High speed downlink packet access (HSDPA) data structure optimizing and processing method of user equipment (UE) side
Technical field
The present invention relates to mobile communication system HSDPA (High Speed Downlink Packet Access, high speed downlink packet inserts) technical field, relate in particular to high speed downlink packet access (HSDPA) data structure optimizing and processing method of user equipment (UE) side.
Background technology
In third generation wireless mobile communications standard, Media Access Control MAC is positioned on the physical layer, mainly be at transmission channel that physical layer provides and provide to rlc layer between the logic channel of service and carry out the channel mapping, also be the suitable transformat (TF) of logic channel selection simultaneously.The MAC layer includes three functional entity: MAC-d, MAC-c/sh, MAC-b.They utilize SAP to realize corresponding channel mapping.These entities are by controlling the configuration messages that SAP accepts to come from RRC, simultaneously to rrc layer feedback states measurement report.
The major function of each entity is as follows:
MAC-b mainly is responsible for handling cell broadcast messages;
MAC-c/sh mainly is responsible for handling common signal channel and the shared channel in the residential quarter;
MAC-d mainly is responsible for handling the dedicated channel of distributing to UE under connection mode.
On the realization position of these entities, UE side and UTRAN side are different.The UE side has only MAC-d, a MAC-c/sh, one or more MAC-b entity.Because they in same equipment, therefore communicate by internal interface each other.For the UTRAN side, MAC-b is arranged in Node B, and has only one among each CELL; MAC-c/sh is arranged in CRNC; MAC-d is arranged in SRNC, and provides a MAC-d entity for each UE.Because according to the division of functions of the equipments, these entities adhere to separately in the different equipment, must communicate by standard interface between each entity.As MAC-d and MAC-c/sh not under the situation of same RNC, the two undertaken by the Iur interface alternately.These interfaces have been designed to standard open, so that the equipment of different manufacturers is compatible each other.
Because the rapid increase of wireless data service, the asymmetry of up-downgoing traffic carrying capacity can highlight more.In addition, system itself must have more and be fit to some characteristics of transmitting data service, as high data volume, high sudden, high reliability etc.To this, in the evolution of 3G (Third Generation) Moblie technology, introduced an important enhancing technology in 3GPP Release 5 Version Specs---HSDPA (High Speed Downlink PacketAccess).HSDPA is the set of some wireless enhanced technology, utilizes the HSDPA technology that the downlink data peak rate is improved a lot.The HSDPA technology is applicable to WCDMA FDD, UTRA TDD and three kinds of different modes of TD-SCDMA simultaneously.Implementation in different system is quite similar.
HSDPA strengthens systematic function by revising air interface, and considers to operate fast and dispatch.So main operation is all finished at physical layer and the MAC layer of UE and Node B.Because the introducing of HSDPA technology has added a functional entity MAC-hs again at the MAC layer, be responsible for carrying out the function of relevant HSDPA.No matter be in the UE side or in Node B side, the MAC-hs entity of increase is a very important entity, and the MAC layer operation of relevant HS-DSCH is all here finished.Except the Flow Control and processed of necessity, the associative operation that will finish the HARQ agreement at first comprises scheduling, re-transmission, rearrangement etc.
UE side MAC-hs mainly comprises and mixes automatic repeat requests (HARQ) entity and reset entity.
The HARQ entity is responsible for handling the HARQ agreement.In each TTI, each HS-DSCH should have a HARQ process.The HARQ functional entity is handled all tasks relevant with HARQ, for example is responsible for handling ACK or NACK.The detailed configuration information of HARQ is that RRC is undertaken by MAC-Control SAP.
Reset entity and sort according to the data block of TSN that receives, and data block in proper order is submitted to high level successively.But being to use timer can will not be that continuous data delivery is given high-rise.In the UE side, for each priority and transmission channel the entity that reorders is arranged.
