevant system all narrated.

Fig. 2 is the data flow diagram of expression by a non-simulcasting system.This system is shown is purpose, provide explanation preferably to simulcast method of the present invention simultaneously in order to illustrate.Non-simulcast method shown in Figure 2 is an embodiment based on an airlink transaction card APC (as shown in Figure 1).In general, each digital signal processor and processing are all closely linked together via host interface HPI from the each task of the data of the protocol Data Unit PDU of each digital signal processor.As shown in Figure 2, the task of the data of processing protocol data cell PDU is exactly related control program ACP, RAP RAP and voice channel procedure VCP.By message queue MQ data are sent to related control program ACP and RAP RAP from a message router MR.By a function F UNC data are sent to voice channel procedure VCP.Data arrive message router MR by formation Q from a digital signal processor reader task RT and a digital signal processor write device task WT.Data arrive reader task RT and write device task WT from host interface HPI.Data more particularly, arrive host interface HPI from parts DSP-Tx and parts DSP-Rx from digital signal processor.Data arrive parts DSP-Tx and parts DSP-Rx from parts FPGA, above-mentioned FPGA to/provide/receive original I/Q from radio blade.

With reference to Fig. 2, data transmit beginning based on so-called time slot request.The time slot request is the request that digital signal processor proposes in order to send a new protocol Data Unit PDU in next time slot.In one embodiment, every 15ms just receives a time slot request by digital signal processor reader task from digital signal processor parts 22.Arrival from an I/Q sample packet of radio blade triggers this time slot request event.Then, digital signal processor reader task RT sends a piece of news to message router MR, and this message call appropriate functional is obtained protocol Data Unit PDU.Take out protocol Data Unit PDU via funcall, and send via message queue.Then, message router MR sends to digital signal processor write device task WT with protocol Data Unit PDU, and the latter is written into digital signal processor.In one embodiment, the time between twice time slot request is 15ms, and, there is the time of about 14ms to be used for the receiving slot request message, respond to digital signal processor with a protocol Data Unit PDU simultaneously.Digital signal processor is converted to an I/Q sample packet with protocol Data Unit PDU, then it is sent to the usefulness of radio blade RB for radio-frequency transmissions.

Except the time slot request, also has the departure protocol Data Unit PDU that is sent to integrated site controller ISC from digital signal processor, handover message, the message of immigration integrated site controller ISC, every periodic monitoring incident, and the UDP configuration messages, these all are to wish usually to handle in the time of this 14ms.Many these additional incidents all are nonsynchronous, and they can occur at any time and in any combination.In the system of Fig. 2, also has an Ethernet switch (not shown) usually, in order to the suitable route of each Ethernet grouping to be provided.

In one embodiment of the invention, one of design object is exactly the restriction that obtains each 3 control channel frequency of system, and this may make C/I be difficult to control with technology via the honeycomb of standard again.When the symbol that is sending as a given radio blade RB did not match with other sub-districts sent, this just became a problem.If can not in the whole coverage of this system, concentrate coordination, then there is a kind of like this possibility to the special symbol of bearer traffic and control channel message.With reference to Fig. 3 (illustrate a pseudo-simulcasting system) and with reference to Fig. 5-9 (signal is according to true simulcasting system of the present invention), these problems are discussed below.

Fig. 3 represents a pseudo-simulcasting system, wherein, as shown in the figure, produces second layer message in each airlink chassis units ACU.Fig. 3 has each parts that is similar to Fig. 1.As shown in Figure 3, concerning each airlink chassis units ACU, the transmission from each airlink transaction card APC to Ethernet switch ES comprises the 3rd layer of message/VSELP grouping/IQ grouping.Transmission in each airlink transaction card APC comprises second layer message/VSELP grouping, (it is controlling RAP RAP from microprocessor PROC second layer parts for they, related control program ACP and voice control program VCP) be sent to be used for SPAM L2 (FEC, MAC), L1.Comprise the 3rd layer of message/VSELP divide into groups (15ms) from the transmission of the Ethernet switch ES of the Ethernet switch ES of network chassis unit NCU and airlink chassis units ACU, they are the identical message of type of message that is transmitted between the Ethernet switch ES that follows at network chassis unit NCU and the integrated site controller ISC.

The distributed second layer of Fig. 3 is handled and has been simplified network design, but has opened up the possibility that the different grouping of second layer symbol enters each frame of airlink.In this script, each the radio frequency unit RFU that sends the symbol be different from the symbol that is sent from the radio frequency unit RFU that serves this portable terminal will generate the common-channel interference at random at this portable terminal effectively.In true simulcasting system, second layer message is provided with by the concentrated area, and be assigned to each airlink chassis units ACU (being assigned to each radio frequency unit RFU subsequently) with the method for synchronization, in fact feasible each air-link frames from each radio frequency unit RFU contains identical various symbols.This also can produce and divide into groups and finish with distributed second layer symbol, but needs a basic synchronization agreement, so that guarantee the synchronous of symbol and airlink grouping in the whole system scope.

What Fig. 4 represented a single sub-district/3 sectors uses figure again.This diagrammatic representation is used for sector A, each element of B and C.Fig. 4 represents from how the to be added to example of mobile radio station at the middle part that is positioned at sector A of the signal that each the different radio frequency unit RFU that is positioned at sector A sends.Result according to an experimental group, (the deterioration that does not comprise channel fading and increased of sub-district framework, do not comprise the influence of floor and ceiling yet) a simple edge show, in some cases, pseudo-simulcasting system may have only the C/I of 9.5dB with respect to each symbol that is total to the asynchronous control channel of channel.This is usually less than the operation threshold of strengthening network (iDEN) waveform at integrated dispatch, and supposes that one is planned well and normal topology network, and in certain embodiments, this may not accomplish.This data point encourages is used true radio hookup control channel (and bearer traffic) message, rather than uses pseudo-radio hookup message.

Comprise because vector signal is offset the dummy status (null) that produces with the relevant problem of radio hookup message, and in system-wide to the synchronisation requirement between each symbol (discussed above) and the timing of RF frame.Each network signal will disturb devastatingly, and this will produce dummy status in the overlay area of system., degree of depth dummy status or the approaching condition of signal cancellation completely are very special.Sending according to the line-of-sight propagation model under the situation of two groups of network signals, unique to make the position of each signal cancellation be exactly mid point, and here, the loss of each paths equates, and and if only if the phase place of RF envelope when just in time opposite, could realize counteracting.Along with the randomness of channel variation and the increase of broadcasting the number of carrier frequency, the counteracting that produce the dummy status of the degree of depth becomes difficult more.Use simply and show based on the path loss model of distance and the system emulation of random carrier PHASE DISTRIBUTION uniformly, use two simulcast pilots, exist the degree of depth to cover the area of danger of dummy status less than 1% of total overlay area, and use 3 simulcast pilots, the probability of danger that this covering dummy status then occurs is less than 0.1%.These simple estimations show that from covering the viewpoint of reliability, it is not a main thing usually that signal covers dummy status.It is also important that it is also noted that in the 850MHz frequency range, signal wavelength is enough short, makes by mobile receiver is only moved 10cm, just can change into and produce the required specific phase cancellation of degree of depth dummy status.Therefore, a mobile receiver will unlikely rest among the degree of depth dummy status for a long time.

From system's viewpoint regularly, simulcast forward link packets will show as a kind of multipath factor concerning mobile receiver.If between first and last received network signal between total time-delay time-delay of surpassing this system signal scatter tolerance, deterioration then will appear.The simple estimation that system delay is scattered tolerance will be in exactly in the 10dB scope of main signal all signal limitations send symbol duration 10% in.Concerning integrated dispatch reinforcing mat (iDEN) waveform, this is interpreted as in 25 μ s, captures all signals in the 10dB scope that is in main signal.Because true simulcast pilots will be frequently as strong, perhaps be better than interested signal (concerning normal multipath system, this is a unlikely script) signal receive, a kind of conservative estimation will be limited in the 15 μ s between the total extra time-delay between first and last signal of interest.Even distribute 5 μ s to change to the flight time, still allowing can have 1 mile extra time between radio frequency unit of serving and every other important radio frequency unit.Even in line-of-sight network, be positioned at each radio frequency unit beyond 1 mile of the radio frequency unit of serving and will do not have enough energy and make signal degradation.Therefore, in this embodiment, it is the permissible value of a conservative extra flight time that 5 μ s show as.This still regularly changes " time-delay is scattered " budget that has kept 10 μ s for the base station to the frame of base station.

