A wireless network protocol header compression method for distinguishing multiple services and optimizing cross-layer

Abstract

The invention relates to the technical field of wireless communication, and relates to a method for distinguishing multi-service cross-layer optimized wireless network protocol header compression, comprising: modeling a compressor as a finite state machine with three states; modeling a decompressor as a finite state machine with three states; modeling a ROHC channel as a Markov channel with K states; based on the modeling of the compressor, the decoder and the channel, modeling the type of the next packet header to be transmitted as a partially observable Markov decision model, and obtaining the type of the next packet header to be transmitted by solving; distinguishing the optimistic parameters of the compression end according to the real-time requirements of the service types of different data streams, and adjusting the state of the ROHC compression end; the invention adaptively decides the state of the ROHC compression end, and can achieve higher efficiency and reliability without introducing a separate feedback channel, distinguishes the services, and enables the real-time services to obtain higher robustness, and the non-real-time services to obtain higher compression ratio.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a method for distinguishing multiple services and optimizing wireless network protocol header compression across layers. Background Art

[0002] Nowadays, IP has become the most widely used Internet interconnection protocol. When IP carries packets, the overhead of the IP protocol header will be very large, even higher than the proportion of the effective load, which greatly reduces the bandwidth utilization of the communication system. Compared with wired communication, the bandwidth of wireless communication is limited, and it is more necessary to reduce the protocol header overhead of the communication protocol to increase the utilization of wireless bandwidth resources. Protocol header compression is the main technical means to reduce the communication protocol overhead. To this end, 3GPP introduced the Robust Header Compression (ROHC) method from version 4 to reduce the protocol encapsulation overhead between wireless base stations and mobile clients and increase bandwidth utilization. ROHC is a protocol header compression algorithm suitable for wireless packet switching links. It is usually a compressor / decompressor module added between the data link layer and the network layer of the sender / receiver. ROHCv2 expands the function of ROHC and can support the compression of various protocols such as IP, IP / TCP, IP / UDP, etc. It has a good compression ratio. At present, ROHC is used to improve the protocol encapsulation efficiency of 4G, 5G, satellite, ad-hoc (military) and WAN networks.

[0003] There are three working modes in ROHC, among which R and O modes require a separate ROHC feedback channel to be established between the compression end and the decompression end to maintain state synchronization at both ends. U mode does not have this part of feedback overhead. The compression end uses the optimistic approximation principle and the periodicity principle to control state transition and achieve state synchronization at both ends. The present invention mainly improves ROHC in U mode.

[0004] At present, most studies are aimed at optimizing ROCH compression parameters and improving the efficiency of ROCH, but there are the following shortcomings: 1) When modeling ROHC, the compression end is rarely considered to be downgraded to the IR state, which is not suitable for wireless communication scenarios with burst errors. Burst errors in wireless communication can cause the static fields of the context to lose synchronization, making the ROHC model unable to work properly. 2) Different requirements for quality of service (QoS) of different services are not taken into account. With the widespread application of new generation mobile communication networks (such as LTE and 5G mobile communications), wireless communication networks need to support more and more types of services. To this end, 4G / 5G mobile communications use QCI (QoS class identifier) ​​attribute values ​​to distinguish different types of services. The 8th version of the standard released by 3GPP has a total of 9 services with different priorities, and the 14th version has increased to 15 services with different priorities. Generally, different services should have different requirements for real-time performance, robustness and compression ratio. Among them, robustness and compression ratio are two important indicators of protocol header compression. It is necessary to dynamically adjust the status of both ends of ROHC according to the QoS requirements of different services to balance the real-time performance, robustness and compression ratio of different services. Summary of the invention

[0005] In order to solve the above problems, ROHC can ensure a better compression ratio and performance in a bad wireless environment. The present invention proposes a method for distinguishing multi-service cross-layer optimized wireless network protocol header compression, which specifically includes the following steps:

[0006] S1. Model the compressor as a finite state machine with three states, including a refresh state, a first-order state, and a second-order state;

[0007] S2, model the decompressor as a finite state machine with three states, including a non-text state, a full context state, and an intermediate static context state;

[0008] S3. Model the ROHC channel as a Markov channel with K states, where For a good channel state, The channel is in a bad state;

[0009] S4. Based on the modeling of the compressor, decoder and channel, how to select the type of the next packet header to be transmitted is modeled as a partially observable Markov decision model, and the type of the next packet header to be transmitted is obtained by solving the model.

