A short packet rate division multiple access optimization method

Abstract

The application discloses a short packet rate division multiple access optimization method, which comprises the following steps: dividing a message into a public message and a private message; encoding the public message and the private message of all users into a public data stream and a private data stream respectively; linearly pre-encoding the public data stream and the private data stream; deriving corresponding traversal and rate lower bounds of the public data stream and the private data stream according to the moving speed of all users and the statistical characteristics of a channel, optimizing a public data stream power allocation factor, a private data stream power allocation factor and a public data stream rate allocation proportion; multiplying each data stream by a corresponding pre-encoding matrix and a power allocation factor, and obtaining a base station transmitting signal after superimposing all signals; decoding the public data stream and the private data stream of the user in sequence, and reconstructing all information of the user. The application can significantly improve the spectrum utilization under non-ideal CSI, and meet the strict service quality requirements in ultra-reliable low-latency communication.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, in particular to a short packet rate division multiple access optimization method. BACKGROUND

[0002] Due to the fact that autonomous vehicles must operate in complex and dynamic environments, the demand for high reliability and low latency communication becomes particularly urgent. Ultra-reliable low-latency communication (xURLLC) in the sixth generation (6G) communication technology can provide almost perfect communication reliability and extremely low end-to-end delay, which is crucial for supporting real-time decision-making of autonomous vehicles. To meet this demand, short data packet and limited block length (FBL) coding technology has become an effective solution, which can significantly reduce communication delay while ensuring reliability, especially suitable for real-time data transmission in autonomous driving environments.

[0003] In practical applications, it is particularly difficult to obtain accurate channel state information (CSI) in high mobility environments, especially in the communication between base stations and vehicles; the Doppler effect and signal delay caused by the high-speed movement of vehicles can cause a significant decline in signal quality, thereby affecting the stability and reliability of communication. In this case, advanced technologies such as rate splitting multiple access (RSMA) can effectively deal with interference and optimize signal transmission. In existing technologies related to short packet rate division multiple access, such as the short packet rate division multiple access method and system suitable for high mobility scenarios disclosed in patent publication CN117858234A, which aims to solve the problem that existing technologies cannot effectively deal with the decline in communication quality caused by high mobility of vehicles in short packet rate division multiple access vehicle networking communication. The problem with this solution is that the system model assumes that the large-scale fading of all vehicles is the same, which ignores the changes in distance between vehicles and base stations in actual applications, as well as the signal fading effect at different locations; it also uses an exhaustive method to optimize resource allocation, resulting in high computational complexity and making it difficult to be applied efficiently in actual systems. In addition, it does not consider the non-uniform allocation of private data streams and the allocation proportion of public data streams, which may lead to uneven traffic allocation, further affecting communication efficiency and fairness. If the large-scale fading of different vehicle locations and dynamic channel conditions are not specifically designed, the communication quality will be severely affected.

[0004] Therefore, the existing technology has not yet provided a practical, effective and operable solution, especially in terms of handling vehicle location changes, channel fading and traffic allocation in high mobility systems, which still needs further improvement. SUMMARY

[0005] In order to overcome the defects and deficiencies existing in the prior art, the present application provides a short packet rate division multiple access optimization method, which adopts a short packet based rate division multiple access technical solution, perfects the parameter settings such as power allocation and rate allocation of public and private data streams, solves the technical problems of ultra-low latency and ultra-low block error rate in communication, and achieves the technical effect of significantly improving the spectrum utilization rate under non-ideal CSI. It can effectively improve the system performance and rate under the premise of meeting the strict quality of service (QoS) requirements in ultra-reliable low latency communication (URLLC).

[0006] In order to achieve the above purpose, the present application adopts the following technical solutions:

[0007] The present application provides a short packet rate division multiple access optimization method, comprising the following steps:

[0008] Obtain the message sent by the base station to the user, and divide the message into public message and private message;

[0009] Encode the public message of all users into a public data stream, and encode the private message of each user into a private data stream;

[0010] Linearly precode the public data stream and the private data stream to obtain the corresponding precoding matrix;

[0011] According to the statistical characteristics of the moving speed and the channel of all users, the corresponding traversal and rate lower bound of the public data stream and the private data stream are derived, and the public data stream power allocation factor, the private data stream power allocation factor, and the public data stream rate allocation ratio are optimized to obtain the corresponding power allocation factor of each data stream and the rate allocation of the public data stream;

[0012] Multiply each data stream by the corresponding precoding matrix and power allocation factor, and superimpose all signals to obtain the base station transmitting signal;

[0013] After receiving the signal, the receiving end decodes the public data stream by regarding all private data streams as noise, and separates the public part information of the receiving end itself;

[0014] Subtract the corresponding part of the public data stream from the received signal, and decode the private data stream of the receiving end by regarding the private data stream from other receiving ends as noise according to the subtracted signal, to obtain the private part information;

[0015] Reconstruct the public part information and the private part information of the user to obtain all information of the user.