Fig. 5-9 expression is according to each different embodiment of each simulcasting system of the present invention.In the embodiment of Fig. 5-6, on network chassis unit NCU, carry out the second layer (L2) and handle, carry out ground floor (L1) processing on the link rack unit ACU aloft simultaneously.In the embodiment of Fig. 7-8, on network chassis unit NCU, carry out all departures and the second layer (L2) and handle, aloft enter a country and ground floor (L1) processing on the link rack unit ACU simultaneously.In the embodiment of Fig. 9, all handle and all on network chassis unit NCU, finish.

In order to realize simulcasting system of the present invention, can finish the generation of protocol Data Unit PDU an independent center, and, by using a multicast address, just protocol Data Unit can be assigned to simultaneously all airlink transaction card APC.A solution is exactly to allow the airlink transaction card APCs of task run on network chassis unit NCU such as related control program ACP, RAP RAP and voice channel procedure VCP, makes the latter become a network processing card NPC thus.Will there be a network processing card NPC each sector.Though voice channel procedure VCP does not have independent task, the processing of voice channel procedure VCP is to finish as the part of time slot request, and above-mentioned time slot request still awaits centralization.

In the embodiment of Fig. 5, the processing of half storey on all second layers (CPU, RAP, related control program and voice channel procedure) is placed among one or more network processing card NPC on the network chassis unit NCU.(FEC, processing MAC) is placed among one or more airlink transaction card APC on the airlink chassis units ACU half storey under the ground floor and the second layer.From each network processing card NPC each protocol Data Unit PDU to all (respectively) airlink chassis units ACU radio hookup departures, each IQ grouping is then produced and is broadcasted to all radio frequency unit RFU that are connected to (respectively) airlink chassis units ACU.On the immigration direction, each airlink chassis units ACU receives each IQ grouping from (respectively) radio frequency unit RFU, and selects best grouping to be used for handling.If one group of effective message is decoded, then it will be sent to network chassis unit NCU, carry out the last selection between each grouping that receives from several airlink chassis units ACU here.

One of advantage of the embodiment of Fig. 5 is exactly, because short and not frequent second layer protocol Data Unit, feasible traffic carrying capacity between network chassis unit NCU and airlink chassis units ACU seldom.In addition, the second layer of the centralization of stamp service time (timeslot number) guarantees realizing synchronously between each airlink chassis units ACU in immigration and departure message.And the second layer protocol Data Unit of the immigration that all are detected all is sent to network processing card NPC, this means in order to locate the mobile radio station in this system, does not need mobile management.At last, aspect non-simulcasting system of conversion, each processing module of CPU and digital signal processor is not needed to make change, only need make change in a small amount at each interface between digital signal processor and CPU.In addition, by improving the reliability of the link between network chassis unit NCU and (respectively) airlink chassis units ACU, just can guarantee that the important information (table being set as channel) that is kept by network chassis unit NCU can be sent to each airlink chassis units ACU reliably, thereby improve the reliability of native system further.

Fig. 6 is the every functional task figure among the embodiment of presentation graphs 5 (and with Fig. 2 similar a bit).In this embodiment, the message queue MQ interface between digital signal processor reader/writer task and message router MR is replaced by network interface, and the latter comprises the local area network (LAN) LAN that goes into from any side joint by each parts of TNETTASK.In an embodiment of the method, interrupt producing every time slot request by a 15ms clock on the network processing card NPC, and it is directly sent to message router MR.Then, use a multicast address,, the protocol Data Unit PDU that is produced is sent to all airlink transaction card APC via this network at a specific sector.Equally, receive each protocol Data Unit PDU of immigration via network.

Fig. 5 and 6 embodiment are positioned heavy IQ packet traffic for link between airlink chassis units ACU and radio frequency unit RFU, keep symmetry and simple with seasonal layer 2 interface.Even CPU and (respectively) digital signal processor are independent of outside NCU network chassis unit and the airlink chassis units ACU, the time-delay of message is acceptable normally.To/collect each protocol Data Unit PDU from several blades formula radio station RB, concentrate in together, in one section message, they are once sent then, just can reduce the number of message considerablely.In one embodiment, be the time slot of a 15ms of the time-delay of the message between CPU and digital signal processor distribution, this is normally enough.By improving the reliability of the link between network chassis unit NCU and each airlink chassis units ACU, just can improve the reliability of native system further.Compare with some other embodiment, this architecture has less traffic carrying capacity, makes it might to go to develop the reliable every agreement between network chassis unit NCU and each airlink chassis units ACU.

In the embodiment of Fig. 7, the second layer of all departures and ground floor are handled among one or more network processing card NPC that all is set among the network chassis unit NCU.From each network processing card NPC each IQ grouping to all radio frequency unit RFU radio hookup departures.The processing of half storey (CPU) is set among one or more network processing card NPC on the network chassis unit NCU on the second layer of immigration.The processing of half storey and ground floor (digital signal processor) is set among one or more airlink transaction card APC of airlink chassis units ACU under the second layer.On the immigration direction, each airlink chassis units ACU receives each IQ grouping from each radio frequency unit RFU, and selects best grouping to be used for handling.If one group of effective message is decoded, then it will be sent to network chassis unit NCU, carry out the last selection between each grouping that receives from several airlink chassis units ACU here.

One of advantage of the embodiment of Fig. 7 is exactly, and the second layer of centralization/ground floor is handled and can be guaranteed synchronous between each radio blade RB.In addition, the processing of half storey means in order to locate the mobile radio station in this system on the second layer of the immigration of centralization, does not need mobile management.And if a system clock is arranged on network chassis unit NCU, then it can anticipate the time slot of respectively leaving the country, and system delay can shorten (for the buffering capacity that the variation that compensates the immigration time-delay is required less).In addition, by between network chassis unit NCU and (respectively) airlink chassis units ACU, increasing a reliable link, to guarantee to be sent to each airlink chassis units ACU (half storey and ground floor are needed in handling under the second layer of immigration) reliably, just can improve the reliability of native system further by the important information (table being set) that network chassis unit NCU is kept as channel.

The embodiment of Fig. 5 is exactly that in the embodiment of Fig. 5, on the departure direction, the traffic carrying capacity from network chassis unit NCU to (respectively) airlink chassis units ACU seldom sometimes than one of more welcome reason of embodiment of Fig. 7.In addition, because in the embodiment of Fig. 7, immigration and departure are handled asymmetric, so concerning some embodiment, also need the interface between CPU and Digital Signal Processing is further revised.

Fig. 8 is similar to Fig. 2 and 6, is the figure of expression at every functional task of the embodiment of Fig. 7.According to the method, directly broadcast original I/Q sample packet to all radio frequency unit RB, and do not need extra network task is all done one time from network processing card NPC.The immigration Message Processing is similar to the processing of Fig. 6.The basic premise of the method for Fig. 8 is by eliminating via the demand of network to another piece plate transmission data, just can shorten the transmission delay of departure protocol Data Unit PDU.Equally, owing to the departure data are broadcasted, so total network traffic is an acceptable.

Fig. 7 and 8 embodiment are particularly advantageous at one in can be with the minimized system of time-delay.In order to improve this architecture, can provide the access of system clock and the portion copy of time slot and frame register to network chassis unit NCU/ network processing card NPC, feasible time slot request from airlink chassis units ACU to network chassis unit NCU becomes usually not to be needed.

In the embodiment of Fig. 9, all second layers and ground floor are handled among one or more network processing card NPC that all is set on the network chassis unit NCU.From each network processing card NPC each IQ grouping to all radio blade RB radio hookup departures, and beam back each IQ grouping of immigration to each network processing card NPC from each radio blade RB, here, each digital signal processor is made last selection in these groupings.

One of advantage of the embodiment of Fig. 9 is exactly that the second layer of centralization/ground floor is handled provides bigger synchronous likelihood score between each radio blade.In addition, because all IQ grouping all is sent to network processing card NPC, so, do not need by mobile management for a mobile unit in the navigation system.Follow the embodiment of Fig. 5-8 to do one relatively the embodiment of Fig. 9, in the embodiment of Fig. 9, between network chassis unit NCU and airlink chassis units ACU, there is sizable traffic carrying capacity, this is because in each time interval, each radio blade RB sends and receives an IQ grouping, in one embodiment, the above-mentioned time interval be every 7.5ms once.In addition, in the embodiment of Fig. 9, digital signal processor must be able to be stored dozens of IQ grouping, is used for selection course, and the storage (capacity) owing to limited on digital signal processor may limit to some extent to this.Therefore, in some cases, the embodiment of Fig. 5-8 may be more more desirable than the embodiment of Fig. 9.