[0010] Furthermore, the compressor is always started in the initialization and refresh state. The initialization and refresh state indicates that the data packet header is not compressed, and context synchronization is established through the data packet. The first-order state indicates that the header of the data packet is partially compressed, and the data packet includes differential information of the dynamic domain, and the data packet partially contains differential information of the static domain. The second-order state indicates that the header of the data packet is completely compressed.

[0011] Furthermore, in the decompressor, if in a non-text state, the decompressor requires a refresh state data packet to establish context synchronization and complete the initialization process; after successful initialization, the decompressor can switch from the non-text state to the full-context state. The full-context state can decompress the refresh state, first-order state and second-order state data packets, and when the full-context state fails to decompress repeatedly, it can switch to the intermediate static context state; the intermediate static context state can also be moved up and converted to the full-text state after successfully receiving the refresh state or first-order state data packet, but when the intermediate static context state fails to decompress repeatedly, it will also be transferred down to the non-text state.

[0012] Furthermore, the decompressor has k1+k2+3 states including a non-text state, k1+1 full-context states and k2+1 intermediate static context states. When the decompressor fails to decompress k1 consecutive data packets in the full-context state, the decompressor switches to the static context state, which can only decompress data packets in the refresh state and the first-order state. When a first-order state data packet or a refresh state data packet is successfully decompressed, the state is switched to the full-context state. If decompression fails for k2 consecutive data packets, the state is switched to the non-text state.

[0013] Furthermore, after solving the next packet header type to be transmitted according to the POMDP solution, the services are differentiated according to the QCI index of the data stream, and different optimistic parameters of U-mode ROHC are used for different services. The ROHC compression end state is adjusted again, that is, if the current task is a non-real-time task, the optimistic parameters used are k1 and k2, and the current decompression end state of the partially observable Markov decision model service is s D ∈{FC 0 ,...,FC k1} or s D ∈{SC 0 ,...,SC k2}, the compression end state is not adjusted; if the task currently being processed is a real-time task, the k1 and k2 parameters are simulated as k1>>1 and k2>>1, if s D ∈{FC 0 , ..., FC k1 / 2} or s D ∈{SC 0,...,SC k2 / 2} does not adjust the compression end state, if s D ∈{FC k1 / 2+1 ,...,FC k1} or s D ∈{SC k2 / 2+1 ,...,SC k2}, then the state is converted from the second-order state to the first-order state or from the first-order state to the refresh state.

[0014] Furthermore, the partially observable Markov decision model is represented by a seven-tuple, expressed as (S, A, T, R, Ω, O, γ), where S is the set of system states, and the system state at a certain moment is the Cartesian product of the decompressor state and the channel state; A is the set of agent actions, T is the set of conditional transition probabilities between states, R is the set of reward functions, Ω is the set of observation values, O is the set of conditional observation probabilities, and γ is the set of discount factors.

[0015] Furthermore, when action a is given, the state transition matrix of the system from state s to state s′ is expressed as:

[0016] T(s,a,s′)=p(s′ H ∣s H )p(s′ D ∣s D ,a,s′ H );

[0017] Among them, T(s,a,s′) represents the state transition matrix of the system from state s to state s′ when action a is given, which belongs to the set of conditional transition probabilities between states; p(s′ H ∣s H ) indicates that the channel is from the current s H The state transfers to the next time s' H The probability of the state; p(s′ D ∣s D ,a,s′ H ) indicates that given action a, the channel state at the next moment is s′ H When , the state transition probability of the decompression end; a belongs to the set of agent actions, and state s belongs to the set of system states.

[0018] Furthermore, after executing action a, the probability of observing o in state s′ is expressed as:

[0019] O(s′,a,o)=p(o H ∣s′ H )p(o T |a,s′ D ,s′ H );

[0020] Among them, O(s′,a,o) represents the probability of observing o in state s′ after executing action a, which belongs to the conditional observation probability set; p(o H ∣s′ H ) indicates that s′ H Indicates that the channel state at the next moment is s′ H The channel state is observed when H The probability of T |a,s′ D ,s′ H ) means that given action a, the channel state at the next moment is s′ H , the decompression end state is s′ D The probability of observing the data packet transmission state at time s′ D represents the set of decompression end states at the next moment; a belongs to the set of agent actions, state s belongs to the set of system states, o belongs to the set of observation values, o H belongs to the set of observed channel states, o T Represents a collection of observed packet transmission states.

