Power allocation and beamforming method for downlink communication system of Internet of Vehicles based on NOMA-OTFS

Abstract

The present invention relates to a NOMA-OTFS-based vehicle networking downlink communication system power allocation and beamforming method; first, high-speed user signals are restored in the delay-Doppler domain by frequency domain linear equalization, and serial interference elimination is used to overcome the influence of high-speed users on low-speed users, so as to accurately restore low-speed user signals; secondly, by analyzing user channel state information, the signal-to-interference-noise ratio of each user at each stage of SIC is calculated, and a joint optimization problem is constructed based on the result to maximize the total system throughput, and the power allocation vector and the beamforming vector of each user are jointly optimized; finally, continuous convex approximation and semi-positive definite programming algorithms are used to solve, and the optimal or suboptimal solution of the power allocation vector and the beamforming vector of each user is output. Compared with the classic NOMA-OTFS vehicle networking communication system, the present invention has higher resource utilization efficiency and communication quality, and can effectively improve system throughput and reliability.

Description

Technical Field

[0001] The present invention relates to a NOMA-OTFS-based vehicle networking downlink communication system power allocation and beamforming method, and belongs to the technical field of wireless communication. Background Art

[0002] With the rapid development of Internet of Vehicles (IoV) technology, it is increasingly important to achieve efficient and reliable inter-vehicle communication. In this context, multi-user multiple input single output (MU-MISO) technology is widely used because it can serve multiple users simultaneously on multiple transmission channels. At the same time, the combination of non-orthogonal multiple access (NOMA) and orthogonal time-frequency space modulation (OTFS) technology provides a new way to improve the comprehensive performance of wireless communication in high-speed mobile environments. NOMA technology improves the system user capacity and spectrum efficiency by completing user multiple access in the power domain, while OTFS technology can effectively combat the adverse effects in high-speed mobile environments, such as frequency selective fading and Doppler effect, by utilizing the structural characteristics of the delay-Doppler domain and performing signal modulation in this domain. It is particularly suitable for dynamic IoV environments. At the same time, reasonable power allocation and beamforming design in the NOMA-OTFS system are crucial to improving system throughput. Accurate power allocation can ensure that NOMA users access the network at different power levels and maximize the system spectrum efficiency and capacity. Beamforming can control the signal directionality by adjusting the weight vector of the base station transmitting antenna, enhance the signal transmission quality, and reduce interference to other users. As a widely used mathematical method for solving resource allocation problems, convex optimization algorithms are very suitable for solving the optimal value of complex optimization problems. By constructing an optimization model, the NOMA-OTFS system beamforming and power allocation vectors can be used as optimization variables to maximize the system throughput and thus solve the problem. This method can ensure that the global optimal solution is found, and the calculation process is efficient and easy to implement.

[0003] Ding Zhiguo et al. ([1]Z. Ding, "Robust Beamforming Design for OTFS-NOMA," in IEEE Open J. Commun. Soc., vol. 1, pp. 33-40, 2020.) proposed a new NOMA method for users with different mobility. The method combines OFDM and OTFS technologies to provide services for low-speed users in the TF domain and high-speed users in the DD domain, and uses frequency domain linear equalization technology to eliminate inter-symbol interference for high-speed user signals at the receiving end. At the same time, the OTFS-NOMA system beam design is considered when there are errors in the channel state information. It is proved that the proposed beamforming scheme can significantly improve the fairness of system users and the minimum rate of each NOMA user. Although the scheme has been optimized in terms of system beamforming, it adopts the strategy of all users sharing the same beamforming vector, without considering the geographical differences of each user, which may lead to inaccurate beam coverage, thereby affecting the signal directional performance and communication efficiency. At the same time, the scheme does not fully consider the impact of power allocation on the overall performance of the system, which is also an important factor that cannot be ignored in the vehicle network communication system.

[0004] With the rapid development of Internet of Vehicles technology and the increasing demand for data, the problem of how to effectively improve the communication efficiency and coverage quality of the Internet of Vehicles system needs to be solved urgently. Therefore, it is particularly important to conduct in-depth research on the joint design of power allocation and beamforming based on the NOMA-OTFS system in Internet of Vehicles communication. Summary of the invention

[0005] In order to solve the data transmission reliability and system capacity problems in IoV communication, the present invention proposes a NOMA-OTFS-based IoV downlink communication system power allocation and beamforming method. When the transmission signal-to-noise ratio is 30dB, the system throughput of this method is improved by about 5bps / Hz compared with the solution in the prior art.