Main difference between the method for the method of Fig. 5 and 6 and Fig. 7 and 8 is, in the method for Fig. 5 and 6, produce each protocol Data Unit PDU in network processing card NPC, and the latter is sent to airlink transaction card APC.Then, aloft the digital signal processor among the link processing card APC is converted to original I/Q sample with these protocol Data Units PDU, so that send to each radio blade RB.In the method for Fig. 7 and 8, digital signal processor on network processing card NPC is converted to original I/Q sample with these protocol Data Units PDU, and directly send to each radio blade RB, walk around airlink transaction card APC thus from network processing card NPC.

Be an importance of simulcast method of the present invention regularly, in one embodiment, speech performance allows to handle a time slot request with about 14ms, and responds every departure request with protocol Data Unit PDU.Concerning the protocol Data Unit PDU of immigration, an important requirement is exactly that before next protocol Data Unit PDU arrived after 15ms, these data were removed from digital signal processor.Compare with non-simulcasting system, the method for Fig. 5-8 also has two additional difference.Total one is the voice packet delay time that needs shortening, be exactly the requirement of another grouped data.

People wish to shorten the amount of voice delay in system.For example, reportedly send in the system that runs on 15ms (time slot request) and 15ms+7.5ms (half time slot) timing reference in a number, one section time-delay shortening that is at least 7.5ms has special value.Here to point out two zones, may occur time-delay therein and shorten.Concerning first zone of shortening that may occur delaying time, in one embodiment, to do the time of output time slot request from digital signal processor and must be sent to these data the time of digital signal processor, this departure regularly allows the grouping time-delay of a 14ms.This can be shortened the time-delay of a 7.5ms.Concerning second zone that may shorten time-delay, in one embodiment, digital signal processor is used for 15ms SRC filtering at first, and it may shorten to 7.5ms with this time.

Concerning a simulcasting system, a time may be represented in a time slot request border, and in this time, each grouping of departure must arrive network processing card NPC, so that handle in next time slot.In one embodiment, distribute 7.5ms be used to handle departure each protocol Data Unit, they are sent to airlink transaction card APC, and each protocol Data Unit PDU is written to digital signal processor.

Each packets of voice that arrives from integrated site controller ISC also is synchronized to identical boundary of time slot, and itself will arrive in for the suitably consistent time with respect to time slot request border at one.In order to have benefited from the minimum amount of delay that voice channel procedure message need be waited for before sending the time slot request, can carry out the JITA process.

Grouped data has additional requirement to voice.Concerning speech processes, the departure data do not rely on the immigration data, and they can be used as two independently process move.For grouped data PRAP message, the departure data depend on the immigration data, and before it had received and handled the immigration data, it did not usually send its departure data.

Concerning the method for the method of Fig. 5 and 6 and Fig. 7 and 8, immigration may be identical regularly.Because according to the type of message (RAP RAP, related control program ACP, or the like), the processing of digital signal processor takies the time of one section variable number possibly, so immigration regularly is complicated more.Equally, transmit data to airlink transaction card APC and depend on also when the IQ grouping arrives digital signal processor.Any amount of timing at input changes the timing variation that all will cause when these data are sent to airlink transaction card APC.

In order to reduce the quantity of the internet message that is sent to network processing card NPC, as long as protocol Data Unit PDU arrives airlink transaction card APC, it just is not sent to network processing card NPC usually.The substitute is, it is tied and be sent out in particular time interval.To the binding of protocol Data Unit PDU be described in detail below.

Because PRAP message is sent usually as soon as possible, and other message may no longer keep after 7ms, so, according to the interval of 7.5ms, send the protocol Data Unit PDU of binding to network processing card NPC from airlink transaction card APC.Any protocol Data Unit that occurs in first 7.5ms (particularly each PRAP protocol Data Unit) can be sent out at half crack boundary, and any protocol Data Unit that occurs in second 7.5ms can be sent out at the boundary of time slot place.

Though may occur time-delay in certain embodiments, this normally a kind of rare worst-case up to 40ms.Similar all message except related control channel ACC and grouped data, all will appear at first half crack.Equally, transmit and handle required time from airlink transaction card APC typically about 1ms to network processing card NPC.

In one embodiment, have 7 radio blade of reaching and carry out work in any given time, wherein, 6 are exclusively used in voice, and 1 is exclusively used in grouped data.Because some task need be used UDP, makes network activity that the transmission/reception of important grouping is postponed till after the key activities, so this need be considered to a kind of potential time-delay.Estimating that a problem in the time of implementation is exactly that many incidents may occur randomly with asynchronously.Every time slot request and immigration data appear at the border of each 15ms, but many other incidents are considered to occur at any time sometimes and the chance event of appearance together.For example, in the interval of 15ms, a packets of voice may arrive from integrated site controller ISC there, a packets of voice may be sent to integrated site controller ISC, the control channel DCC grouping of a special use may arrive from integrated site controller ISC there, and a handover message may be sent to integrated site controller ISC.

In order to reduce randomness, in the time of can beginning at the boundary of the time slot request of every 15ms to time slot request handle, and transmit from the data of digital signal processor and can be set at the time slot of every 15ms and the boundary in half crack.Handover message from digital signal processor can divide into groups to send in company with each half crack.Consequently, most of chance events will be some asynchronous incident and pseudo-incident.This distribution has such effect: in interval, a half crack, handles the departure data, and in interval, second half crack, handles the data of entering a country, but except the accidental related control channel ACC message.This method can also be used for the timing specification of grouped data.

For the timing in the method that realizes Fig. 5 and 6, in one embodiment, can't help digital signal processor produces the time slot request, would rather, be positioned at clock on the network processing card NPC, that be synchronized with the 15ms of integrated site controller ISC from one and produce the time slot request.This means and not read every time slot request from digital signal processor.In one embodiment, suppose that sending a grouping needs 50 μ s, and receive a grouping and need 120 μ s.In order more effectively to utilize message, adopt the grouping binding, this will be described in detail below.

Put it briefly, in the grouping binding, each grouping that is used for all radio blade RB will be combined, and make the as a whole airlink transaction card APC that is sent to.Consequently, only call muxSend function (being used for sending grouping) and call a filter function (being used to receive data).This is applied to network processing card NPC and airlink transaction card APC on both direction.Only mail to/individually sent from the message of integrated site controller ISC.This means only to have a network to send and receive for all 7 radio blade RB, rather than 7 times other transmission/receptions.The time of implementation of network processing card NPC is considered to a limiting factor usually.Owing to go over whole network required extra time, institute thinks to finish enters a country and required total time of implementation of departure incident increases to some extent.Yet some such processing is aloft finished on the link processing card APC, and some is finished on network processing card NPC.Because network processing card NPC must handle additional asynchronous events, so its time of implementation is important for finishing the work in the window of 7.5ms.Handover message usually with the immigration data combination together.

For the method for Fig. 7 and 8, in one embodiment, use a multicast address in the Ethernet grouping, directly send each IQ sample to all radio blade RB from network processing card NPC.Immigration is handled and is delayed time and is similar to the situation of Fig. 5 and 6 usually.Two importances of Fig. 7 and 8 method are exactly, time of implementation of network processing card NPC and in the influence of the excessive data on the network between network processing card NPC and the airlink transaction card APC.

For the timing in the method that realizes Fig. 7 and 8, in one embodiment, can't help digital signal processor produces the time slot request, would rather, be positioned at clock on the network processing card NPC, that be synchronized with the 15ms of integrated site controller ISC from one and produce the time slot request.This means and not read every time slot request from digital signal processor.In one embodiment, suppose that sending a grouping needs 50 μ s, and receive a grouping and need 120 μ s.Have only the immigration data that extra transmission/reception is just arranged.In order more effectively to utilize message, on the immigration direction, adopt the grouping binding, this will be described in detail below.In one embodiment, immigration is handled aloft and is distributed between the link processing card APC and network processing card NPC.Owing to go over whole network required extra time, institute thinks to finish enters a country and required total time of implementation of departure incident increases to some extent.Because some such processing is aloft finished on the link processing card APC, some is then finished on network processing card NPC, so the allowance of resulting network processing card NPC should only be reflected in the processing of finishing on the network processing card NPC.Handover message usually with the immigration data combination together.

In an embodiment of the method for Fig. 5-8, all radio blade RB are mutually synchronously in 10 μ s.The timing signal of 1pps (1 pulse-translator of per second) from integrated site controller ISC will be provided to all radio blade RB.It is the incident of 1pps that this 1pps timing signal also can be taken as.Also have, will provide a time slot and frame number to one or more radio blade RB, it will be assigned to each radio blade RB by Ethernet, as synchronous (SYNC) message of Ethernet.All radio blade RB that receive the Ethernet synchronization message are set to their time slot and frame counter the numerical value that provided by the Ethernet synchronization message in next 1pps incident.This shows this 1pps incident, and perhaps timing signal must be assigned to all radio blade RB in the time that differs 10 μ s each other.To in network chassis unit NCU, finish the generation of Ethernet synchronization message by FPGA parts or by CPU.Then, by using a multicast address, broadcast this Ethernet synchronization message to all radio blade RB.Can supply alternatively, can before abutting against a specific 1pps incident, send one group of Ethernet synchronization message to one or more radio blade, and send other Ethernet synchronization message to one or more other radio blade after a while, come all radio blade RB synchronously in mode optionally.Time slot that is comprised in what Ethernet synchronization message in office and frame number all must accord with the current flowing water counting of time slot and frame number, make any radio blade can individually be synchronized with the remainder of this system.