[0021] Furthermore, after executing action a, the reward function of the system from state s to state s′ is expressed as:

[0022]

[0023] Among them, R(s,a,s′) represents the reward function of the system from state s to state s′ after executing action a, which belongs to the reward function set; s′ D Indicates the decompressor state at the next moment; L P Indicates the total length of the data packet before compression; L i Indicates the total length of the compressed package when the decompressor is in state i.

[0024] Furthermore, the PBVI algorithm is used to solve the partially observable Markov decision model, that is, to maximize the expected return for each belief, and the optimization objective is expressed as:

[0025]

[0026] in, represents the optimization goal of belief b(s); a belongs to the set of agent actions; ρ(b,a)=∑ s b(s)R(s,a) represents the total expected reward, b(s) represents the confidence when the agent is in state s, and R(s,a) represents the reward for selecting action a in state s; P(o|b,a) represents the probability of obtaining observation o when the current confidence is b and the action is a; It represents the optimal expectation of confidence b′ when the decision step is n-1.

[0027] Furthermore, under the condition that the current confidence is b, the state is s, and action a is performed to obtain observation o, the probability of the next state being s' is updated to b', and the confidence b'(s') is expressed as:

[0028]

[0029] Among them, P(o|a,b) represents the probability of obtaining observation o when the current belief is b and the action is a; b(s) is the probability of obtaining observation o when the current belief is b and the state is s.

[0030] The new ROHC model proposed in the present invention uses the POMDP method to maintain a good compression ratio and efficiency even under poor wireless channel conditions, and it is obvious that the worse the channel conditions, the better the compression ratio of the model compared to existing research. When distinguishing services, for real-time services, a certain compression ratio is sacrificed in exchange for faster recovery of context confidence, which meets the real-time requirements of the service and has more advantages in throughput. For non-real-time services, there is no need to transfer to the downward state earlier, and more consideration is given to its compression ratio, and its simulation solution meets expectations. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is the ROHC protocol stack diagram in the present invention;

[0032] Figure 2 is a state transition diagram of the compressor in the present invention;

[0033] Figure 3 is a state transition diagram of the medium voltage device of the present invention;

[0034] Figure 4 is the Markov channel state transition diagram of this invention;

[0035] Figure 5 This is a functional block diagram of cross-layer optimization of robust protocol header compression of the present invention;

[0036] Figure 6 The channel error of the present invention and the prior art is 10 -3 When testing the ACL diagram;

[0037] Figure 7 The channel error of the present invention and the prior art is 10 -4 When testing the ACL diagram;

[0038] Figure 8 The channel error of the present invention and the prior art is 10 -5 When testing the ACL diagram;

[0039] Fig. 9 This is a throughput comparison chart between the present invention and the prior art. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] The present invention proposes a method for distinguishing multiple services and optimizing cross-layer wireless network protocol header compression, which specifically includes the following steps:

[0042] S1. Model the compressor as a finite state machine with three states, including a refresh state, a first-order state, and a second-order state;

[0043] S2, modeling the decompressor as a finite state machine with three states, including no context state, full context state and static context state;

[0044] S3. Model the ROHC channel as a Markov channel with K states, where For a good channel state, The channel is in a bad state;

[0045] S4. Based on the modeling of the compressor, decoder and channel, how to select the type of the next packet header to be transmitted is modeled as a partially observable Markov decision model, and the type of the next packet header to be transmitted is obtained by solving the model.

[0046] ROHC has three modes to choose from, namely, unidirectional U mode, bidirectional optimistic O mode, and bidirectional reliable R mode. Among them, O and R modes need to introduce a feedback mechanism, and the need for a separate feedback channel will increase bandwidth pressure. The unidirectional U mode does not use a feedback channel, and the decompression party cannot send feedback information to the compression party. In order to ensure compression robustness and compression rate, the compression party adopts the optimistic approximation principle and the periodicity principle for state transfer. The present invention mainly discusses the optimization of the ROHC-U mode and the use of ROHC in the U mode. The present invention predicts the state of the decompression end by using useful information obtained from the lower layer, and adaptively adjusts the ROHC compressor, thereby achieving dual-end synchronization. The present invention is suitable for use in harsh wireless environments, and the computational complexity is within the range that the system can bear, and is suitable for use in actual wireless packet links.