[0016] As a preferred technical solution, the linear pre-coding of the public data stream and the private data stream specifically includes:

[0017] The public data stream is precoded by maximizing eigenvalue, and the private data stream is precoded by zero-forcing, to obtain corresponding precoding matrix.

[0018] As a preferred technical scheme, the corresponding ergodic and rate lower bounds of the public data stream and the private data stream are derived according to the moving speed of all users and the statistical characteristics of the channel, and specifically include:

[0019] Ergodic and rate lower bounds of the public data stream are expressed as:

[0020]

[0021] wherein, denotes the expectation of a random variable, C denotes the Shannon capacity, V denotes the channel dispersion, Q -1 denotes the inverse Gaussian Q function, Γ c，k denotes the signal-to-interference-and-noise ratio of the public data stream at the user k, β c,k denotes the block error rate of the public data stream, l c denotes the block length of the public data stream, k denotes the kth user, ε = J0(2πf D T) denotes the time correlation coefficient subject to the Jakes model, f D = v k f c / c is the maximum Doppler frequency of a given carrier frequency f c and user speed v k , and c is the speed of light, ζ k denotes the large-scale fading from the base station to the user k, P denotes the maximum transmission power, t denotes the public power coefficient, K denotes the number of users, denotes the generalized n-order exponential integral;

[0022] Ergodic and rate lower bounds of the private data stream are expressed as:

[0023]

[0024] wherein, Γ p，k denotes the signal-to-interference-and-noise ratio of the private data stream at the user k, l k denotes the block length of the private data stream, β p,k denotes the block error rate of the private data stream, is the spatially uncorrelated Rayleigh flat fading channel vector between the user k and the base station at time m, N t denotes the number of base station antennas, P denotes the precoding matrix, μ kdenotes a private power coefficient, denotes a parameter information of a gamma distribution derivation,

[0025] As a preferred technical solution, the public data stream power allocation factor, the private data stream power allocation factor and the public data stream rate allocation proportion are optimized, specifically including:

[0026] Taking the maximum communication rate as the optimization target, the minimum communication rate index is introduced, and the optimization problem is constructed as:

[0027]

[0028] Wherein, C k denotes a public data stream rate part allocated to user k, R c and R k denote the expected achievable rate of public data stream and private data stream respectively, R min denotes the quality of service QoS constraint of each vehicle, 1-t denotes the power proportion of public data stream, t·μ k denotes the power proportion of the kth private data stream, t denotes the public power coefficient, μ=[μ1,μ2,...,μ K ] denotes a private power coefficient, c=[C1,C2,...,C K ] denotes a public data stream rate allocation proportion;

[0029] First, the public power coefficient under the average condition of the private data stream is optimized, and whether the current state is started according to the value of the public power coefficient. The optimization problem is solved as:

[0030]

[0031] The sequential quadratic programming method is used to find the local optimal solution t * of the problem.

[0032] If t * <<1, the optimization problem is solved as:

[0033]

[0034] The initial point is set ζ k denotes the large-scale fading from the base station to user k;

[0035] The minimum communication rate QoS requirement is realized by maximizing the minimum ergodic rate, and the optimization problem is solved as follows:

[0036]

[0037]

[0038] where, j is reduced from K to 1, if the determination is made, the loop exits;

[0039] If t * →1, set t * =1, and solve the optimization problem as:

[0040]

[0041] Set the initial point μ k =1 / K, K represents the number of users, and the sequential quadratic programming method is used to solve the optimal solution of the optimization problem.

[0042] As a preferred technical scheme, each data stream is multiplied by a corresponding precoding matrix and a power allocation factor, and all signals are superimposed to obtain a base station transmitting signal, which is specifically represented as:

[0043]

[0044] wherein x represents the base station transmitting signal, p c represents linear precoding of the common data stream, p k represents linear precoding of the private data stream, P represents the maximum transmission power, t represents the common power coefficient, μ k represents the private power coefficient, s c represents the common data stream, and s k represents the private data stream.

[0045] As a preferred technical scheme, the signal received by the receiving end is represented as:

[0046]

[0047] wherein, represents the noise-normalized received signal at user k, is the spatial uncorrelated Rayleigh flat fading channel vector between user k and the BS at time m, N t represents the number of base station antennas, ζ k represents the large-scale fading from the base station to user k, n k ~CN(0,1) is the additive white Gaussian noise at user k.

[0048] As a preferred technical scheme, after receiving the signal, the receiving end decodes the common data stream by regarding all private data streams as noise, and the signal-to-interference-and-noise ratio of the common data stream is represented as:

[0049]

[0050] where P denotes the maximum transmit power, t denotes the common power coefficient, ζ k denotes the large-scale fading from the base station to user k, p c denotes the linear precoding of the common data stream, is the spatially uncorrelated Rayleigh flat-fading channel vector between user k and the BS at time m, N t denotes the number of base station antennas, μ j denotes the private power coefficient, p j denotes the linear precoding of the private data stream, K denotes the number of users.