In one embodiment, base when NCU and each ACU provide preferably, it comprises the flowing water counting of time slot and frame.Base will produce according to the following step synchronously during NCU.If network processing card NPC still is unrealized synchronously, then it contains wait the reception of Ethernet synchronization message of the adjustment information of time slot and frame.The every 15ms of network processing card NPC ISR will respond periodic interruptions one time, and read out in a register on the FPGA parts.When the 1pps bit on parts FPGA becomes SM set mode (only just be set at once interrupting, this interrupts corresponding to a 1pps incident), network processing card NPC will bring into use the timeslot number from the time slot adjustment information, as new timeslot number.Each 15ms interrupts making timeslot number add 1.Base is synchronous when using a process to make ACU, and this process is substantially the same in the synchronous employed process of NCU.

In non-simulcasting system, each radio blade RB sends an I/Q grouping that has the immigration of timeslot number to digital signal processor.Then, digital signal processor sends to airlink transaction card APC with this timeslot number, as the part of time slot request.Then, airlink transaction card APC responds with a protocol Data Unit based on timeslot number.For the method for Fig. 5-8, usually can be by the amount of delay of being introduced to time slot request of network processing card NPC transmission from airlink transaction card APC greater than the time-delay of expection.The time-delay of calcaneus rete network is combined, and this will produce an appreciable time slot request message that is changing at network chassis unit NCU.Solution is exactly to produce the time slot request by the synchronous 15ms clock on the network chassis unit NCU, produces the time slot request and can't help digital signal processor.Remaining issues is exactly that network processing card NPC need know that the part which time slot is taken as the time slot request handles.Network processing card NPC is in order to realize that with correct timeslot number synchronous method is exactly to allow network processing card NPC send a pulse width modulation (PWM) signal from the FPGA parts that are positioned on the LAN adapter LC to the FPGA parts that are positioned on the network processing card NPC.This pulse width modulating signal will contain the clock information and the 1pps information of 15ms (also may be 7.5ms).

A problem that is accompanied by an embodiment of non-simulcasting system is exactly that when the channel setting (table) from integrated site controller ISC arrived, Tx and Rx digital signal processor cores get were provided with table with this part channel simultaneously and upgrade.The processing of departure time slot lags behind the immigration time slot and handles a pair of time slots.This means that the processing of departure time slot has a table that does not upgrade as yet.Concerning voice, be enough to hold one section time-delay between the actual use owing to being set to, so this can not become a problem usually from channel.Concerning grouped data, be connected owing between each protocol Data Unit PDU of immigration and departure, exist more closely, so be such situation no longer just.A solution is exactly to allow digital signal processor remove the phase of history of memory channel table, and still, because the restriction in space, this can not be as a kind of selection.Another kind of solution is exactly to send a new channel table with each protocol Data Unit of each group time slot request.This has the benefit of several respects: its signaling channel table of the first keeps synchronously with every time slot request, and upgrades along with every time slot request.It two is when network processing card NPC being proposed channel request is set, and it has been eliminated primary network and has sent and call, and thus, has shortened expensive channel the time is set.

As mentioned above, the method utilization of Fig. 5-8 binding technology is so that more effectively use network.Therefore, in one embodiment, be combined in together usually, and be grouped in together as an Ethernet and send at each protocol Data Unit PDU of each radio blade RB.Consequently have only a muxSend to be used to send data, and only call filter function one time, in order to receive data.Each radio frequency unit RFU finishes once binding, and the latter is handled by an independent airlink transaction card APC.The binding of protocol Data Unit PDU can be used to send data from network processing card NPC to airlink transaction card APC, the time slot request that is used to leave the country, and send data to network processing card NPC from airlink transaction card APC, the data that are used to enter a country show.Binding manages to protocol Data Unit by the digital signal processor reader task on message router MR on the network processing card NPC and the airlink transaction card APC.

The method of Fig. 5-8 relies on the network interface that is used for protocol Data Unit PDU exchange and is used to send protocol Data Unit PDU or original I/Q sample.In one embodiment, a Mo Tuoluola FCC Ethernet driver is all passed through in the all-network activity on the link processing card APC aloft, and is dispatched by tNetTask.In tNetTask, for a " work " is arranged in each grouping that institute receives and sends.When the time that is used to carry out this work arrived, data just were sent to driver for sending and receiving.As long as call muxSend, just can be sent to/send from all data of airlink transaction card APC.Because whole network stacks is operation under same tNetTask management all, so any TCP/IP activity also all need be dispatched by the activity of this protocol Data Unit PDU, and, consequently, because it need wait for the TCP/IP packet transaction, so the longer time-delay than expection may appear in the transmission/reception of protocol Data Unit PDU grouping.In one embodiment, concerning all synchronously and the asynchronous activity, be 120 μ s at the possible maximum delay of the once independent conflict in minute group of received, be 170 μ s at possible maximum delay from an ICMR activity of pseudo-vital task.The allowance of network processing card NPC is enough to hold these time-delays.

Because the transmission of protocol Data Unit PDU occurs among the distributed network, so a kind of consideration just provides a reliable link.In order to produce a reliable link, the detection to packet loss or link failure is provided, and when grouping is lost efficacy, had retransmitted.

The switch configuration that is used for the system of Fig. 5-8 is provided with like this, and feasible 0 * 1180 grouping from each radio blade RB does not send from the interface between airlink chassis units and network chassis unit.On the departure direction, the Ethernet that has bundled the grouping of all protocol Data Unit PDU all has an associated multicast address.On the immigration direction, the Ethernet that has bundled the grouping of all protocol Data Unit PDU all has an associated unicast address.

Be each airlink transaction card APC and each network processing card NPC give information form and type, they are expressed as the accumulation of a plurality of protocol Data Unit PDU (binding) of the link processing card APC of the aerial portion of each piece appointment.The code that is used for airlink transaction card APC and network processing card NPC is realized like this, makes identical message arrive suitable message queue.Because the Ethernet program uses a multicast address to replace unicast address, is different from non-simulcasting system so be sent to the message source of message queue.Message router task MR accumulates each protocol Data Unit mutually.In case receive each protocol Data Unit PDU, just measure and select an independent protocol Data Unit PDU based on RSSI.This protocol Data Unit PDU is sent to appropriate tasks just.

Each digital signal processor is carried out the filtering of immigration message.In order to accomplish this point, each digital signal processor is apprised of interpolation or the removal of the radio blade RB that they need handle.This usually need one-period be sent to a sign or bit field be set in the message of digital signal processor.

Compare with non-simulcasting system, concerning the method for Fig. 5-8, each digital signal processor does not produce the time slot request on 15ms time slot request border.Yet by interruption is provided, which timeslot number what this system just can check that this digital signal processor handling is, and system thinks which timeslot number it should handle, and thus, informs in these cases and makes mistakes.

Digital signal processor also sends a RSSI who is provided with and measures in each section immigration message.This is to be used to enter a country a required standard of message selection in network chassis unit NCU.In order to simplify interface, digital signal processor all sends one group of immigration message with RSSI in each slot time, but only for effective each protocol Data Unit PDU effective synchronous mark is set.This makes this interface often all send the definite type of each information, and needn't go the message of compensate for variable length.

Concerning the method for Fig. 5-8, in general, network management comprises network processing card NPC is connected with one group of airlink transaction card APC and the radio frequency unit RFU that relates to particular sector.This comprises to suitable sector and correctly distributes multicast address.The code library that is used for airlink transaction card APC and network processing card NPC can be identical, but each can both discern it self, so that begin to carry out specific task.Each radio blade RB that network management also is responsible for adding informs suitable network processing card NPC and airlink transaction card APC.

Concerning the method for Fig. 5-6, the reader of number of support word signal processor and the accumulation of the code of write device are from each protocol Data Unit PDU of each digital signal processor, subsequently, carry out one section muxSend to replace msgQSend, in order to send each protocol Data Unit PDU.In order to receive, digital signal processor write device WT receives one section message from filter function, to replace the message from message router task MR, as what done in non-simulcasting system.Compare with non-simulcasting system further, in the method for Fig. 5-6, message router task MR sends by muxSend, to replace msgQSend.The data that received continue to arrive via the message queue from filter function.Network management knows all that usually each digital signal processor only is positioned on the airlink transaction card APC, and usually not on network processing card NPC.