[0047] This embodiment models the ROHC compressor as a finite state machine (FSM) with three states, each state is represented by the type of data packet transmitted in this state, including the refresh (IR) state, the first-order (FO) state and the second-order (SO) state, and the data packets in each state are represented as refresh data packets (IR packets), first-order data packets (FO packets) and second-order data packets (SO packets). The fields of the data packet header can be roughly classified into static ones that do not change during the entire packet session (such as IP addresses), and dynamic ones that change regularly and mainly in a specific pattern (such as data packet length). The compressor always starts in the initialization and IR states, and establishes context synchronization by transmitting IR packets (whose headers are not compressed). FO packets are usually partially compressed, containing differential information of dynamic domains and rarely containing differential information of static domains, so as to effectively transmit irregular information in data packets; the header of SO packets is fully compressed.

[0048] The one-way U mode does not use a feedback channel. In this case, the decompressor cannot send feedback information to the compressor. To ensure compression robustness and compression rate, the compressor uses the optimistic approximation principle and periodicity principle for state transfer. Figure 2 In U mode, the compression state transition is controlled by the timer and counter. The compressor state is achieved by sending multiple packets in each lower state (i.e., IR and FO) and then transitioning upward to a higher-order state (FO and SO). For ease of calculation, the state of the compression end is represented as S C ={0,1,2}, 0,1,2 represent IR, SO, FO data packets respectively, P C,yx represents the probability of transitioning from state x to state y in the compressor, x, y = {0, 1, 2}.

[0049] At the beginning of the decompressor, the decompressor needs to be initialized. After initialization, it is in the No Context (NC) state, which can only decompress IR packets. Once at least one IR packet is successfully received, the context of the static and dynamic fields is successfully established by default, and the decompressor switches up to the Full Context (FC) state, where all three types of (IR, FO, SO) packets can be successfully decompressed. Figure 3 In the case of k1 consecutive data packets decompression failure, the decompressor switches to the static context (SC) state, which can only decompress IR and FO data packets. If one FO or IR packet is decompressed successfully, it is sufficient to re-establish context synchronization to move the decompressor state up to the FC state again. If k2 consecutive data packets fail to be decompressed, the static field of the decompression end is lost by default, and it is converted down to the NC state. The present invention models the decompression end as a finite state machine with k1+k2+3 states, represented as S D={0,1,...,k1,k1+1...,k1+k2+1,k1+k2+2}, where [0,k1] represents the FC in the figure 0 -FC k1 state, [k1+1, k1+k2+1] represents SC 0 -SC k2 state, k1+k2+2 represents the NC state, s T =1 and s T =0 represents the success and failure of packet transmission respectively.

[0050] The present invention models the ROHC channel as a Markov channel with K states, such as Figure 4 ,in For a good channel state, The channel state is bad, and the channel state is equivalent to S H =m,m∈{1,...,K}, where t represents the time variable. Then the transition probability of state k transforming into state m is:

[0051]

[0052] Therefore, the channel transfer probability matrix P can be expressed as a K×K matrix:

[0053]

[0054] The present invention predicts the state of the decompression end based on partial information introduced from the lower layer, and adaptively decides the state of the compression end of ROHC, thereby achieving higher efficiency and reliability without introducing a feedback channel. Whether the previous data packet was successfully transmitted can be evaluated by introducing HARQ feedback from the MAC layer, and the ROHC channel state can be estimated based on information such as CQI reports and link adaptation from the PHY layer. HARQ is related to the CRC check code, and the success rate of the current transmission can be estimated based on the number of retransmissions.

[0055] In the U-type ROHC model, the compressor does not know the explicit state information of the decompressor and cannot be sure whether the states of both ends are synchronized. The purpose of using POMDP is to allow the compressor to evaluate the state of the decompressor based on the available low-level information without the feedback of the decompressor. The basic principle is that based on the initial state of the decompressor, the transmission state and the channel state estimation, the belief in the state of the ROHC system can be continuously updated, and the optimal decision can be made on the type of packet header to be transmitted next. This problem can be reduced to a partially observable Markov decision process, that is, the agent takes a series of actions to maximize the reward under environmental uncertainty. POMDP can be represented by a seven-tuple: (S, A, T, R, Ω, O, γ), where S is the set of system states, A is the set of agent actions, T is the conditional transition probability between states, R is the reward function, Ω is the observation value, O is the conditional observation probability, and γ∈[0,1] is the discount factor.