[0051] As a preferred technical solution, the private data stream from other receiving ends is regarded as noise to decode the private data stream of the receiving end, and the signal-to-interference-plus-noise ratio of decoding the private data stream at user k is denoted as:

[0052]

[0053] where P denotes the maximum transmit power, t denotes the common power coefficient, ζ k denotes the large-scale fading from the base station to user k, p k denotes the linear precoding of the private data stream, is the spatially uncorrelated Rayleigh flat-fading channel vector between user k and the BS at time m, N t denotes the number of base station antennas, μ j denotes the private power coefficient, K denotes the number of users.

[0054] As a preferred technical solution, the common part information and the private part information of the user are reconstructed to obtain all information of the user, and the instantaneous maximum achievable rate of the common data stream and the private data stream is:

[0055]

[0056] where Γ c，k denotes the signal-to-interference-plus-noise ratio of the common data stream at user k, C denotes the Shannon capacity, V denotes the channel dispersion, Q -1 (·) denotes the inverse Gaussian Q function, l c denotes the block length of the common data stream, β c,k denotes the block error rate of the common data stream, Γ p，k denotes the signal-to-interference-plus-noise ratio of the private data stream at user k, l k denotes the block length of the private data stream, β p,k denotes the block error rate of the private data stream.

[0057] The application also provides a short packet rate division multiple access optimization system, comprising a base station provided with multiple antennas, a receiving end, a message division module, a data stream coding module, a linear precoding module, a power allocation optimization module, a sending signal construction module, a common data stream decoding module, a private data stream decoding module and a reconstruction module.

[0058] The base station sends a message to the receiving end.

[0059] The message division module is used for dividing the message into a common message and a private message.

[0060] The data stream coding module is used for coding the common message of all users into a common data stream and coding the private message of each user into a private data stream.

[0061] The linear precoding module is used for linearly precoding the common data stream and the private data stream to obtain corresponding precoding matrices.

[0062] The power allocation optimization module is used for deriving corresponding ergodic and rate lower bounds of the common data stream and the private data stream according to the moving speed of all users and the statistical characteristics of the channel, optimizing the common data stream power allocation factor, the private data stream power allocation factor and the common data stream rate allocation proportion to obtain the corresponding power allocation factor of each data stream and the rate allocation condition of the common data stream.

[0063] The sending signal construction module is used for multiplying each data stream by the corresponding precoding matrix and power allocation factor, and obtaining the base station transmitting signal after superimposing all signals.

[0064] The common data stream decoding module is used for decoding the common data stream by regarding all private data streams as noise and separating the common part information of the receiving end itself.

[0065] The private data stream decoding module is used for subtracting the corresponding part of the common data stream from the receiving signal, decoding the private data stream of the receiving end by regarding the private data stream from other receiving ends as noise according to the signal after subtraction, and obtaining the private part information.

[0066] The reconstruction module is used for reconstructing all information of the user by using the common part information and the private part information of the user.

[0067] Compared with the prior art, the application has the following advantages and beneficial effects:

[0068] (1) The application adopts a short packet based rate division multiple access technical solution, solves the technical problems of ultra-low latency and ultra-low block error rate in communication, and achieves the technical effect of significantly improving the spectrum utilization rate under non-ideal CSI.

[0069] (2) The present invention can implement the short packet rate segmentation scheme with low complexity. For complex mobility environments, the present invention improves the parameter settings such as power allocation and rate allocation of public and private data streams. Compared with the average setting of RSMA, traditional NOMA, SDMA and other multiple access technologies, the present invention can meet the strict quality of service (QoS) requirements in ultra-reliable low latency communication (URLLC) while effectively improving system performance and rate. Attached Figure Description

[0070] Figure 1 This is a flowchart illustrating the short packet rate segmentation multiple access optimization method of the present invention;

[0071] Figure 2 This is a schematic diagram of the overall architecture of the present invention applied to an in-vehicle scenario;

[0072] Figure 3 A schematic diagram showing the system and rate comparison of the RSMA of the present invention with RSMA, conventional SDMA, and NOMA under different transmit power conditions, and at average settings;

[0073] Figure 4 A schematic diagram showing the system and rate comparison of the RSMA of the present invention with the RSMA, conventional SDMA, and NOMA under average settings at different vehicle speeds;

[0074] Figure 5 This diagram illustrates a comparison of the system and rate of RSMA of the present invention with RSMA, conventional SDMA, and NOMA under average settings, under different block lengths and block error rates. Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0076] Example 1