Concerning the method for Fig. 7-8, the reader of number of support word signal processor and the accumulation of the code of write device are from each protocol Data Unit PDU of each digital signal processor, subsequently, carry out one section muxSend to replace msgQSend, in order to send each protocol Data Unit PDU to network processing card NPC, as what in the embodiment of non-radio hookup, done.In both cases, digital signal processor write device WT is similar.Equally, each load of digital signal processor is set up, so that support the departure that an independent immigration is handled and independent that is used for each core to handle.This embodiment may need two departure loads that are positioned on the network processing card NPC, and two immigration loads that are positioned on the airlink transaction card APC, makes each airlink transaction card APC have the configuration that is different from each network processing card NPC.Network management all know usually each the departure digital signal processor be positioned on the network processing card NPC, and each the immigration digital signal processor be positioned on the airlink transaction card APC so that be provided with and dispose.

Concerning the embodiment of Fig. 5-8, run on one independently on the network processing card NPC from the grouped data of speech processes.This network processing card NPC provides the interruption of a 15ms to microprocessor PROC, and the latter is synchronized with the 1pps clock of integrated site controller ISC.Network processing card NPC can also support reading with the relevant timeslot number of this time slot request.

Concerning the embodiment of Fig. 5-8, it is a kind of like this method that immigration divides group selection, by means of this kind method, be combined into a piece of news from the message of respectively entering a country of each radio frequency unit RFU and each airlink chassis units ACU, it represents that this sector (message) is sent to integrated site controller ISC.A plurality of airlink chassis units ACU and a plurality of radio frequency unit RFU can be dispensed within the sector.Concerning a plurality of radio frequency unit RFU, digital signal processor is responsible for selecting suitable immigration grouping.Consequently, represent a specific sector from one of each airlink chassis units ACU independent immigration message.

Concerning a plurality of airlink chassis units ACU, network processing card NPC receives an immigration message there from each piece airlink transaction card APC, and it represents a specific time slot and sector.Whether no matter there is actual message to provide, also no matter whether have the message that a time slot request border can be used for related control channel ACC type, immigration message all appears on each border, half crack.Be attached on each section immigration message is channel quality measurement values at this particular time-slot.This channel quality measures comprises RSSI, I+N and synchronous error flag.Synchronously error flag is used at first compare with each time slot, and has invalid synchronous any message and will be left in the basket.Then, RSSI and I+N numerical value are used to calculate at the SQE from the particular time-slot of each airlink chassis units ACU.Airlink chassis units ACU with maximum SQE makes its immigration message be selected as suitable message for processing, and sends it to integrated site controller ISC.Except selecting immigration message, the channel quality set that is selected also is to be used for this particular time-slot to form the set of handover message.In one embodiment, this metric set is sent to the new base station management entity BRM of network chassis unit, and here, it has accumulated 24 groups of measurement results, and sends as a handover measurement report subsequently.

According to the time of implementation of network processing card NPC, when following the method for Fig. 5-6 method of Fig. 7-8 to compare, in certain embodiments, be unloaded to other processor, a kind of more effectively processing procedure just can be provided by time-consuming digital signal processor is write.Though in one embodiment, grow 170 μ ss than the method for Fig. 5-6 total circulation timei, owing to needn't be at war with CPU on many other things, the possibility of time-delay is reduced.The efficient of the method for Fig. 5-6 is achieved, because the CPM module of network processing card NPC is more faster with the speed that protocol Data Unit PDU is written to each digital signal processor parts than the CPU of network processing card NPC to the speed that airlink transaction card APC sends Ethernet protocol data cell PDU sometimes.

In certain embodiments, the method for the method of Fig. 5-6 and Fig. 7-8 all obtains the total processing time-delay of required 15ms.Because immigration is handled with being separated on the different circuit boards of departure, so the software of the method for Fig. 7-8 is complicated a little.In addition, in certain embodiments, because the time slot request produces on network processing card NPC, rather than produce from each digital signal processor, so the method for the method of Fig. 5-6 and Fig. 7-8 is compared with non-simulcasting system, the time-delay that has all obtained 160 additional μ s is shortened.

Handle the time that protocol Data Unit PDU is required by being improved to each radio blade, can also improve the ability of the method and system of Fig. 5-8 usually.Total be exactly because the processing time that the number of each radio blade RB multiply by this time whole system, so saving all changes the saving of doubly taking advantage of into any time.In addition, when digital signal processor obtained that request message is set, the channel of system was provided with and with what upgrade digital signal processor table is set respectively.If this channel configuration information is sent out together in company with every time slot request, then it has eliminated the primary network transmission.

Concerning the method for Fig. 5-8, aloft the message that transmits between link processing card APC and the digital signal processor is Service Access point (SAP) message.Such message is usually with iDEN functional specification cooperating.Each protocol Data Unit PDU is combined with additional information, to set up Service Access point.That the time slot request is produced and that be written into digital signal processor just in time is exactly this Service Access point SAP.

In order to realize the method for Fig. 5-8, one section message ad hoc is used for representing the Service Access point SAP that is sent to the Service Access point SAP of airlink transaction card APC and is sent to network processing card NPC.This message table is shown on the specific airlink transaction card APC, at the accumulation of the Service Access point SAP of each radio frequency unit RFU.According to configuration, in an embodiment with 7 radio blade RB, each bar message can be represented nearly 7 Service Access point SAP, and each is corresponding to a radio blade RB.Each Service Access point SAP follows a different base station sign to link together.

Each radio blade RB or each radio frequency unit RFU can dynamically be inserted among the system of Fig. 5-8 in system start-up or after system moves.The beginning of each radio blade RB and the setting that stops and management are controlled by the iRBS and the iRBC task that run on an interface card IC and the LAN adapter LC respectively.Task CS is set controls setting and the management that base station part BRS and airlink transaction card APC channel are provided with table by being positioned at the base station part BRS on the LAN adapter LC and being positioned at channel on the network processing card NPC.Several different scripts that are provided with may occur, wherein each may need certain part of system is taked different a little actions.These scripts are all based on start a plurality of radio frequency unit RFU on a plurality of airlink chassis units ACU of same carrier frequency.To discuss 4 scripts below in more detail, wherein preceding 3 are shown in Figure 10.

As shown in figure 10, concerning first script, the first radio frequency unit RFU is activated in the first airlink chassis units ACU, and it is represented as ACU#1 in Figure 10.The situation of Here it is the first radio blade RB, it is designated as RB#1 in Figure 10.In this case, a base station part BR need be started by integrated site controller ISC, channel table need be informed that network processing card NPC channel is provided with CS, and airlink transaction card APC needs also to be provided with first with each channel table.In the embodiment of Figure 10, all each radio blade RB are on the same carrier, and are assigned to identical base station BR and identical sector.

As Figure 10 will illustrate further, concerning the crus secunda basis, the second radio frequency unit RFU is activated in the first airlink chassis units ACU, in this case, only need that another radio blade RB (being designated as RB#2 in Figure 10) has been started this situation and inform this airlink transaction card APC.This airlink transaction card APC will be provided with digital signal processor like this, make it know that another radio frequency unit RFU awaits handling.In this embodiment, no matter be base station part BRS, or network processing card NPC channel is provided with CS, do not need to know this situation usually.In this script, this base station BR is not unlocked as yet.

As Figure 10 will illustrate further, this, after base station BR is unlocked, the first radio frequency unit RFU that is positioned on second airlink chassis units will be activated concerning tripod.In this case, network processing card NPC channel management person CM need be arranged on each channel table on the new airlink transaction card APC.

Concerning the 4th script (not shown), the second radio frequency unit RFU is activated on an airlink chassis units ACU who has moved.In this case, only need have another radio blade RB to await handling this situation it and inform digital signal processor.In this case, digital signal processor need divide into groups to make decision with regard to accepting which immigration I/Q.

The message about adding an additional radio blade RB from the iRBS task can arrive at any time.In one embodiment, if base station part RBS still is lockable, then channel is provided with CS and must waits for that usually before airlink transaction card APC had been set, it received till the release triggering message from base station part RBS.In case base station RB is in released state, all additional radio blade RB will be provided with CS and manage by being positioned at channel on the network processing card NPC.