[0006] Terminology explanation:

[0007] 1. NOMA: Non-orthogonal multiple access (NOMA) is a multiple access technology that allows multiple users to communicate on the same time-frequency resources. It distinguishes users in the power domain (using different power levels) so that multiple users can share the same channel. NOMA improves spectrum utilization and can provide fairer quality of service among users by using super-position coding and layer-by-layer decoding technology.

[0008] 2. OTFS: Orthogonal Time-Frequency-Space (OTFS) is a modulation technique designed to take full advantage of the orthogonality of the time and frequency domains, and is particularly suitable for dealing with channel effects in high-speed mobile environments, such as Doppler shift and time-varying fading. OTFS can effectively resist these adverse effects and provide stable and reliable signal transmission by mapping data in the so-called delay-Doppler domain.

[0009] 3. Beamforming: Beamforming is a technology that uses multiple antennas (arrays) to send or receive signals. By adjusting the signal phase and amplitude of each antenna, the signal can be intentionally enhanced in a specific direction. This technology can improve the quality of signal reception, reduce interference, and increase the directionality and distance of communication.

[0010] 4. FD-LE: Frequency Domain Linear Equalization (FD-LE) is a commonly used equalization technology in communication systems, especially suitable for high-speed data transmission and multipath propagation environments. It converts the received signal into the frequency domain through fast Fourier transform, adjusts the signal using preset equalization coefficients, compensates for channel distortion and interference, and restores the processed signal to the time domain output through inverse transformation.

[0011] 5. SIC: Serial Interference Cancellation (SIC) is an efficient signal processing technique used to reduce multi-user interference at the receiving end. SIC first identifies and demodulates the strongest user signal in the signal, then eliminates the influence of the demodulated signal from the total signal, and gradually repeats this process to recover the signals of other users. This method effectively improves the reception quality of the signal and the error rate performance of the system by eliminating interference sources one by one.

[0012] 6. CSI: Channel state information (CSI) refers to the knowledge about the channel transmission characteristics in the wireless communication system, including channel gain, phase offset, delay spread and other parameters. By measuring these parameters, the system can understand the various influences on the signal during the process from the transmitter to the receiver, so that the transmitter can adjust its transmission strategy more accurately.

[0013] 7. SINR: Signal to Interference and Noise Ratio (SINR) is a key performance indicator in wireless communications, used to describe the ratio of the expected signal strength to the sum of interference and noise at a specific time and frequency. SINR is an important indicator for evaluating wireless signal quality and communication reliability, which affects data transmission rate, error rate and system capacity.

[0014] The technical solution of the present invention is as follows:

[0015] A power allocation and beamforming method for a downlink communication system of an Internet of Vehicles based on NOMA-OTFS, wherein the downlink communication system of the Internet of Vehicles includes a base station, a distributed vehicle unit and a signal processing module; the base station is equipped with a multi-element transmitting antenna array; the distributed vehicle unit includes a high-speed vehicle user using DD domain modulation, i.e., a high-speed user, and a plurality of low-speed vehicle users using time-frequency domain modulation, i.e., a low-speed user; the signal processing module includes a serial interference canceller and a frequency domain linear equalizer;

[0016] Assume that the received noise is additive white Gaussian noise; including:

[0017] First, the high-speed user signal in the DD domain is restored at the high-speed user receiving end, and then the low-speed user signal is restored at each low-speed user receiving end;

[0018] Secondly, by analyzing the CSI corresponding to each user, the SINR of each user is calculated, and based on the SINR of each user, a joint optimization problem is constructed to maximize the total throughput of the downlink communication system of the Internet of Vehicles, and then the power allocation vector and the beamforming vector of each user are jointly optimized;

[0019] Finally, the optimal beamforming and power allocation vectors are output through continuous convex approximation and semi-positive definite programming algorithms.