Fig. 1 shows MAC-hs PDU data structure among the 3GPP, and Fig. 2 shows MAC-hs functional module division among the 3GPP.According to 3GPP and CCSA related protocol, for HSDPA, a MAC-hs PDU is made up of MAC-hs packet header and a plurality of MAC-hs SDU, and a MAC-hs SDU comprises a MAC-d PDU, and a MAC-d PDU comprises a MAC-d SDU.Generally speaking, because the existence in MAC-hs packet header, the MAC-d SDU among the MAC-hs SDU is not byte-aligned.The formation distribution of MAC-hs, formation are reset and are unpacked to handle and be to use software to handle.
According to the scheme that related protocol 3GPP and CCSA is described, MAC-hs PDU unpacks the bit shifting function that mass data unavoidably occurs.Under some extreme case, the MIPS in the time of operation (abbreviation of MillionInstructions Per Second), 1,000,000 grades the machine language instruction number that per second is handled.This is an index weighing CPU speed.Have rapid variation.In order to satisfy these opposite extreme situations, just need the more MCU of high primary frequency (Microprocessor Control Unit) microcontroller, must bring the increase of cost and the increase of power consumption like this, otherwise certain instantaneous stability of system can be subjected to very big influence even become unstable.The disposition of this HSDPA not only is present in FDD, also is present in LCR-TDD, and 3.84Mcps TDD and 7.86Mcps TDD not only are present in single carrier LCR TDD, also are present in multicarrier LCR-TDD.
Summary of the invention
The purpose of this invention is to provide high speed downlink packet access (HSDPA) data structure optimizing and processing method of user equipment (UE) side, solve according to the scheme according to 3GPP and CCSA related protocol, the data processing amount that occurs during the HSDPA data are handled has the big and problem of unstable of change.
Technical scheme of the present invention is that a kind of high speed downlink packet access (HSDPA) data structure optimizing and processing method of user equipment (UE) side is characterized in that, may further comprise the steps:
A1, acquisition MAC-hs PDU determine MAC-hs head length degree, send and mix automatic repeat requests;
A2, described MAC-hs PDU is stored in MAC-hs right-aligned mode;
A3, to described MAC-hs PDU split successively, queue and rearrangement distribution and resetting, obtain MAC-dPDU and be sent to MAC-d.
Described MAC-hs PDU is split in the described steps A 3 comprised described MAC-hsPDU packet header is resolved and the byte-aligned of MAC-hs SDU.
In described steps A 2, if contain the C/T territory, after described MAC-hsPDU is stored in MAC-hs right-aligned mode, the MAC-hs SDU of half quantity among the MAC-hs PDU is carried out the displacement of 4 bits among the described MAC-hs PDU.
Serve as that the processing unit carries out described steps A 3 to described MAC-hs PDU with default MAC-hs PDU quantity one.
A kind of high speed downlink packet access (HSDPA) data structure optimizing and processing method of user equipment (UE) side is characterized in that, may further comprise the steps:
B1, acquisition MAC-hs PDU determine MAC-hs head length degree, send and mix automatic repeat requests;
B2, described MAC-hs PDU is stored in MAC-hs right-aligned mode;
B3, described MAC-hs PDU is carried out queue and rearrangement distribution successively, resets and splits, obtain MAC-dPDU and be sent to MAC-d.
Described MAC-hs PDU is split among the described step B3 comprised described MAC-hsPDU packet header is resolved and the byte-aligned of MAC-hs SDU.
In described step B2, if contain the C/T territory, after described MAC-hsPDU is stored in MAC-hs right-aligned mode, the MAC-hs SDU of half quantity among the MAC-hs PDU is carried out the displacement of 4 bits among the described MAC-hs PDU.
Serve as that the processing unit carries out described step B3 to described MAC-hs PDU with default MAC-hs PDU quantity two.