Must be noted that in the above-described embodiments each radio frequency unit RFU can have nearly 7 ones radio blade RB, but all follow a different base station part BRS to link together at each radio blade RB of a radio frequency unit RFU the inside.Though figure 10 illustrates an independent base station part BRS,, should be appreciated that every other base station part all will follow identical sequence independently.In Figure 10, the additional radio frequency unit RFU that is positioned on the independent airlink chassis units ACU represents to have added next radio blade RB, and it is in relation to initialized base station BR.In this embodiment, suppose that as long as have a radio blade RB at least in operation, then base station BR will be considered to be in operation.Must be noted that also that in this embodiment before having taken place synchronously of link rack unit ACU and network chassis unit NCU, radio blade RB usually can not take place synchronously aloft.

Between the model of Figure 10 and a difference between the non-simulcasting system is exactly message Add from the interpolation radio blade RB of iRBS task.Channel is provided with task CS and determines its whether first radio frequency unit RFU.If yes, then it is provided with the channel table that is positioned on the new airlink transaction card APC.If it is the second radio frequency unit RFU that is positioned on the first airlink transaction card APC, then it awaits handling this situation with additional radio blade RB and informs digital signal processor.Time slot/the synchronizing process of global position system GPS also is different, and this will discuss below in more detail.

After energized, network processing card NPC determines its whether network processing card NPC by reading a hardware register.In one embodiment, in case network processing card NPC has determined that it is a network processing card NPC, except digital signal processor reader RT (RFN_DSPR_TASK), digital signal processor write device (RFN_DSPW_TASK), and beyond message router MR (RFN_DSPX_TASK) task, all tasks all will be activated.

After energized, airlink transaction card APC determines its whether airlink transaction card APC by reading a hardware register.In case determined that it is an airlink transaction card APC, except random access procedure RAP, related control procedure ACP, immigration message task INTASK, departure message task OUTTASK, beyond the message router MR, all tasks and all grouped datas will be activated.

Except normal initiation sequence, in one embodiment, before allowing radio blade RB initialization, interface card IC also will confirm airlink chassis units ACU calcaneus rete network rack unit NCU foundation synchronously.The timing part FPGA calcaneus rete network interface card LC that is positioned on the interface card IC shakes hands, so that itself is set up synchronously.In one embodiment, be positioned at the state of the cpu cycle property ground inspection timing part FPGA on the interface card IC, and when detecting a synchronised clock, it proceeds the initialization of radio blade.In this embodiment, do not send immigration message to network chassis unit NCU usually, unless realized synchronously from airlink chassis units ACU.

In non-simulcasting system, be arranged on timeslot number on each radio blade RB by base station part BRS.This sequence starts from sending a global positioning system time slot/frame request by base station part BRS, and waits for the global positioning system time slot/frame response with timeslot number.When one of integrated site controller ISC transmission had the message of current global positioning system time slot/frame response, base station part BRS just sent an Ethernet message to radio blade RB, is used to be provided with timeslot number.When only having one one during with the relevant radio blade RB of base station BR, this method is an acceptable.Yet, concerning the simulcast method of Fig. 5-8, because each base station BR has a plurality of radio blade RB, so radio blade RB management role sends message to radio blade RB, so that timeslot number is set.

Figure 11 is a figure, and expression is as the part event sequence of global positioning system time slot/synchronizing process of carrying out for the insertion of each radio blade RB of.In Figure 11, all radio blade RB are in identical carrier frequency, and are assigned to identical base station BR and identical sector.As shown in figure 11, enter after global positioning system time slot/synchronous regime, base station part BRS just sends global positioning system time slot/frame sync message to the iRBS task.This iRBS task of this message trigger sends a global positioning system time slot/frame request to integrated site controller ISC, and integrated site controller ISC responds with frame and timeslot number in global positioning system time slot/frame response.If this is first radio blade, then the iRBS task is waited for this triggering from base station part BRSA, thereby it can keep correct state.

Along with additional radio blade RB is initialised at a specific base station BR, the iRBS task is to time slot/frame number of integrated site controller ISC request.This global positioning system time slot/frame response is a multicast address, and is sent to all the base station BR in the system.Usually under these circumstances, distributor (Distributor) task can transmit a piece of news by route to each base station part BRS, but because the iRBS task is responsible for all radio blade RB, so in this embodiment, utilization is sent to an independent message of this task, and is considered to enough.In the next 1pps that is following slot timing adjustment message, time slot and frame number are set up in the FPGA parts.

Concerning the method for Fig. 5-8, run on channel on the network processing card NPC and task CST is set is responsible for all airlink transaction card APC management channel tables in the system.Channel in non-simulcast method is provided with among the task CST, and the method is receiving logical channel assignment, and channel is provided with CS, and after the base station BR unlock trigger, just to the digital signal processor writing information.In the simulcast method of Fig. 5-8, each digital signal processor is not placed on the network processing card NPC, and this just need send to this information airlink transaction card APC, so that with its airlink transaction card APC that packs into.

With reference to Figure 10,0 * 1106 message be expanded be comprise be sent to/from the message of airlink transaction card APC, in order to this identical information to be set.0 * 1106 message is used to channel allocation and base station BR unlock trigger, and in this embodiment, it only occurs once in initialization procedure usually.Any channel that receives from integrated site controller ISC is provided with request CS and all is sent to airlink transaction card APC as the part of time slot request bundled messages usually, to reduce unnecessary network traffic.Receiving after a channel is provided with request, the channel table of local (network processing card NPC) is updated, to reflect various variations.

Channel is provided with task CST in response to many dissimilar message.Between network processing card NPC and system local area network interface card LC, 0 * 1106 message is arranged usually, between airlink transaction card APC and network processing card NPC, 0 * 1186 message is arranged usually.One of them is exactly a trigger that is used for release base station BR for these message.It is sent to network processing card NPC from base station part BRS.Another message is used for every system parameters is sent to airlink transaction card APC, and it sends to airlink transaction card APC from network processing card NPC.Another message is used to confirm the setting of system parameters, and it is sent to network processing card NPC from airlink transaction card APC.Another message is used to confirm release, and it is sent to base station part BRS from network processing card NPC.

Article one, additional message is used to confirm logical channel assignment, and it is sent to base station part BRS from network processing card NPC.Another message is a trigger, is used for base station BR is locked on the network processing card NPC, and it is sent to network processing card NPC from base station part BRS.Another message is a trigger, is used for base station BR is locked on the airlink transaction card APC, and it is sent to airlink transaction card APC from network processing card NPC.Another message is used to the master control channel that signaling is provided, and it is sent to airlink transaction card APC from network processing card NPC.Another message is used to confirm the setting of master control channel, and it is sent to network processing card NPC from airlink transaction card APC.The message of another kind of type is used to notify additional radio blade RB ready, it is sent to network processing card NPC from the rRBS task, it should be noted that, though this is not one group of 0 * 1106 message, but message from the iRBS task, it is provided with task CST to channel and shows that additional radio blade RB is ready now.

With reference to Figure 10, channel is provided with task will be provided with the channel information of network processing card NPC according to from one of the base station part BRS message of the base station BR of release.If this is an initial airlink transaction card APC, then it will be each system parameter setting on the specific airlink transaction card APC of this piece.Filter function on the network processing card NPC is responsible for that 0 * 1106 and 0 * 1186 all message is sent to channel task CS is set.Aloft on the link processing card APC, filter function is responsible for 0 * 1186 all message is sent to message router task.

Channel is provided with task CS and also is responsible for the channel table that provides suitable, prepares after each time slot request of processing it to be sent to digital signal processor.In one embodiment, at the channel table that is sent to the Tx core and be sent between the channel table of Rx core, need one section time-delay usually.This is because the Rx core is all being handled from one of the Tx core different time slot at any given time.In order to realize this point, adopt the channel table of a circle, as shown in figure 12.

With reference to Figure 12, an example is provided, therein, the different timeslot number of each numeral in circular buffer RBUF.Channel is provided with request CS will (for example, #1) begin, from this channel table of new configuration from a specific timeslot number.In this example, list item #9 is exactly the current previous list item that is using of Rx core, and #6 then is the current time-delay of using of Tx core each list item as N (being 3 in this example).By in each time slot request buffer pointer being added 1, each list item that is used for Tx and Rx core just is controlled.

Figure 13 is a flow chart of message router task MR.On the responsible link processing card APC aloft of message router task MR, set up the interface between this network and digital signal processor reader and write device task WT.This task is mainly in response to 3 incidents.First incident is exactly at a time slot request timer, and it is such incident, and therein, every 7.5ms just sends one group of bundled messages to network processing card NPC.Second incident is exactly at trigger message, and it is 0 * 1186 a base station BR control messages from network processing card NPC.The 3rd incident is at the departure bundled messages, and it is the departure bundled messages that is sent to the digital signal processor write device.