[0020] Preferably, according to the present invention, the construction process of the vehicle networking downlink communication system includes:

[0021] Assume that there are K+1 users U in the downlink communication system of the Internet of Vehicles i , i = 0,…,K, where the high-speed user is denoted as U0 and the low-speed user is denoted as U j , j=1,…,K; Assume that each time-frequency resource block in the downlink communication system of the Internet of Vehicles includes N OFDM symbols and K+1≤M, where M is the number of system subcarriers; each low-speed user independently occupies subcarrier resources in the TF domain, and then the low-speed user information symbols and the high-speed user information symbols are superimposed through the NOMA technology, and at the transmitting end, the signal is customized for each user through a precise beamforming strategy to optimize the overall signal transmission quality and system performance.

[0022] Further preferably, in the Internet of Vehicles downlink communication system;

[0023] The signal sent by the TF domain of the vth antenna of the base station is: where γ i and v,i ,i=0,…,K, respectively, are the power allocation coefficient of user i in the system and its beamforming weight on the vth path, is the signal sent by user i in the TF domain;

[0024] The input and output relationship of the high-speed user DD domain is: Among them, y0 represents the received signal vector of the high-speed user, x i ,i=0,…,K, represents the signal vector sent by the base station to user i in the system, that is, when i=0, x i represents the signal vector sent by the base station to the high-speed user, when i=j and j=1,…,K i represents the signal vector sent to the jth low-speed user, V is the number of transmitting antennas, represents the channel matrix from the vth antenna of the base station to the high-speed user, z0 is the additive white Gaussian noise vector at the high-speed user;

[0025] The received signal vector of the jth low-speed user is in represents the channel matrix from the vth antenna of the base station to the jth low-speed user, z j is the additive white Gaussian noise vector at the low-speed user j;

[0026] The high-speed user channel equalization matrix is ​​set as Among them, F N and F M is the Fourier transform matrix of N and M points, (·) H represents the complex conjugate transpose operation, is the equivalent channel matrix from the vth antenna of the base station to the high-speed user, is a diagonal matrix, and The (kM+l+1)th diagonal element of in, for The (nM+m+1)th row, first column element;

[0027] Equalize y0 to restore the high-speed user signal in the DD domain, and the equalized signal is:

[0028] Among them, the first term on the right side of the equation is the signal sent by the base station to the high-speed user, the second term is the interference term of the low-speed user, and the third term is the additive Gaussian white noise; assuming that the signal amplitude transmitted by all users is the same and the noise power is normalized, the signal-to-noise ratio sent by each user is expressed as By calculating the expected signal, interference and noise power, the SINR at the high-speed user is derived as Among them, g k,l is the base station to high-speed user channel matrix, and

[0029] Preferably, according to the present invention, the low-speed user signal is restored at each low-speed user receiving end; comprising:

[0030] Perform DD domain detection on the equalized high-speed user signal, perform SIC at each low-speed user to eliminate the interference of the high-speed user, and then restore the signal of each low-speed user; specifically:

[0031] Assume that the channel of the low-speed user is a time-invariant channel, that is, the low-speed user does not experience Doppler frequency shift. Therefore, the channel matrix from the vth antenna to user j is is a block diagonal matrix, plus the detection matrix After that, the low-speed user signal is effectively balanced and restored, among which, is the equivalent channel matrix from the base station's vth antenna to low-speed user j, is a diagonal matrix, and its (l+1)th diagonal element is in, for The (m+1)th row, first column element of .

[0032] Preferably, according to the present invention, calculating the SINR of each user comprises:

[0033] Assuming that the high-speed user signal is perfectly demodulated and eliminated, the high-speed user U0 is obtained after the jth low-speed user U j The SINR under interference is: Among them, h j,l is the channel matrix from the base station to the jth low-speed user, so the TF domain low-speed user U after SIC is obtained j The received signal at in, is the time-frequency channel response from the vth antenna of the base station to the jth low-speed user, is the time-frequency domain transmitted signal of user j; assuming that all NOMA user channels are time-invariant channels, the SINR corresponding to the jth NOMA user is

[0034] Preferably, according to the present invention, under the conditions of satisfying the QoS of high-speed users, perfect execution of SIC by low-speed users and total power constraints, the NOMA-OTFS throughput is maximized to achieve efficient resource utilization and optimization of user experience. The established joint optimization problem of beamforming and power allocation is as follows:

[0035]

[0036]

[0037]

[0038]

[0039] Where P is the total transmission power of the base station; constraint c 1ais the high-speed user QoS and perfect SIC constraints, R0 is the minimum data rate required to meet the high-speed user QoS; constraint c 1b represents the base station beamforming power constraint; constraint c 1c is the total system power constraint.