[0077] like Figure 1 As shown, this embodiment provides a short packet rate segmentation multiple access optimization method, which can be used to handle problems such as vehicle location changes, channel fading, and traffic allocation in high mobility systems. Figure 2 As shown, a system with K users will be further explained. A single-layer rate division multiple access (RDMA) scheme is used to serve K users. The downlink MISO broadcast system includes a system equipped with N... t A single base station (BS) with one antenna and K single-antenna high-mobility users, the user set is as follows: K represents the number of users, and the method specifically comprises the following steps:

[0078] S1: the base station sends K messages W to K users k , the message W k is divided into two parts, namely a public message and a private message;

[0079] Specifically, the message W k is divided into two parts {W c,k , W p,k}, which are respectively named as a public message and a private message, the public messages {W c,1 ,..., W c,K} of the K users are encoded into a common data stream s c , which is decoded by all K users, and the private message W p,k of the user k is encoded into a private data stream s k , and the entire data stream vector to be sent is represented by s = [s c , s1, s2,..., s K ] T ;

[0080] In this embodiment, linear precoding is applied to all data streams, and the precoding matrix satisfies ||p c || 2 = 1, ||p k || 2 = 1, Zero-forcing precoding (ZF-precoding) is adopted for the private data stream, and eigenvalue maximization precoding is adopted for the common data stream, p c represents linear precoding of the common data stream, and p k represents linear precoding of the private data stream of the user k.

[0081] S2: the base station solves the optimization problems P1-P4 by deducing the corresponding traversal and lower bound of the rate of the public data stream and the private data stream according to the channel information of the high-mobility user, determines the power allocation factor, the beamforming scheme of each data stream, and the rate allocation proportion of the common data stream;

[0082] In this embodiment, the public part of the user is combined into a common data stream, and the encoding proportion is set according to the calculated rate allocation proportion of the common data stream, and the private part of each user is encoded into a private data stream;

[0083] In this embodiment, the base station divides the information of the high-mobility user into a public part and a private part and performs short packet processing;

[0084] In the embodiment, the base station derives the lower bound of the ergodic sum rate and the lower bound of the sum rate of the system according to the moving speed of the K users and the statistical characteristics of the channel, and optimizes the power allocation factor of the common data stream, the power allocation factor of the private data stream, and the rate allocation proportion of the common data stream in turn according to the derived lower bound of the ergodic sum rate and the lower bound of the sum rate, to obtain the power allocation factor corresponding to each data stream and the rate allocation of the common data stream.

[0085] In step S2, the specific operation of optimization according to the derived lower bound of the ergodic sum rate and the lower bound of the sum rate is as follows:

[0086] First, the private data stream adopts average power allocation, and the common power allocation factor is solved by using the sequential quadratic programming (SQP) method according to the lower bound of the ergodic sum rate and the lower bound of the sum rate.

[0087] According to the solved common power allocation factor, the scheme automatically determines whether to start the common data stream.

[0088] If the common data stream is started, the power allocation factor of the private data stream is first optimized according to the lower bound of the ergodic sum rate and the lower bound of the sum rate of the private data stream, and then the rate allocation proportion of the common data stream is optimized according to the given power setting of the common and private data stream scheme, to complete the configuration of all parameters.

[0089] If the common data stream is closed, the related parameters of the common data stream do not need to be optimized, and at this time, the power allocation factor of the private data stream of the scheme is obtained by optimizing the closed expression of the lower bound of the private data stream, to complete the configuration of all parameters.

[0090] In the embodiment, the steps of optimizing the power of the common data stream, the power of the private data stream, and the rate allocation of the common data stream will be specifically explained in combination with the lower bound of the ergodic sum rate and the lower bound of the sum rate:

[0091] The final lower bound of the ergodic sum rate and the sum rate of the common data stream can be expressed as:

[0092]

[0093] wherein the function Q -1 (·) represents the inverse Gaussian Q function, E(X) represents the expectation of the random variable, and the generalized n-order exponential integral is represented by

[0094] The lower bound of the ergodic sum rate and the sum rate of the private data stream can be finally expressed as:

[0095]

[0096] wherein the second row, the parameter

[0097] And then the fourth line of expression, denotes the derivative of the parameter information of the gamma distribution ,

[0098] Further, in order to cope with the actual vehicle communication scene, the total maximum communication rate of all vehicles is taken as the optimization target, and the minimum communication rate index of each vehicle is introduced in order to meet the communication demand of each vehicle.