As shown in figure 13, at square frame 310, program makes each function initialization of variable.In decision block 312 and 314, program enters a circulation that comprises the 7.5ms time delay interval, and this circulation continues always, up to as in decision block 314, judged receive a piece of news till.In case receive a piece of news, program just enters decision block 320, and here, whether the message that programmed decision has received one group of immigration bundled messages.If this message is one group of immigration bundled messages, then program enters decision block 332.

At decision block 332, whether the afterbody of programmed decision bundled messages arrives.If the afterbody of bundled messages arrives, then program is returned.If the afterbody no show still of bundled messages, then program enters square frame 334.At square frame 334, program copy Mblk is as an independent protocol Data Unit PDU message.At square frame 336, program is sent to the signal processor write device with this message, turns back to decision block 332 subsequently.

At decision block 320, if the message that has received is not one group of immigration bundled messages, then program enters into decision block 322.At decision block 322, the message that programmed decision has received is one group of trigger message, rather than a slot timing device incident.If the message that has received is one group of trigger message, then program enters square frame 342.At square frame 342.Program is handled this trigger, returns then.

At decision block 322, if the message that programmed decision has received is not one group of trigger message, but a slot timing device incident, then program enters square frame 352.At square frame 352, program sends this bundled messages.At square frame 354, program is provided with one group of new bundled messages, returns then.

In one embodiment, in order to eliminate from digital signal processor reader task to one of message router task MR extra copy, an API is provided,, requires to return the pointer that a sensing should write the Mblk position of data with call number signal processor reader task.This just illustrates how the local copy of a departure bundled messages is realized sharing between digital signal processor reader task and message router task.The digital signal processor write device receives the message that contains a Mblk, wherein contains the protocol Data Unit of preparing to be written to digital signal processor.In order to reduce the original required change of digital signal processor write device version, use a function that a kind of method is provided, the data among a Mblk are copied to another Mblk.This program is copies data not in fact, would rather, it allows the reference count of original Mblk add 1, and this Mblk must not discharge, up to each of this Mblk quote all be released after till.This allows multiple messages to be sent to the digital signal processor write device, and each bar message is all quoted original Mblk, has eliminated the copy of protocol Data Unit PDU thus.

Digital signal processor reader task RT is responsible for reading immigration message from digital signal processor, and they are added among the bundled messages.In order to eliminate a copy that arrives the protocol Data Unit of message router from the digital signal processor reader via a message queue, in one embodiment, the digital signal processor reader directly writes to a Mblk, and this Mblk is shared by message router and digital signal processor reader.An API is provided, and it allows to carry out funcall, with the lock buffer device, and obtains writing position.

Digital signal processor write device task WT receives one group of independent message with Mblk, wherein contains protocol Data Unit PDU to be written.Digital signal processor is written to protocol Data Unit PDU digital signal processor and discharges this Mblk.

Message router task MR handles departure time slot request or the request of immigration time slot independently.As the result that need select from a plurality of airlink chassis units ACU, some complexity that seems is handled in immigration.In one embodiment, message router task MR is broken down into two independently tasks, is called departure message task OUTTASK and immigration message task INTASK.In one embodiment, OUTTASK is on time slot request border, and every 15ms receives an independent time slot request message.This message is from time slot request timer.Then, OUTTASK checks the base station BR of the activity that each has been set up, and calls suitable time slot request function, simultaneously resulting protocol data cell PDU is added among the departure message.

Figure 14 is a flow chart of exit message Processing tasks OUTTASK.Square frame " departure SDB processing procedure " is followed done similar in non-simulcasting system.As shown in figure 14, at square frame 410, program makes each function initialization of

variable.In decision block

412 and 414, program enters one by the timer-operated circulation of time slot request, when receives a piece of news to judge.When decision block 414 judgements had received a piece of news, program entered square frame 420.

At square frame 420, the base station of program acquisition activity sign.At square frame 422, the routine processes timeslot number makes it to become time slot type information (base station sign baseRld).At square frame 424, each protocol Data Unit PDU of routine processes departure.At decision block 426, whether the afterbody in programmed decision operation base radio station arrives.If no show still, then program turns back to square frame 420.At decision block 426, arrive if judge the afterbody in operation base radio station, then program enters square frame 428.At square frame 428, program sends one group of bundled messages to airlink transaction card APC, turns back to decision block 412 then.

In simulcasting system of the present invention, owing to having multi-section radio blade RB, a specific channel is in the H.D of a mobile device, so a piece of news that sends from this mobile device may be received by multi-section radio blade RB.Consequently, send identical message from two sources to immigration message task INTASK.In one embodiment, before related control procedure ACP or random access procedure RAP task were carried out their processing, the message of these repetitions was reduced to an independent message.On two positions, need to select immigration message.A position is exactly the digital signal processor that is used for specific radio frequency unit RFU that discuss the front.Second position is exactly the immigration message task INTASK that is used for from the message of each airlink chassis units ACU, and this will discuss below in more detail.

When receiving an immigration message, immigration message task INTASK receives has the Ethernet message of a bundle protocol Data Unit PDU.This is a bundle message from each airlink chassis units ACU in the system.If have only an airlink chassis units ACU, first bundle that has then received is with processed.If in the system a plurality of airlink chassis units ACU are arranged, then allow immigration message task INTASK wait for usually, till receiving whole bundled messages.An immigration message timer is used to anti-locking system and occurs indefinitely blocking because of losing a bundle.

In case receive whole bundled messages, just all protocol Data Unit PDU that have identical base station designator BRID and timeslot number in this bundle compared.At first check each protocol Data Unit PDU, so that be sure of its synchronous mark for effective, and, if it is not that effectively then this protocol Data Unit PDU is abandoned; Otherwise it is used to comparison.In one embodiment, will calculate the SQE of this time slot according to RSSI and I+N.Protocol Data Unit with maximum SQE will be selected and be used for further processing, and each remaining protocol Data Unit will fail to be elected.The rssi measurement set that is accompanied by the protocol Data Unit PDU that is selected is sent to suitable base station part BRS, in order to produce handover message.

Figure 15 is the flow chart of an immigration message handling task INTASK.In one embodiment, a timer is only arranged, and it is activated when receiving the message of the first airlink chassis units ACU.If timer event occurs, then use current available bundled messages, and all will be abandoned from any bundled messages subsequently of identical time slot.This system also reports and makes mistakes.The resolution of the timer of Figure 15 is usually above resolution that system clock provided.

As shown in figure 15, at square frame 510, program makes each function initialization of variable.In decision block 512 and 514, program enters a circulation with timer, and this timer is activated in square frame 524 (this will discuss below in more detail), and this circulation continues always, till receiving a piece of news.At decision block 514, in case judgement has received a piece of news, program just continues to enter decision block 516.At decision block 516, the message that programmed decision has received is one group of immigration bundled messages, rather than one group of immigration message timer event.At decision block 516, if the message that programmed decision has received is not one group of immigration message, but one group of immigration timer event, then program enters square frame 532.At square frame 532, report makes mistakes, and program enters square frame 540, will be discussed in more detail below.

At decision block 516, if the message that programmed decision has received is immigration message, then program continues to enter decision block 522.At decision block 522, whether the last airlink chassis units ACU of programmed decision arrives.If last airlink chassis units ACU no show still, then program enters square frame 524, and here timer is activated, and program turns back to square frame 522.

At decision block 522, if the last airlink chassis units ACU of programmed decision arrives, then program continues to enter square frame 540.At square frame 540, the timer anergy.At square frame 542, suitable immigration protocol Data Unit PDU of procedure Selection.At square frame 544, program partly sends the HO measurement report to base station.At square frame 546, the SDB of routine processes immigration.At decision block 550, whether programmed decision has additional protocol Data Unit PDU in tabulation.If additional protocol Data Unit PDU is arranged in tabulation, then program turns back to square frame 542.If there is not remaining protocol Data Unit PDU in tabulation, then program turns back to decision block 512.

Network and radio blade RB management role are responsible for specific radio blade RB, radio frequency unit RFU, airlink transaction card APC, airlink chassis units ACU and base station part BRS are linked together.In simulcasting system of the present invention, can there be a plurality of radio frequency unit RFU and a plurality of airlink chassis units ACU to link together with a specific sector.When detecting a radio blade RB, the iRBS task notifies suitable network processing card NPC to claim, additional radio blade RB is added.This message authentication airlink transaction card APC and radio frequency unit RFU set up contact with radio blade RB.As mentioned above, radio blade task RBT is responsible for GPS time slot/frame synchronization is put on each radio blade RB.