[0040] Preferably, according to the present invention, the solution is performed by continuous convex approximation and semi-positive definite programming algorithm, and the optimal beamforming and power allocation vector are output; the power allocation-beamforming joint optimization algorithm specifically includes:

[0041] Assume that the number of NOMA users is 2 and the beamforming vector w0,w is fixed. j ,j=1,2, solve the power vector γ as follows:

[0042] First, the power allocation optimization sub-problem is transformed into:

[0043]

[0044] Convert the subproblem into a concave difference function, then:

[0045]

[0046] Using continuous convex approximation technology to further transform the function Transformed into a convex function, transformed into and

[0047] Secondly, by introducing auxiliary variables The original objective function is transformed into And further by fixing the variables Transform the objective function into Continue to introduce auxiliary variables Transform the original objective function into The objective function is used when optimizing the beamforming vectors. The power allocation vector γ and the beamforming vectors w1 and w2 of the low-speed users are fixed. The optimization sub-problem for w0 is:

[0048]

[0049] The problem is transformed into a convex problem using the following transformation, that is, And use Tr(W0)≤1,Rank(W0=1,W 0,n,n ≥0,n=1,...,V rewrites constraint c 4b , where Tr(·) is the trace function, W 0,n,n are the main diagonal elements of the W0 matrix; by relaxing the constraints, the original optimization problem is transformed into:

[0050]

[0051] Tr(W0)≤1 (c 5b )

[0052] W 0，n，n ≥0,n=1,...,V (c 5c )

[0053] The Gaussian randomization technique is used to obtain the suboptimal solution. The power allocation vector γ and the beamforming vectors w0 and w2 are fixed, and the subproblem of optimizing w1 is as follows:

[0054]

[0055] Let the matrix And after relaxing the rank 1 matrix constraint, we have:

[0056]

[0057] Tr(W1)≤1 (c 7d )

[0058] W 1,n,n ≥0,n=1,...,V (c 7e )

[0059] Among them, (c 7e ) 1,n,n are the main diagonal elements of the W1 matrix; Gaussian randomization technique is used to recover the suboptimal solution of the problem.

[0060] A computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the processor implements the steps of a NOMA-OTFS-based vehicle network downlink communication system power allocation and beamforming method.

[0061] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a NOMA-OTFS-based vehicle network downlink communication system power allocation and beamforming method.

[0062] The beneficial effects of the present invention are:

[0063] 1. Compared with the traditional NOMA-OTFS system using beamforming technology, the present invention proposes a power allocation-beamforming joint optimization algorithm. By jointly optimizing the power allocation in the system and the beamforming vector of each user, it has higher resource utilization efficiency and communication quality, and greatly improves the system throughput and reliability.

[0064] 2. The power allocation-beamforming joint optimization algorithm based on convex optimization proposed in the present invention can effectively ensure that the system throughput converges to a predetermined range within 10 iterations to meet the performance requirements of the communication system for throughput. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a schematic block diagram of the structure of the vehicle networking downlink communication system based on NOMA-OTFS of the present invention.

[0066] Figure 2 The figure is a schematic diagram showing a simulation comparison of the system throughput performance of the power allocation-beamforming joint optimization algorithm used in the present invention and the optimization method in the literature [1] under the configuration of the number of carriers M = 8, the number of subcarriers N = 4, and the number of base station antennas V = 3.

[0067] Figure 3 It is a convergence diagram of the power allocation-beamforming joint optimization algorithm in the present invention.

[0068] Figure 4 It is a schematic diagram of the power allocation-beamforming joint optimization algorithm in the present invention. DETAILED DESCRIPTION

[0069] The present invention will be further defined below in conjunction with the accompanying drawings and embodiments, but is not limited thereto.

[0070] Example 1

[0071] The power allocation and beamforming method of the NOMA-OTFS-based vehicle networking downlink communication system is implemented through an optimization algorithm based on user channel state information;

[0072] like Figure 1 As shown, the downlink communication system of the Internet of Vehicles includes a base station, a distributed vehicle unit and a signal processing module; the base station is equipped with a multi-element transmitting antenna array; the distributed vehicle unit includes a high-speed vehicle user using DD domain modulation, namely a high-speed user, and a plurality of low-speed vehicle users using time-frequency (TF) domain modulation, namely a low-speed user; the signal processing module includes a serial interference canceller and a frequency domain linear equalizer;

[0073] Assume that the received noise is additive white Gaussian noise; including:

[0074] First, FD-LE is used at the high-speed user receiving end to recover the high-speed user signal in the DD domain. Then, SIC is used at each low-speed user receiving end to overcome the interference of high-speed users on low-speed users in the NOMA system and accurately recover the low-speed user signal.