[0056] In any period of time, the environment is in a certain state s∈S. The present invention takes the Cartesian product of the decompressor state and the channel state as the system state:

[0057] s=(s D ,s H )∈S=S D ×S H ;

[0058] The action taken by the agent is This results in the probability of the environment transitioning to state s′ being, given action a, the state transition matrix from s to s′:

[0059] T(s,a,s′)=p(s′ H ∣s H )p(s′ D ∣s D ,a,s′ H );

[0060] At the same time, the agent receives an observation o = (o T ,o H )∈Ω=Ω T ×Ω H , where o T ∈Ω T ={0,1} and S T They all represent the transmission status, Ω H =S H ={0,...,K-1} represents the channel state. This affects the new state s' of the environment and the action a taken by the agent. The observation function is: After executing action a, the probability of observing o in state s' is:

[0061] O(s′,a,o)=p(oH ∣s′ H )p(o T |a,s′ D ,s′ H )

[0062] Finally, the agent receives a reward R(s,a,s′) whose function is:

[0063]

[0064] like Figure 3 , when the decompression is successful D The status is FC 0 , and s D =0 means the decompression is successful, L P Represents the total length of the data packet before compression, L i , i=1,2,3 represent s respectively c In IR, FO, SO state, the load of ROHC data packet, and L P / L i represents the compression ratio of the current data packet. D = 0, indicating successful decompression, l(s′ D =0) is the D =0, the function value is 1, otherwise it is 0.

[0065] When the POMDP model is known, the agent can update its belief b′(s′) when interacting with the environment. Under the condition that the current belief is b and the agent takes action a and obtains observation o, the probability of the next state being s′ is:

[0066]

[0067] The idea behind the PBVI algorithm is to approximate the solution of the POMDP model. The algorithm flow of PBVI is as follows:

[0068] The goal of POMDP is to find a sequence of actions {a 0 ,...,a t} to maximize the expected return, expressed as:

[0069]

[0070] Of course, since each state in the POMDP cannot be fully observed, it is necessary to maximize the expected return for each belief. The optimization goal is to maximize the discounted cumulative reward, and its value function can be expressed as:

[0071]

[0072] in

[0073] After n iterations, the value function can be expressed as a series of α-vectors: V n ={α 0 ,α 1 ,...,α m}. Each α-vector defines a value function in a certain interval, expressed as:

[0074]

[0075] In POMDP value iteration, the value function of the nth plateau can be constructed from the value function of the n-1th plateau by backtracking, using operator H:

[0076] V=HV′

[0077]

[0078] To implement the above steps, an intermediate is generated for each action observation set:

[0079] Γ a,* ←α a,* (s) = R(s,a)

[0080]

[0081] Since only a limited number of belief points need to be considered, only a limited set of points need to be calculated, and the complexity is greatly simplified. The formula is:

[0082]

[0083] Finally, we get the optimal action for each belief point:

[0084]

[0085] like Figure 3 As shown in the figure, the decompression end is modeled into k1+k2+3 states. When the decompressor is in the FC state, it fails to decompress k1 packets in a row and is converted to the static context (SC) state. If it fails to decompress k2 packets in a row, the static field of the decompression end is lost by default and it is converted to the NC state. The values ​​of k1 and k2 determine the speed of restoring synchronization of the context at both ends and the compression ratio of the ROHC module. The smaller the k1 and k2, the earlier the system performs state transfer when the quality of the wireless link is worse, ensuring that the two ends will not lose synchronization and improve the decompression success rate. The larger the k1 and k2, the fewer times the ROHC module will transfer to the next state, and the system will have a higher bandwidth utilization.

[0086] After solving POMDP, the services are divided into SRT =1 for non-real-time services and S RT =0 real-time business.

[0087] When S RT =1, that is, the current service type is non-real-time service, and the parameters used are k1 and k2. After POMDP prediction, if s D ∈{FC 0 ,...,FC k1} or s D ∈{SC 0 ,...,SC k2}, the compression end state is not adjusted, ensuring that the service has a higher compression ratio while maintaining a certain level of robustness.