[0099] The optimization problem is constructed as:

[0100]

[0101] Where C k denotes the public data flow rate part allocated to user k, R c and R k denote the expected reachable rate of public data flow and private data flow respectively, R min denotes the QoS (Quality of Service) constraint of each vehicle, K represents the number of users, and the optimization variable c = [C1, C2,..., C K ] determines the rate allocation among users in the public data flow. Variables t and μ = [μ1, μ2,..., μ K ] represent parameters related to power allocation ratio, where 1-t represents the power ratio of public data flow, and t·μ k represents the power ratio of the kth private flow. The public power coefficient t, the private power coefficient μ = [μ1, μ2,..., μ K ], and the public data flow rate allocation ratio c = [C1, C2,..., C K ] need to be optimized to maximize the objective function. The mutual dependence of variables and their ratios in the objective function makes the problem essentially non-convex, and the objective function lacks a direct closed-form expression, which can only be approximated by the sample average result of Monte Carlo, and is not conducive to solving.

[0102] Next, using the derived closed-form solution, some simple methods for solving the problem are provided. First, optimize the public power coefficient t under the average condition of private data flow, and according to the value of the public coefficient, judge whether to start the public data flow in the current state. For this purpose, the following optimization problem is solved alternatively:

[0103]

[0104] Although there is only one variable, it is difficult to obtain a direct closed-form solution by directly deriving the objective function, which is not an elementary function. Considering the continuous first derivative of the objective function, we use the "fmincon" function in MATLAB to solve it, fixing The starting point t = 0.5 is solved for the premise of the average allocation of private data streams. The function configuration is based on gradient information, and the sequential quadratic programming (SQP) method can efficiently find the local optimal solution t * .

[0105] According to the two results of optimizing the public coefficient, the following two schemes are respectively as follows:

[0106] 1) t * <<1, in this case, most of the power will be allocated to the public data stream. Therefore, it can be assumed that the public data stream has a large enough rate to be allocated to meet the QoS requirements of the vehicle. Further, the following optimization problem can be solved alternatively:

[0107]

[0108] Since the objective function is a non-elementary function and the variables are coupled with each other, the sequential quadratic programming algorithm of MATLAB is used to obtain a solution. The initial point is set This initial setting can achieve faster convergence compared to optimization starting from the average.

[0109] After completing the allocation of public and private coefficients, the actual Shannon achievable rate of all data streams is determined, and by maximizing the minimum traversal rate, the minimum communication rate QoS requirement can be indirectly achieved, that is, the following is solved:

[0110]

[0111] This problem is essentially a water filling problem, which can be calculated by the following expression:

[0112]

[0113] where, j is reduced from K to 1, if Loop exit, R c , R k can be obtained by the sample average method.

[0114] 2) t * →1, in the actual system, it means that the public data stream is turned off, and the RSMA scheme is degraded to the SDMA scheme. Therefore, t * =1 can be directly set, and the rate allocation of the public data stream is no longer considered, and the following problem is directly optimized:

[0115]

[0116] Setting initial point μ k = 1 / K, Continue to use the SQP method to solve the problem.

[0117] S3: After multiplying the entire data stream vector by the precoding matrix P and the power allocation factor, all signals are superimposed and sent to the above K users by the base station;

[0118] Specifically, the base station transmit signal can be represented as: The power sharing coefficient 0≤t≤1 controls the power allocation between the common data stream and the total private data stream, where 1-t is the power ratio of the common data stream. The parameter 0≤μ k ≤1, and The private data stream power allocation ratio allocated to the kth vehicle is specified, and the vector μ=[μ1, μ2,..., μ K ] is defined, so t·μ k is the power ratio of the kth private data stream. Assuming that all symbols have unit power, The transmit signal is subject to the maximum transmit power P constraint.

[0119] In this embodiment, the noise-normalized received signal at user k can be represented as: where is the spatial uncorrelated Rayleigh flat fading channel vector between user k and the BS at time m, each element of which is independent and subject to CN(0, 1) distribution, ζ k represents the large-scale fading from the base station to user k, n k ~ CN(0, 1) is the additive white Gaussian noise at user k. Considering the scenario of non-ideal CSIT due to user mobility and delay in user reporting CSI, the channel vector at time m is represented as e k [m] each element is independent and subject to CN(0, 1) distribution. ε=J0(2πf D T) represents the time correlation coefficient subject to Jakes model, f D = v k f c / c is the maximum Doppler frequency given the carrier frequency f c and user speed v k (c is the speed of light), and T represents the time interval of channel instantiation. In other words, h k[m] is the instantaneous channel vector observed at time m. Due to the high mobility of users, the transmitter only knows the channel vector h k [m-1] observed at time m-1 and calculates the precoding matrix with h k [m-1] at time m.

[0120] S4: After receiving the signal, the user first decodes the common data stream s

[0121] Specifically, after receiving the signal, the user first decodes the common data stream s c and separates the common part information of the user k. The signal-to-interference-and-noise ratio (SINR) of the common data stream at the user k is:

[0122] S5: After successfully decoding the common data stream s c , the user k then subtracts the part corresponding to the common data stream from the received signal using successive interference cancellation (SIC) technique, and then decodes the private data stream s k , decodes the private data stream of the user k from the private data streams of other users as noise according to the subtracted signal, and further obtains the private part information.