[0075] Secondly, by analyzing the CSI corresponding to each user, the SINR of each user is calculated, and based on the SINR of each user, a joint optimization problem is constructed to maximize the total throughput of the downlink communication system of the Internet of Vehicles, and then the power allocation vector and the beamforming vector of each user are jointly optimized;

[0076] Finally, the optimal beamforming and power allocation vectors are output through continuous convex approximation and semi-positive definite programming algorithms.

[0077] Example 2

[0078] The power allocation and beamforming method of the vehicle networking downlink communication system based on NOMA-OTFS described in Example 1 is different in that:

[0079] The construction process of the Internet of Vehicles downlink communication system includes:

[0080] In the downlink of the NOMA-OTFS-based IoV downlink communication system, the base station serves one high-speed user and multiple low-speed users at the same time, and both the high-speed user and the low-speed user are configured with a single receiving antenna. Specifically, the high-speed user is served in the DD domain, and its signal is modulated using OTFS. The low-speed user, also known as the NOMA user, is served in the TF domain, and its signal is modulated using traditional OFDM. Assume that there are K+1 users U in the IoV downlink communication system. i , i = 0,…,K, where the high-speed user is denoted as U0 and the low-speed user is denoted as U j , j=1,…,K; Assume that each time-frequency resource block in the downlink communication system of the Internet of Vehicles includes N OFDM symbols and K+1≤M, where M is the number of system subcarriers; each low-speed user independently occupies subcarrier resources in the TF domain, and then the low-speed user information symbols and the high-speed user information symbols are superimposed through the NOMA technology, and at the transmitting end, the signal is customized for each user through a precise beamforming strategy to optimize the overall signal transmission quality and system performance.

[0081] In the downlink communication system of the Internet of Vehicles, the TF domain sending signal of the vth antenna of the base station is: where γ i and v,i ,i=0,…,K, respectively, are the power allocation coefficient of user i in the system and its beamforming weight on the vth path, is the signal sent by user i in the TF domain;

[0082] At the high-speed user, the information symbol of the high-speed user is directly mapped in the DD domain, and the low-speed user signal is regarded as interference. Therefore, the input and output relationship of the high-speed user DD domain in the system is Among them, y0 represents the received signal vector of the high-speed user, x i,i=0,…,K, represents the signal vector sent by the base station to user i in the system, that is, when i=0, x i represents the signal vector sent by the base station to the high-speed user, when i=j and j=1,…,K i represents the signal vector sent to the jth low-speed user, V is the number of transmitting antennas, represents the channel matrix from the vth antenna of the base station to the high-speed user, z0 is the additive white Gaussian noise vector at the high-speed user;

[0083] Similarly, the received signal vector of the jth low-speed user is in represents the channel matrix from the vth antenna of the base station to the jth low-speed user, z j is the additive white Gaussian noise vector at the low-speed user j; at the high-speed user receiving end, FD-LE is used to reduce the inter-symbol interference of the high-speed user in the DD domain and restore the high-speed user signal;

[0084] The high-speed user channel equalization matrix is ​​set as Among them, F N and F M is the Fourier transform matrix of N and M points, (·) H represents the complex conjugate transpose operation, is the equivalent channel matrix from the vth antenna of the base station to the high-speed user, is a diagonal matrix, and The (kM+l+1)th diagonal element of in, for The (nM+m+1)th row, first column element;

[0085] Equalize y0 to restore the high-speed user signal in the DD domain, and the equalized signal is:

[0086] The first term on the right side of the equation is the signal sent by the base station to the high-speed user, the second term is the interference term of the low-speed user, and the third term is the additive Gaussian white noise. Assuming that the signal amplitude transmitted by all users is the same and the noise power is normalized, the signal-to-noise ratio of each user is expressed as ρ = E{|x0[k,l|] 2}=E{|x j (n)| 2}, by calculating the expected signal, interference and noise power, the SINR at the high-speed user is derived as Among them, g k,l is the base station to high-speed user channel matrix, and