Specifically, because forming form, MAC-hs PDU data structure can be divided into two classes:
1) do not comprise the C/T territory among the MAC-hs PDU;
2) contain the C/T territory among the MAC-hs PDU;
To above-mentioned first kind MAC-hs PDU, adopt MAC-hs right-aligned mode after because MAC-hsSDU itself is byte-aligned, data do not need to be shifted, and can directly obtain required data sequence.For the second class MAC-hs PDU, adopt MAC-hs right-aligned mode after because the existence in C/T territory, MAC-hs SDU length is the integral multiple of 4bit always.Therefore only need the MAC-hs SDU of half quantity among the MAC-hs PDU is carried out the 4bit displacement, just can obtain required data sequence.
The scheme that the present invention proposes will receive that the data structure of MAC-hs PDU carries out suitable optimization process, deposit overall data in MAC-hs right-aligned mode, receive that so whenever a MAC-hs PDU just substantially effectively avoids or reduced the operation of data bit displacement, thereby can solve the consumption of MIPS and the demand of reduction extreme case MIPS of following moment effectively.
Description of drawings
Fig. 1 shows MAC-hs PDU data structure schematic diagram among the 3GPP
Fig. 2 shows MAC-hs functional module division schematic diagram among the 3GPP
Fig. 3 shows the MAC-hs process flow figure of one embodiment of the present of invention
Fig. 4 shows MAC-hs PDU does not have C/T numeric field data form schematic diagram
Fig. 5 shows MAC-hs PDU even number C/T numeric field data form schematic diagram
Fig. 6 shows MAC-hs PDU odd number C/T numeric field data form schematic diagram
Fig. 7 shows improvement MAC-hs PDU does not have C/T numeric field data form schematic diagram
Fig. 8 shows improvement MAC-hs PDU even number C/T numeric field data form schematic diagram
Fig. 9 shows improvement MAC-hs PDU odd number C/T numeric field data form schematic diagram
Embodiment
Elaborate below in conjunction with the embodiment of accompanying drawing to technical solution of the present invention.
Suppose that the MAC-hs PDU that receives comprises three MAC-hs SDU (the C/T territory exists when there be multiplexing the relation in many dedicated logical channel and a MAC-d stream, has increased the complexity that data are handled the bit displacement).The present invention does not limit the composition mode of MAC-hs PDU.
During no C/T territory, it is as follows to adopt the solution of the present invention data structure to handle:
1) finish the parsing of MAC-hs head, MAC-hs SDU has been byte-aligned.
2) with MAC-hs SDU1,2,3 (being MAC-d PDU) are issued MAC-d.
It is as follows to adopt the solution of the present invention data structure to handle when the C/T territory is arranged:
1) finishes the parsing of MAC-hs head
2) MAC-hs SDU1 byte-aligned is issued MAC-d with MAC-hs SDU1 (being MAC-d PDU).
3) with the displacement of MAC-hs SDU24 bit, make MAC-hs SDU byte align.
4) MAC-hs SDU2 (being MAC-d PDU) is issued MAC-d.
5) MAC-hs SDU3 byte-aligned is issued MAC-d with MAC-hs SDU3 (being MAC-d PDU).
Can certainly and need the MAC-hsSDU of bit displacement to handle respectively with the MAC-hs SDU of byte-aligned, the present invention limit concrete processing mode.
Processing method is similar when a MAC-hs PDU comprises a plurality of MAC-hs SDUs or during a MAC-hs pdu header length variations or during MAC-hs SDU length variations.In order to further specify the essence technical scheme of above-mentioned processing method, having provided MAC-hs PDU in Fig. 4 does not have the example of C/T numeric field data form, having provided MAC-hs PDU in Fig. 5 has the example of even number C/T numeric field data form, having provided MAC-hs PDU among Fig. 6 has the example of odd number C/T numeric field data form, Fig. 7 has provided the example that adopts MAC-hs PDU after the technical solution of the present invention not have C/T numeric field data form, Fig. 8 has provided the example that adopts MAC-hs PDU after the technical solution of the present invention that even number C/T numeric field data form is arranged, and Fig. 9 has provided the example that adopts MAC-hs PDU after the technical solution of the present invention that odd number C/T numeric field data form is arranged.