[0123] In this embodiment, the SINR of decoding the private data stream s k at the user k is:

[0124] S6: The common part and private part information of the user obtained in steps S4 and S5 are reconstructed, and finally all the information of the user is obtained;

[0125] In this embodiment, after successfully decoding the common data stream and the private data stream, the user k reconstructs the original message by extracting W c from the decoded W c,k and combining W c,k and W p,k .

[0126] In short packet communication, for a large enough block length (l>100), at time m, the instantaneous maximum achievable rate (the maximal achievable rate) of the common data stream and the private data stream can be approximated as:

[0127]

[0128]

[0129] To ensure that the public data stream is successfully decoded by all users, the maximum achievable rate of the public data stream is calculated as the signal-to-interference ratio. Wherein the Shannon capacity C(x)=log2(1+x), the channel dispersion V(x)=(1-(1+x) -2 )(log2e) 2 , l c ,l1,...,l K represent the block lengths of streams s c ,s1,s2,...,s K , respectively, and beta represents the block error rate (BLER);

[0130] In this embodiment, the channel state information is obtained by a pilot-based channel estimation method.

[0131] The present application considers the influence of user mobility on the accuracy of channel state information (CSIT) and derives the ergodic sum rate lower bound closed-form expression related to user speed. By jointly optimizing the expression, the power allocation and rate scheduling of public and private data streams are determined, effectively improving the sum rate of the system. In addition, the present application considers the influence of limited packet length in actual communication, further improving the system reliability and reducing the delay. Through accurate power and rate optimization, the present application realizes higher transmission efficiency and lower delay in high-speed mobile environment.

[0132] Through the detailed description of the foregoing embodiments, it can be clearly seen that the RSMA scheme proposed by the present application improves the system performance by optimizing the public data stream power allocation coefficient, the private data stream power allocation coefficient and the public data stream rate allocation proportion. As shown in Figures 3-5 , the trend of the ergodic sum rate of the system under different parameters with transmission power, vehicle speed and block length is shown. In all test scenarios, the proposed RSMA scheme is always superior to the traditional scheme (including RSMA, SDMA and NOMA under average setting). Especially in high-speed mobile environment, through accurate optimization and power allocation of private data stream, the robustness and transmission efficiency of the system are further improved.

[0133] Specifically, Figure 3 shows that as the transmission power changes, the proposed RSMA scheme shows higher sum rate than the traditional scheme. Figure 4 Emphasizes the influence of vehicle speed on system performance. Although the ergodic sum rate decreases with the increase of speed, the RSMA scheme of the present application can still maintain strong robustness and significantly exceed the traditional scheme under average setting. Figure 5The performance advantage of the short packet rate division multiple access optimization system is further verified from the block length and the block error rate, and the optimal performance gain is ensured under the strict block error rate and short code transmission conditions.

[0134] Embodiment 2

[0135] The embodiment provides a short packet rate division multiple access optimization system, which is used for implementing the short packet rate division multiple access optimization method in the embodiment 1, and the system comprises a base station provided with a plurality of antennas, a receiving end, a message division module, a data stream coding module, a linear precoding module, a power allocation optimization module, a sending signal construction module, a common data stream decoding module, a private data stream decoding module and a reconstruction module.

[0136] In the embodiment, the base station sends a message to the receiving end.

[0137] In the embodiment, the message division module is used for dividing the message into a common message and a private message.

[0138] In the embodiment, the data stream coding module is used for coding the common message of all users into a common data stream and coding the private message of each user into a private data stream.

[0139] In the embodiment, the linear precoding module is used for linearly precoding the common data stream and the private data stream to obtain corresponding precoding matrices.

[0140] In the embodiment, the power allocation optimization module is used for deriving corresponding ergodic and rate lower bounds of the common data stream and the private data stream according to the moving speeds of all users and the statistical characteristics of channels, optimizing a common data stream power allocation factor, a private data stream power allocation factor and a common data stream rate allocation proportion, and obtaining a corresponding power allocation factor of each data stream and a rate allocation condition of the common data stream.

[0141] In the embodiment, the sending signal construction module is used for multiplying each data stream by a corresponding precoding matrix and a power allocation factor, and obtaining a base station transmitting signal after superimposing all signals.

[0142] In the embodiment, the common data stream decoding module is used for decoding the common data stream by regarding all private data streams as noise and separating the common part information of the receiving end itself.

[0143] In the embodiment, the private data stream decoding module is used for subtracting the part corresponding to the common data stream from the receiving signal, decoding the private data stream of the receiving end by regarding the private data streams from other receiving ends as noise according to the subtracted signal, and obtaining the private part information.

[0144] In the embodiment, the reconstruction module is configured to reconstruct all information of the user by using the public part information and the private part information of the user.