[0087] SIC is used at each low-speed user receiving end to overcome the interference of high-speed users on low-speed users in the NOMA system and accurately restore the low-speed user signal; including:

[0088] Perform DD domain detection on the equalized high-speed user signal, perform SIC at each low-speed user to eliminate the interference of the high-speed user, and then restore the signal of each low-speed user; specifically:

[0089] Assume that the channel of the low-speed user is a time-invariant channel, that is, the low-speed user does not experience Doppler frequency shift. Therefore, the channel matrix from the vth antenna to user j is is a block diagonal matrix, similar to the detection of high-speed user signals mentioned above, plus the detection matrix After that, the low-speed user signal is effectively balanced and restored, among which, is the equivalent channel matrix from the base station's vth antenna to low-speed user j, is a diagonal matrix, and its (l+1)th diagonal element is in, for The (m+1)th row, first column element of .

[0090] Calculate the SINR for each user; including:

[0091] Assuming that the high-speed user signal is perfectly demodulated and eliminated, the high-speed user U0 is obtained after the jth low-speed user U j The SINR under interference is: Among them, h j,l is the channel matrix from the base station to the jth low-speed user, so the TF domain low-speed user U after SIC is obtained j The received signal at in, is the time-frequency channel response from the vth antenna of the base station to the jth low-speed user, is the time-frequency domain transmitted signal of user j; assuming that all NOMA user channels are time-invariant channels, the SINR corresponding to the jth NOMA user is

[0092] Under the constraints of high-speed user QoS, perfect SIC execution for low-speed users, and total power, the NOMA-OTFS throughput is maximized to achieve efficient resource utilization and optimize user experience. The joint optimization problem of beamforming and power allocation is established as follows:

[0093]

[0094] Where P is the total transmission power of the base station; constraint c 1a is the high-speed user QoS and perfect SIC constraints, R0 is the minimum data rate required to meet the high-speed user QoS; constraint c1b represents the base station beamforming power constraint; constraint c 1c is the total system power constraint.

[0095] The solution is obtained through continuous convex approximation and semi-positive programming algorithm, and the optimal beamforming and power allocation vectors are output; Figure 4 As shown, the power allocation-beamforming joint optimization algorithm specifically includes:

[0096] The joint optimization problem is non-convex, and the high-speed user beamforming vector w0 and the j-th low-speed user beamforming vector w j and the power allocation vector γ are coupled with each other and are difficult to solve directly. Therefore, it can be decomposed into two optimization sub-problems, which are solved from the aspects of beamforming and power allocation respectively.

[0097] Assume that the number of NOMA users is 2 and the beamforming vector w0,w is fixed. j ,j=1,2, solve the power vector γ as follows:

[0098] First, the power allocation optimization sub-problem is transformed into:

[0099]

[0100] Convert the subproblem into a concave difference function, then:

[0101]

[0102] Using continuous convex approximation technology to further transform the function Transformed into a convex function, thus ensuring the concavity of the original objective function. At this time, due to the constraint c 3a Still non-convex, it can be transformed into and At this point, the problem of optimizing the transmit power has been transformed into a standard convex problem and can be solved using the CVX tool.

[0103] Secondly, by fixing the transmit power, we get the multi-user beamforming optimization subproblem. The objective function is also non-convex. By introducing the auxiliary variable The original objective function is transformed into And further by fixing the variables Transform the objective function into Since the objective function is a non-convex fraction and cannot be solved by the traditional convex optimization algorithm, we continue to introduce auxiliary variables Transform the original objective function into The objective function is used when optimizing the beamforming vectors. The power allocation vector γ and the beamforming vectors w1 and w2 of the low-speed users are fixed. The optimization sub-problem for w0 is:

[0104]

[0105] The objective function of this problem is non-concave, and both constraints are non-convex, so the problem can be transformed into a convex problem using the following transformation: And use Tr(W0)≤1,Rank(W0=1,W 0,n,n ≥0,n=1,...,V rewrites constraint c 4b , where Tr(·) is the trace function, W 0,n,n are the main diagonal elements of the W0 matrix; since the rank of the W0 matrix is ​​1 and the rank 1 constraint is a non-convex constraint, the original optimization problem can be transformed into:

[0106]

[0107] Tr(W0)≤1 (c 5b )

[0108] W 0,n,n ≥0,n=1,...,V (c 5c )

[0109] Since this problem is a convex problem, but its solution may not satisfy the rank 1 constraint, it is necessary to use Gaussian randomization technology to obtain its suboptimal solution; the power allocation vector γ and the beamforming vectors w0 and w2 are fixed, and the subproblem of optimizing w1 is as follows:

[0110]

[0111] Let the matrix And after relaxing the rank 1 matrix constraint, we have:

[0112]

[0113] Tr(W1)≤1 (c 7d )

[0114] W 1,n,n ≥0,n=1,...,V (c 7e )

[0115] Among them, (c 7e ) 1,n,n is the main diagonal element of the W1 matrix; the transformed problem is now a convex problem. Since the solution to this problem may not satisfy the rank 1 constraint, Gaussian randomization technology is required to recover the suboptimal solution to the problem. In addition, since users 1 and 2 are both NOMA users and user 2 does not perform additional processing, the steps for optimizing user 2's beamforming vector w2 are similar to those for w1.

[0116] Figure 2The figure is a schematic diagram of the simulation comparison of the power allocation-beamforming joint optimization algorithm used in the present invention and the optimization method in the literature [1] in terms of system throughput performance under the configuration of the number of carriers M = 8, the number of subcarriers N = 4, and the number of base station antennas V = 3. The horizontal axis is the transmission signal-to-noise ratio, and the vertical axis is the system throughput. It can be seen from the figure that under each transmission signal-to-noise ratio, the system throughput of the method of the present invention is improved by about 5bps / Hz compared with that in [1].

[0117] Figure 3 This is a schematic diagram of the convergence of the power allocation-beamforming joint optimization algorithm. The horizontal axis is the transmit signal-to-noise ratio, and the vertical axis is the system throughput. Figure 3 It can be seen that the algorithm can effectively converge after 4 iterations.

[0118] Set the maximum number of iterations, test the convergence of the algorithm, and obtain the optimal or suboptimal solution of the power allocation vector and each user's beamforming vector; at the same time, test the system throughput at different total transmit powers to evaluate the system optimization performance.

[0119] After the system optimization algorithm converges, the online deployment is implemented. First, the number of system antennas is set to 3, the number of OFDM symbols is set to 4, the number of subcarriers is set to 8, the maximum number of iterations is set to 50, and the convergence error threshold is set to 0.001. Based on the algorithm output γ,w0,w j , j = 1, 2 and Shannon’s limit theorem, the maximum throughput of the NOMA-OTFS-based IoV communication system is calculated, and the performance of the system is further evaluated by comparing with the optimization algorithm in reference [1].

[0120] Example 3

[0121] A computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the power allocation and beamforming method of the vehicle network downlink communication system based on NOMA-OTFS described in Example 1 or 2 are implemented.

[0122] Example 4

[0123] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the NOMA-OTFS-based vehicle network downlink communication system power allocation and beamforming method described in Example 1 or 2.

Claims

1. A power allocation and beamforming method for a downlink communication system of an Internet of Vehicles based on NOMA-OTFS, characterized in that: The vehicle networking downlink communication system includes a base station, a distributed vehicle unit and a signal processing module; the base station is equipped with a multi-element transmitting antenna array; the distributed vehicle unit includes a high-speed vehicle user using DD domain modulation, namely a high-speed user, and a plurality of low-speed vehicle users using time-frequency domain modulation, namely a low-speed user; The signal processing module includes a serial interference canceller and a frequency domain linear equalizer; The receiving noise is additive white Gaussian noise; including: First, the high-speed user signal in the DD domain is restored at the high-speed user receiving end, and then the low-speed user signal is restored at each low-speed user receiving end; Secondly, by analyzing the CSI corresponding to each user, the SINR of each user is calculated, and based on the SINR of each user, a joint optimization problem is constructed to maximize the total throughput of the downlink communication system of the Internet of Vehicles, and then the power allocation vector and the beamforming vector of each user are jointly optimized; Finally, the solution is obtained through continuous convex approximation and semi-positive definite programming algorithm, and the optimal beamforming and power allocation vectors are output; The construction process of the Internet of Vehicles downlink communication system includes: There are K+1 users U in the downlink communication system of the Internet of Vehicles i , i = 0,…,K, where the high-speed user is denoted as U0 and the low-speed user is denoted as U j , j=1,…,K; each time-frequency resource block in the downlink communication system of the Internet of Vehicles includes N OFDM symbols and K+1≤M, where M is the number of system subcarriers; each low-speed user independently occupies subcarrier resources in the TF domain, and then the low-speed user information symbols and high-speed user information symbols are superimposed through the NOMA technology, and at the transmitting end, the signal is customized for each user through precise beamforming strategy to optimize the overall signal transmission quality and system performance.