[0088] For S RT =0, that is, real-time services. For services with high real-time requirements, the real-time performance of the data packet is considered more than the compression ratio, and the downward state transition is performed earlier. The k1 and k2 parameters are simulated as k1>>1 and k2>>1. That is, judging s D ∈{FC 0 ,...,FC k1 / 2} or s D ∈{SC 0 ,...,SC k2 / 2} does not adjust the compression end state, such as s D ∈{FC k1 / 2+1 ,...,FC k1} or s D ∈{SC k2 / 2+1 ,...,SC k2} state, then s C Adjusting from SO, FO status to FO, IR status enables both ends to return to synchronization faster, ensuring the real-time nature of the business.

[0089] The present invention proposes a new ROHC model and a new cross-layer ROHC optimization algorithm for distinguishing multiple services in U mode. The algorithm not only introduces underlying information feedback to achieve adaptive protocol header compression, but also distinguishes different services, so that high-priority services (high real-time performance) have faster recovery capabilities and low-priority services have higher compression ratios.

[0090] In order to verify the effectiveness of the algorithm, the present invention will be compared with the traditional U-ROHC and Markov prediction model to verify the effectiveness of the algorithm of the present invention. The present invention uses a PC (Intel Core i5-7500 CPU, 16GBDDR3) to open the Ubuntu16 operating system of VMare Workstation 15, and integrates the ROHC module into the NS-3 open source network simulator to perform cross-layer simulation analysis on ROHC. The transition probabilities in the POMDP instance can all be calculated using statistical methods. The simulation uses the average header compression length (ACL) to evaluate the efficiency, where T represents the total number of packet bytes, P represents the number of bytes of the payload, and C represents the total number of packets sent. The smaller the average protocol header length, the higher the compression ratio.

[0091]

[0092] The simulation is performed under the condition that the channel bit error rate is 10 -3 ,10 -4 ,10 -5 , 10,000 packets are sent, each of which is 40 bytes long, and the first sampling point is to send 100 packets, and then the sampling analysis is performed at each point where 1,000 packets are sent. The simulation results are as follows: Figure 6-8 .

[0093] like Figure 6 As shown, when the channel bit error rate is 10 -3 When the bit error rate is high, the simulation results show that the average compression length of the traditional U-ROHC when transmitting the 10,000th packet is 10.85 bytes, and the ACL predicted by Markov to be stable during transmission is 9.47 bytes. The POMDP model of real-time services will perform state transitions relatively quickly due to real-time reasons, resulting in the ACL being 0.25 bytes higher than the result predicted by Markov when it is stable. The ACL of the POMDP model of non-real-time services when it is stable is about 0.6 bytes less than the ACL predicted by Markov and about 1.7 bytes less than the traditional U-ROHC.

[0094] like Figure 7 As shown, when the channel bit error rate is 10 -4 The simulation results show that the average compression length of the traditional U-ROHC when transmitting the 10000th packet is about 7.10 bytes, and the ACL that tends to be stable during transmission is about 6.10 bytes. The number of downward state transitions of the POMDP model of real-time services under the condition of this bit error rate is relatively low compared to 10 -3When the compression ratio is less, the compression ratio mainly benefits from the POMDP model. When it tends to be stable, it is about 0.08 bytes lower than the ACL of the Markov prediction model. The POMDP model for non-real-time services takes the compression ratio into consideration more. When it tends to be stable, the ACL is about 0.3 bytes less than the ACL of the Markov prediction model and about 1.3 bytes less than the traditional U-ROHC.

[0095] Figure 8 As shown, when the channel bit error rate is 10 -5 The simulation results show that the average compression length of the traditional U-ROHC when transmitting the 10000th packet is about 6.19 bytes, and the ACL that tends to be stable during transmission is about 5.45 bytes. The number of downward state transitions of the POMDP model of real-time services under the condition of this bit error rate is relatively low compared to 10 -4 When the ACL is stable, it is about 0.06 bytes lower than the ACL of the Markov prediction model. The POMDP model for non-real-time services takes the compression ratio into consideration more. When the ACL is stable, it is about 0.2 bytes lower than the ACL of the Markov prediction model and about 0.94 bytes lower than the traditional U-ROHC.

[0096] The present invention conducts a throughput test on the algorithm to verify whether the real-time requirements for distinguishing multiple services meet expectations. Fixed bytes are sent, and each data packet relies on retransmission to ensure successful transmission. The throughput is calculated after obtaining the simulation completion time. The channel bit error rate is set to 10 -5 , 300 data streams are sent at the sending end, each data stream is grouped into 10,000 packets, and the payload of each packet is 40 bytes. The sending time is calculated, where the total amount sent is:

[0097] 40B×1000×300=12MB;

[0098] The throughput is calculated as:

[0099]

[0100] The throughput is closely related to the end-to-end delay of the system.