[0145] The above embodiment is the preferred embodiment of the present application, but the embodiment of the present application is not limited to the above embodiment, and any change, modification, substitution, combination, simplification, which does not deviate from the spirit and principle of the present application, should be an equivalent replacement, and is included in the protection scope of the present application.

Claims

1. A method for optimizing short packet rate division multiple access, characterized by, The method comprises the following steps: obtaining a message sent by a base station to a user, and splitting the message into a common message and a private message; encoding the common message of all users into a common data stream and encoding the private message of each user into a private data stream; linearly precoding the common data stream and the private data stream to obtain corresponding precoding matrices; deriving corresponding ergodic and rate lower bounds of the common data stream and the private data stream according to the moving speed of all users and the statistical characteristics of a channel, optimizing a common data stream power allocation factor, a private data stream power allocation factor and a common data stream rate allocation proportion, and obtaining a corresponding power allocation factor of each data stream and a rate allocation condition of the common data stream; optimizing the common data stream power allocation factor, the private data stream power allocation factor and the common data stream rate allocation proportion, and specifically comprising: taking the maximum communication rate as an optimization target, introducing an index of the minimum communication rate, and constructing an optimization problem as follows: s.t. 0≤t≤1, where C k denotes the common data stream rate portion allocated to user k, R c and R k denote the expected achievable rates of the common data stream and the private data streams, respectively, R min denotes the quality of service (QoS) constraint of each vehicle, 1-t denotes the power proportion of the common data stream, t·μ k denotes the power proportion of the kth private data stream, t denotes the common power coefficient, μ=[μ1,μ2,...,μ K ] denotes the private power coefficients, and c=[C1,C2,...,C K ] denotes the common data stream rate allocation proportion. multiplying each data stream by the corresponding precoding matrix and the power allocation factor, and obtaining a base station transmitting signal after superimposing all signals; after receiving the signal, regarding all private data streams as noise to decode the common data stream, and separating the common part information of the receiving end itself; subtracting the part corresponding to the common data stream from the received signal, regarding the private data stream from other receiving ends as noise to decode the private data stream of the receiving end, and obtaining the private part information; reconstructing the common part information and the private part information of the user to obtain all information of the user.

2. The method of claim 1, wherein, The linear precoding of the common data stream and the private data stream specifically comprises: maximizing eigenvalue precoding for the common data stream and zero-forcing precoding for the private data stream to obtain corresponding precoding matrices.

3. The method of claim 1, wherein, Deriving corresponding ergodic and rate lower bounds of the common data stream and the private data stream according to the moving speed of all users and the statistical characteristics of a channel, specifically comprising: Traversal and rate lower bound of public data streams is represented as: wherein, denotes the expectation of a random variable, C denotes the Shannon capacity, V denotes the channel dispersion, Q -1 denotes the inverse Gaussian Q-function, Γ c，k denotes the signal-to-interference-plus-noise ratio of the common data stream at user k, β c,k denotes the block error rate of the common data stream, l c denotes the block length of the common data stream, k denotes the k-th user, ε = J0(2πf D T) denotes the time correlation coefficient following the Jakes model, f D = v k f c / c is the maximum Doppler frequency for a given carrier frequency f c and user speed v k , c is the speed of light, ζ k denotes the large-scale fading from the base station to user k, P denotes the maximum transmit power, t denotes the common power coefficient, K denotes the number of users, denotes the generalized n-th order exponential integral; Traversal and rate lower bound of private data streams is represented as: where Γ p，k denotes the signal-to-interference-and-noise ratio of the private data stream at user k, l k denotes the block length of the private data stream, β p,k denotes the block error rate of the private data stream, is the spatial uncorrelated Rayleigh flat-fading channel vector between user k and the base station at time instant m, N t denotes the number of base station antennas, P denotes the precoding matrix, μ k denotes the private power coefficient, denotes the derivation of the parameter information of the gamma distribution ​ 4. The method of claim 1, wherein, first optimizing the common power coefficient under the average condition of the private data stream, judging whether the current state starts the common data stream according to the value of the common power coefficient, and solving the optimization problem as follows: s.t. 0≤t≤1 A sequential quadratic programming method is used to find a local optimum solution t * ; If t * <<1, solve the optimization problem for: Initial point setup ζ k denotes the large scale fading from the base station to user k; maximizing the minimum ergodic and rate to realize the requirement of the minimum communication rate QoS, and solving the optimization problem as follows: wherein j is decreased from K to 1, if the decision then the loop exits; If t * →1, set t * =1, solve the optimization problem as: Setting initial point μ k = 1 / K, K represents the number of users, and the optimal solution of the optimization problem is solved by using a sequential quadratic programming method.