For the beneficial effect that adopts technical scheme of the present invention is described, be example with single carrier LCR TDD 2.8MbpsHSDPA, provide the example of an extreme case.This example is just in order to illustrate a kind of extreme case, but do not represent all extreme cases, and this example does not limit the scope of application of this invention yet.Example is specific as follows:
T1 is made as 400ms;
TSN_rx-window_size is made as 32;
MAC-hs PDU size is 14043bits,
TSN=0 does not correctly receive;
TSN=1,2 until TSN31 correctly receive;
TSN=0 receives again then, the current reception correctly;
For each MAC-hs SDU, all need to do the bit displacement, so that MAC-hs SDU is byte-aligned.
Like this, according to the protocol description of 3GPP, there are continuous 32 MAC-hs PDU to unpack at most.Being located at the MIPS that unpacks that finishes a MSC-hs PDU in the 5ms separately is m, and then continuous 32 MIPS demands that unpack processing are 32 * m in 5ms.Attention: the consumption of this part MIPS is mainly caused by the bit displacement that data are handled.Real system also needs to consider the processing of other partial function modules, therefore must dispose in the longest 2.5ms, and so only needed MIPS of this step is minimum is 2 * 32 * m.M increases along with the increase of downstream rate in the real system, and the value of m is about 10 in the time of 2.8Mbps, and then required MIPS of this step is 2 * 32 * 10=640 at least.Obviously, in order to satisfy this extreme case, adopt high performance MCU like this not only to improve cost and also improved power consumption, but if do not improve the performance of MCU, then will inevitably cause the instantaneous MIPS of system not enough, reduce the stability of system.
Adopt the present invention can significantly reduce the m value.When not having the C/T territory, because data need not the bit displacement substantially, MIPS was about 2 * 32 * 0=0 when m leveled off to 0,2.8Mbps, and totally consuming MIPS can ignore.M was half of original value when the C/T territory was arranged, and MIPS is 2 * 32 * 5=320 in the time of 2.8Mbps, effectively reduces peak value and handles required MIPS.
Here the value of m is got 10 and just to be provided an example in order saying something, and not mean that m must be 10.The 2.5ms here also just provides an example, and does not mean that it must is 2.5ms, and how long reality also depends on implementation.
As seen according to technical solution of the present invention, UE side HSDPA data are handled and can be significantly reduced the possibility that the bit displacement occurs, and improve correlated performance.
Considering further at MAC-hs SDU has under the situation of C/T territory, has the possibility of a half data being carried out the 4bit displacement, and according to protocol description, MAC-hs entity handles flow process such as a mistake! Do not find Reference source.2, the distribution of formation and rearrangement process are not partly taked any operation to the SDU of MAC-hs.Resolve and MAC-hs SDU bit displacement and formation distribution and the no any relation of formation rearrangement in the packet header of MAC-hsDisassembly.
Therefore, as Fig. 3 mistake! Do not find Reference source.Shown in, it is as follows to adopt data procedures to handle:
Among the step S301, finish the HARQ function of 3GPP agreement;
Among the step S302, finish two step of MAC-hs PDU Disassembly, i.e. the byte-aligned of MAC-hs pdu header parsing and MAC-hs SDU splits MAC-hs PDUs, does the bit shifting processing, obtains MAC-d PDUs;
Among the step S303, finish the formation distribution function;
Among the step S304, finish formation and reset function;
Among the step S305, the 3rd step of finishing MAC-hs Disassembly is about to the function that MAC-d PDU issues MAC-d.
In the above step, the output of step S302 is the input of step S303 just, need satisfy the requirement of the MAC-d SDU byte-aligned among the MAC-hsSDU, also can determine whether to be kept at the packet header analysis result of MAC-hs among the step S302 simultaneously as required and preserve form to avoid in step S305, continuing to resolve the MAC-hs pdu header.
When the C/T territory was arranged by this method, MIPS was 2 * 1 * 5=10 in the time of 2.8Mbps, further reduced peak value greatly and handled required MIPS.