[0101] like Fig. 9 As shown, at a bit error rate of 10 -5 When the probability of continuous packet loss is small, the throughput of real-time services is less increased by POMDP than that of non-real-time services. Although the average compression length of POMDP for non-real-time services is lower than that of the Markov prediction model [5], it has no advantage in throughput due to the complexity of the algorithm. -4When continuous packet loss occurs occasionally, the throughput of POMDP for real-time services increases significantly compared with that of non-real-time services. -3 When continuous packet loss often occurs, the POMDP for real-time services has a faster downward state transition and faster recovery of context confidence, and its throughput is significantly increased compared with the non-real-time services and the Markov prediction model. However, the POMDP for non-real-time services is not as good as the Markov prediction model in terms of throughput due to its complexity, but it has a higher compression ratio.

[0102] The simulation analysis shows that the new ROHC model proposed by the present invention can maintain a good compression ratio and efficiency even when the wireless channel conditions are bad by using the POMDP method. It is obvious that the worse the channel conditions are, the better the compression ratio of the model is compared with the existing research. When distinguishing services, for real-time services, a certain compression ratio is sacrificed in exchange for faster recovery of context confidence, which meets the real-time requirements of the services and has more advantages in throughput. For non-real-time services, there is no need to transfer to the downward state earlier, and the compression ratio is considered more. The simulation solution meets expectations.

[0103] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for distinguishing multiple services and optimizing cross-layer wireless network protocol header compression, characterized in that: The specific steps include: S1. Model the compressor as a finite state machine with three states, including a refresh state, a first-order state, and a second-order state; S2, model the decompressor as a finite state machine with three states, including non-text state, full context state and static context state; S3. Model the ROHC channel as a Markov channel with K states, where For a good channel state, The channel is in a bad state; S4. Based on the modeling of the compressor, decoder and channel, how to select the type of the next packet header to be transmitted is modeled as a partially observable Markov decision model. By solving the model, the type of the next packet header to be transmitted, that is, the compression end state, is obtained. The partially observable Markov decision model is represented by a seven-tuple, expressed as (S, A, T, R, Ω, O, γ), where S is the set of system states, and the system state at a certain moment is the Cartesian product of the decompressor state and the channel state; A is the set of agent actions, T is the set of conditional transition probabilities between states, R is the set of reward functions, Ω is the set of observation values, O is the set of conditional observation probabilities, and γ is the set of discount factors; S5. Distinguish the data flow service type according to the QCI index, and adjust the ROHC compression end state again, that is, after solving the next packet header type that needs to be transmitted according to the POMDP solution, distinguish the service according to the QCI index of the data flow, use different U-mode ROHC optimistic parameters for different services, and adjust the ROHC compression end state again, that is, if the current task is a non-real-time task, the optimistic parameters used are k1 and k2, and the state of the current decompression end of the partially observable Markov decision model service is s D ∈{FC0,...,FC k1 } or s D ∈{SC0,...,SC k2 }, the compression end state is not adjusted; if the task currently being processed is a real-time task, the k1 and k2 parameters are simulated as k1>>1 and k2>>1, if s D ∈{FC0,...,FC k1 / 2 } or s D ∈{SC0,...,SC k2 / 2 } does not adjust the compression end state, if s D ∈{FC k1 / 2+1 ,...,FC k1 } or s D ∈{SC k2 / 2+1 ,...,SC k2 }, then the second-order state is converted down to the first-order state or from the first-order state is converted down to the refresh state.

2. According to claim 1, a method for distinguishing multiple services and optimizing cross-layer wireless network protocol header compression is characterized in that: The compressor is always started in the initialization and refresh states. The initialization and refresh states indicate that the data packet header is not compressed, and context synchronization is established through the data packet. The first-order state indicates that the data packet header is partially compressed, and the data packet includes differential information in the dynamic domain, and the data packet partially contains differential information in the static domain. The second-order state indicates that the data packet header is completely compressed.