5. The method of claim 1, wherein, multiplying each data stream by the corresponding precoding matrix and the power allocation factor, and obtaining a base station transmitting signal after superimposing all signals, which is specifically represented as: where x denotes the base station transmit signal, p c denotes linear precoding of the common data stream, p k denotes linear precoding of the private data stream, P denotes the maximum transmit power, t k denotes the private power coefficient, s c denotes the common data stream, s k denotes the private data stream.

6. The method of claim 5, wherein, the signal received by the receiving end is represented as: wherein, denotes the noise normalized received signal at user k, is the spatial uncorrelated Rayleigh flat fading channel vector between user k and the BS at time instant m, N t denotes the number of base station antennas, ζ k denotes the large scale fading from the base station to user k, n k ~ CN(0, 1) is the additive white Gaussian noise at user k.

7. The method of claim 1, wherein, after receiving the signal, regarding all private data streams as noise to decode the common data stream, and the signal-to-interference-and-noise ratio of the common data stream is represented as: where P denotes the maximum transmit power, t denotes the common power coefficient, ζ k denotes the large-scale fading from the base station to user k, p c denotes the linear precoding of the common data stream, is the spatial uncorrelated Rayleigh flat fading channel vector between user k and the BS at time instant m, N t denotes the number of base station antennas, μ j denotes the private power coefficient, p j denotes the linear precoding of the private data stream, K denotes the number of users.

8. The method of claim 1, wherein, regarding the private data stream from other receiving ends as noise to decode the private data stream of the receiving end, and the signal-to-interference-and-noise ratio of the decoded private data stream at the user k is represented as: where P denotes the maximum transmit power, t denotes the common power coefficient, ζ k denotes the large-scale fading from the base station to user k, p k denotes the linear precoding of the private data stream, is the spatial uncorrelated Rayleigh flat-fading channel vector between user k and the BS at time instant m, N t denotes the number of base station antennas, μ j denotes the private power coefficient, K denotes the number of users.

9. The method of claim 1, wherein, Reconstructing the public part information and the private part information of the user to obtain all information of the user, and the instantaneous maximum reachable rate of the public data stream and the private data stream is: where Γ c，k denotes the signal-to-interference-plus-noise ratio of the common data stream at user k, C denotes the Shannon capacity, V denotes the channel dispersion, Q -1 (·) denotes the inverse Gaussian Q-function, l c denotes the block length of the common data stream, β c,k denotes the block error rate of the common data stream, Γ p，k denotes the signal-to-interference-plus-noise ratio of the private data stream at user k, l k denotes the block length of the private data stream, β p,k denotes the block error rate of the private data stream.

10. A short packet rate division multiple access optimization system, comprising: comprising: a base station with multiple antennas, a receiving end, a message splitting module, a data stream encoding module, a linear precoding module, a power allocation optimization module, a transmitting signal construction module, a common data stream decoding module, a private data stream decoding module and a reconstruction module; the base station sends a message to the receiving end; the message splitting module is used to split the message into a common message and a private message; The data stream encoding module is configured to encode common messages of all users into a common data stream and private messages of each user into a private data stream; The linear precoding module is configured to linearly precode the common data stream and the private data stream to obtain corresponding precoding matrices; The power allocation optimization module is configured to derive corresponding ergodic and rate lower bounds of the common data stream and the private data stream according to mobile speeds of all users and statistical characteristics of channels, and optimize a common data stream power allocation factor, a private data stream power allocation factor, and a common data stream rate allocation proportion to obtain a corresponding power allocation factor of each data stream and a rate allocation condition of the common data stream; The optimization of the common data stream power allocation factor, the private data stream power allocation factor, and the common data stream rate allocation proportion specifically includes: taking a maximum communication rate as an optimization target, introducing an index of a minimum communication rate, and constructing an optimization problem as: s.t. 0≤t≤1, where C k denotes the common data stream rate portion allocated to user k, R c and R k denote the expected achievable rates of the common data stream and the private data streams, respectively, R min denotes the quality of service (QoS) constraint of each vehicle, 1-t denotes the power proportion of the common data stream, t·μ k denotes the power proportion of the kth private data stream, t denotes the common power coefficient, μ=[μ1,μ2,...,μ K ] denotes the private power coefficients, and c=[C1,C2,...,C K ] denotes the common data stream rate allocation proportion. The sending signal construction module is configured to multiply each data stream by a corresponding precoding matrix and a power allocation factor, and obtain a base station transmitting signal by superimposing all signals; The common data stream decoding module is configured to decode the common data stream by regarding all private data streams as noise and separate common part information of the receiving end itself; The private data stream decoding module is configured to subtract a part corresponding to the common data stream from a receiving signal, decode private data streams from other receiving ends by regarding the private data streams as noise according to the subtracted signal, and obtain private part information; The reconstruction module is configured to reconstruct all information of a user by using common part information and private part information of the user.

Images