In the embodiment of non-radio hookup, the time slot request is come message router MR together with timeslot number as message.In the embodiment of radio hookup, though one of difference is that message source is the ISR that is synchronized with integrated site controller ISC clock,, identical time slot request still is sent to departure message task OUTTASK.In one embodiment, interface card IC and LAN adapter LC have FPGA parts, it is programmed the signal that produces a 15ms, be sent to airlink transaction card APC/ network processing card NPC, the signal Synchronization of this 15ms is in the 1pps signal of integrated site controller ISC, and the trigger that is used as immigration and leaves the country and handle.

An immigration message timer is used to trigger such incident, and therein, INTASK expects the message from airlink chassis units ACU, but receives it never.The problem here is that immigration message task INTASK need know that it can satisfy regularly requirement of departure in a rational time interval (for example, about 1ms), even like this, in one embodiment, the resolution of system clock is about 16ms.In one embodiment, solution just is to use an auxiliary clock AUXCLOCK.This auxiliary clock AUXCLOCK is set to have 1000 and takes/second resolution, obtain the interval of 1ms thus.An ISR is added to this system clock, and therefore, it just sends a time to message (Timeout Message) to immigration message task INTASK every 1ms.Only after receiving article one immigration message, auxiliary clock AUXCLOCK just is activated, and is receiving second immigration message or time after message, and it is just by anergy.So just can avoid auxiliary clock AUXCLOCK when it is not required, to walk far and waste valuable cpu cycle.So just can effectively auxiliary clock AUXCLOCK be configured to the timer that a single triggers.

In signal and explanation the preferred embodiments of the present invention, should be appreciated that and under the prerequisite of not leaving spirit of the present invention and scope, can make various changes.

Claims

1. distributed radio system comprises:

A plurality of processing unit and a plurality of radio frequency sending set parts, described processing unit and radio frequency sending set parts pass through ethernet interconnect, described a plurality of processing unit comprises a CPU and a plurality of airlink processing unit at least, described CPU is responsible for setting up being connected between this system and the external environment condition, and each airlink processing unit all is connected to one or more radio frequency sending set parts by Ethernet; And

Wherein, this distributed radio system uses at least two radio frequency sending set parts that are designated as a radio frequency radio hookup group, on a common radio-frequency carrier wave, with the form radio hookup data of common modulation signal;

Wherein, each parts of described radio frequency radio hookup group are configured to receive the grouping of Ethernet sampled signal and send the information that is included in each grouping by modulating described radio-frequency carrier; And

Wherein, this simulcast method comprises:

First group of Ethernet grouping that emission will be received by each parts of described radio frequency radio hookup group, this first group of Ethernet grouping is used to divide into groups to respond as the Ethernet of specifying medium access control (MAC) address to containing this multicast address after each parts of described radio frequency radio hookup group with programme each parts of described radio frequency radio hookup group of a multicast address.

2. distributed radio system according to claim 1, wherein, this simulcast method also comprises:

Second group of Ethernet grouping that emission will be received by each parts of described radio frequency radio hookup group, this second group of Ethernet grouping are used to programme each parts of described radio frequency radio hookup group to use a specific radio-frequency carrier job.

3. distributed radio system according to claim 2, wherein, this simulcast method also comprises:

Periodically launch the grouping of Ethernet sampled signal, this Ethernet sampled signal grouping has a destination-address, it is equivalent to the multicast address of having programmed, each parts of described radio frequency radio hookup group are configured to receive this Ethernet sampled signal grouping, each parts of described radio frequency radio hookup group are modulated described common radio-frequency carrier wave according to the sampled data that is included in the grouping of Ethernet sampled signal.

4. distributed radio system according to claim 1, wherein, this system is positioned at indoor.

5. distributed radio system according to claim 4, wherein, this system is with low-power level work, even make that this system can not disturb existing macrocellular environment yet when working on this system is being used for the similar frequency bands of a given cellular standards.

6. distributed radio system according to claim 1 further comprises the data link layer that concentrates on described CPU, makes that described CPU is configured to send the grouping of multicast Ethernet to described a plurality of airlink processing units.

7. distributed radio system comprises:

A plurality of processing unit and a plurality of radio frequency sending set parts, described processing unit and radio frequency sending set parts pass through ethernet interconnect, described a plurality of processing unit comprises a CPU and a plurality of airlink processing unit at least, described CPU is responsible for setting up being connected between this system and the external environment condition, and each airlink processing unit all is connected to one or more radio frequency sending set parts by Ethernet;

Wherein, data link layer concentrates on described CPU; And

Wherein, the grouping of data link layer Ethernet is sent to a plurality of airlink processing units relevant with described radio frequency radio hookup group by described CPU, described a plurality of airlink processing unit also is processed into waveform with data in link layer, and these waveforms are launched by described radio frequency radio hookup group.

8. distributed radio system comprises:

Wherein, realize the emission of waveform by a kind of double-deck multicasting technology, wherein, data link layer emission data are as the Ethernet grouping of multicast, sent to a plurality of airlink processing units relevant by described CPU with the data link layer form with described radio frequency radio hookup group, afterwards, described emission data are sent to described radio frequency radio hookup group from each airlink processing unit with waveform.

9. distributed radio system comprises:

Wherein, will be tied by at least a portion of the data of described radio frequency radio hookup group emission, make some data at least of described a plurality of radio-frequency channels be set as the individual ethernet grouping.

10. distributed radio system according to claim 9, wherein, this system also comprises Ethernet switch, they are configured to the emission of specific data is filtered, so that restriction emission data volume.

11. distributed radio system according to claim 9, wherein, the parts of described radio frequency radio hookup group are configured to receive the grouping of Ethernet sampled signal, and send the information that is included in the described grouping by the modulated RF carrier wave.

12. distributed radio system according to claim 9, wherein, described system is positioned at indoor.

13. distributed radio system according to claim 12, wherein, described system operates at low-power level, and the system that makes does not disturb existing macrocellular environment, operates in being used for the same frequency band of given cellular standards even work as system.

14. a distributed radio system comprises:

A plurality of processing unit and a plurality of radio-frequency transmitter parts, described processing unit and radio-frequency transmitter parts pass through ethernet interconnect, described a plurality of processing unit comprises a CPU and a plurality of airlink processing unit at least, described CPU is responsible for setting up being connected between this system and the external environment condition, and each airlink processing unit all is connected to one or more radio-frequency transmitter parts by Ethernet;

Wherein, described radio-frequency transmitter parts are configured to receive the signal from a mobile unit;

Wherein, the signal of launching from described mobile unit is detected by a plurality of radio-frequency transmitter parts, and this system selects one of them radio-frequency transmitter parts to be used for handling;

Wherein, this selection course is a distributed process, and this selection course occurs on two or more processing unit of this system;

Wherein, this selection course partly occurs in described CPU, and part occurs in described airlink processing unit; And

Wherein, each airlink processing unit is only selected a received signal from one of them radio-frequency transmitter parts, and only a signal forwarding is arrived described CPU thus, carries out last signal in described CPU and selects.

15. distributed radio system according to claim 14, wherein, the available data rate of this system is depended in the distribution of selection course, in the time can obtaining the higher data rate, compare in the time of can obtaining lower data transfer rate, more signals will be forwarded to described CPU.

16. distributed radio system according to claim 14, wherein, the used distributed treatment of selection course is synchronous, and described CPU and described radio-frequency transmitter parts are synchronous.

17. distributed radio system according to claim 16, wherein, described airlink processing unit also is synchronous.

18. a distributed radio system comprises:

Wherein, this selection course is a distributed process, and this selection course occurs on two or more processing unit in this system; And

Wherein, use a select time window to set up a time restriction that is used to finish selection course in described CPU.

19. distributed radio system according to claim 18, wherein, base when described CPU comprises, itself and described airlink processing unit and radio-frequency transmitter parts are synchronous.

20. distributed radio system according to claim 18, wherein, the distributed process that described selection course is utilized is synchronous, and described CPU and radio-frequency transmitter parts are synchronous.

21. distributed radio system according to claim 20, wherein, described airlink processing unit also is synchronous.

22. a distributed radio system comprises:

Wherein, this selection course is a distributed process, and this selection course occurs on two or more processing unit in this system;

Wherein, data send to described CPU from the airlink processing unit, and be bundled into one or more Ethernets groupings, rather than be that the Ethernet that each bar radio-frequency channel sends separately divides into groups from least a portion that at least one airlink processing unit sends to the data of described CPU.

23. distributed radio system according to claim 22, wherein, this system also comprises Ethernet switch, and these Ethernet switches are configured to reduce and manage the data volume that transmits between the parts throughout by the grouping of the appointment Ethernet on the rejection designated port.

24. distributed radio system according to claim 22, wherein, the distributed process that described selection course is utilized is synchronous, and described CPU and radio-frequency transmitter parts are synchronous.

25. distributed radio system according to claim 24, wherein, described airlink processing unit also is synchronous.