3. According to claim 1, a method for distinguishing multiple services and optimizing cross-layer wireless network protocol header compression is characterized in that: In the decompressor, if it is in a non-text state, the decompressor needs a refresh state data packet to establish context synchronization and complete the initialization process; after successful initialization, the decompressor can switch from the non-text state to the full-context state. The full-context state can decompress the refresh state, first-order state and second-order state data packets, and when the full-context state fails to decompress repeatedly, it can switch to the intermediate static context state; the intermediate static context state can also be moved up and converted to the full-text state after successfully receiving the refresh state or first-order state data packet, but the intermediate static context state fails to decompress repeatedly. It will also move down to the non-text state; the decompressor has k1+k2+3 states including a non-text state, k1+1 full-context state and k2+1 intermediate static context state. When the decompressor fails to decompress k1 consecutive data packets in the full-context state, the decompressor switches to the static context state, which can only decompress data packets in the refresh state and the first-order state. When a first-order state data packet or a refresh state data packet is successfully decompressed, the state is converted to the full-context state; if the decompression of k2 consecutive data packets fails, the state is converted to the non-text state.

4. According to claim 1, a method for distinguishing multiple services and optimizing cross-layer wireless network protocol header compression is characterized in that: When an action a is given, the state transition matrix of the system from state s to state s′ is expressed as: T(s,a,s′)=p(s′ H ∣s H )p(s′ D ∣s D ,a,s′ H ); Among them, T(s,a,s′) represents the state transition matrix of the system from state s to state s′ when action a is given, which belongs to the set of conditional transition probabilities between states; p(s′ H ∣s H ) indicates that the channel is from the current s H The state transfers to the next time s' H The probability of the state; p(s′ D ∣s D ,a,s′ H ) indicates that given action a, the channel state at the next moment is s′ H , the state transition probability of the decompression end; a belongs to the set of agent actions, state s belongs to the set of system states; s′ D Represents the set of decompression end states at the next moment; s D A collection representing the current state of the decompression end.

5. The method for compressing wireless network protocol headers by differentiating multiple services and optimizing cross-layer according to claim 1, characterized in that: After executing action a, the probability of observing o in state s′ is expressed as: O(s′,a,o)=p(o H ∣s′ H )p(o T ∣a,s′ D ,s′ H ); Among them, O(s′,a,o) represents the probability of observing o in state s′ after executing action a, which belongs to the conditional observation probability set; p(o H ∣s′ H ) indicates that the channel state at the next moment is s′ H The channel state is observed when H The probability of T |a,s′ D ,s′ H ) means that given action a, the channel state at the next moment is s′ H , the decompression end state is s′ D The probability of observing the data packet transmission state at , s′ D represents the set of decompression end states at the next moment; a belongs to the set of agent actions, state s belongs to the set of system states, o belongs to the set of observation values, and o H belongs to the set of observed channel states, o T Represents a collection of observed packet transmission states.

6. The method for compressing wireless network protocol headers by differentiating multiple services and optimizing cross-layers according to claim 1, characterized in that: After executing action a, the reward function of the system from state s to state s′ is expressed as: Among them, R(s,a,s′) represents the reward function of the system from state s to state s′ after executing action a, which belongs to the reward function set; s′ D Indicates the decompressor state at the next moment; L P Indicates the total length of the data packet before compression; L i Indicates the total length of the compressed package when the decompressor is in state i.

7. The method for compressing wireless network protocol headers by differentiating multiple services and optimizing cross-layers according to claim 1, characterized in that: The PBVI algorithm is used to solve the partially observable Markov decision model, that is, to maximize the expected return for each belief. The optimization objective is expressed as: in, represents the optimization goal of belief b(s); a belongs to the set of agent actions; ρ(b,a)=∑ s b(s)R(s,a) represents the total expected reward, b(s) represents the confidence when the agent is in state s, and R(s,a) represents the reward for selecting action a in state s; P(o|b,a) represents the probability of obtaining observation o when the current confidence is b and the action is a; It represents the optimal expectation of confidence b′ when the decision step is n-1.

8. The method for compressing wireless network protocol headers by differentiating multiple services and optimizing cross-layers according to claim 7, characterized in that: Under the condition that the current confidence is b, the state is s, and action a is performed to obtain observation o, the probability of the next state being s' is updated to b', and the confidence b'(s') is expressed as: Among them, P(o|a,b) represents the probability of obtaining observation o when the current belief is b and the action is a; b(s) is the probability of the current belief is b and the state is s; O(s′,a,o) represents the probability of observing o in state s′ after executing action a, which belongs to the conditional observation probability set; T(s,a,s′) represents the state transition matrix of the system from state s to state s′ when action a is given, which belongs to the conditional transition probability set between states.

Images