2. According to claim 1, the NOMA-OTFS-based vehicle networking downlink communication system power allocation and beamforming method is characterized in that: In the vehicle networking downlink communication system; The signal sent by the TF domain of the vth antenna of the base station is: where γ i and v,i ,i=0,…,K, respectively, are the power allocation coefficient of user i in the system and its beamforming weight on the vth path, is the signal sent by user i in the TF domain; The input and output relationship of the high-speed user DD domain is: Among them, y0 represents the received signal vector of the high-speed user, x i ,i=0,…,K, represents the signal vector sent by the base station to user i in the system, that is, when i=0, x i represents the signal vector sent by the base station to the high-speed user, when i=j and j=1,…,K i represents the signal vector sent to the jth low-speed user, V is the number of transmitting antennas, represents the channel matrix from the vth antenna of the base station to the high-speed user, z0 is the additive white Gaussian noise vector at the high-speed user; The received signal vector of the jth low-speed user is in represents the channel matrix from the vth antenna of the base station to the jth low-speed user, z j is the additive white Gaussian noise vector at the low-speed user j; The high-speed user channel equalization matrix is ​​set as Among them, F N and F M is the Fourier transform matrix of N and M points, (·) H represents the complex conjugate transpose operation, is the equivalent channel matrix from the vth antenna of the base station to the high-speed user, is a diagonal matrix, and The (kM+l+1)th diagonal element of in, for The (nM+m+1)th row, first column element; Equalize y0 to restore the high-speed user signal in the DD domain, and the equalized signal is: The first term on the right side of the equation is the signal sent by the base station to the high-speed user, the second term is the interference term of the low-speed user, and the third term is the additive Gaussian white noise. Assuming that the signal amplitude transmitted by all users is the same and the noise power is normalized, the signal-to-noise ratio of each user is expressed as ρ = E{x0[k,l] 2 }=E{x j (n) 2 }, by calculating the expected signal, interference and noise power, the SINR at the high-speed user is derived as Among them, g k,l is the base station to high-speed user channel matrix, and 3. The NOMA-OTFS-based vehicle networking downlink communication system power allocation and beamforming method according to claim 2 is characterized in that: Restoring low-speed user signals at each low-speed user receiving end; including: Perform DD domain detection on the equalized high-speed user signal, perform SIC at each low-speed user to eliminate the interference of the high-speed user, and then restore the signal of each low-speed user; specifically: The channel for low-speed users is a time-invariant channel, that is, low-speed users do not experience Doppler frequency shift. Therefore, the channel matrix from the vth antenna to user j is is a block diagonal matrix, plus the detection matrix After that, the low-speed user signal is effectively balanced and restored, among which, is the equivalent channel matrix from the base station's vth antenna to low-speed user j, is a diagonal matrix, and its (l+1)th diagonal element is in, for The (m+1)th row, first column element of .

4. The NOMA-OTFS-based vehicle networking downlink communication system power allocation and beamforming method according to claim 3 is characterized in that: Calculate the SINR for each user; including: The high-speed user signal is perfectly demodulated and eliminated, and the high-speed user U0 is obtained after the jth low-speed user U j The SINR under interference is: Among them, h j,l is the channel matrix from the base station to the jth low-speed user, so the TF domain low-speed user U after SIC is obtained j The received signal at in, is the time-frequency channel response from the vth antenna of the base station to the jth low-speed user, is the time-frequency domain transmission signal of user j; all NOMA user channels are time-invariant channels, then the SINR corresponding to the jth NOMA user is 5. According to claim 4, the NOMA-OTFS-based vehicle networking downlink communication system power allocation and beamforming method is characterized in that: Under the constraints of high-speed user QoS, perfect SIC execution for low-speed users, and total power, the NOMA-OTFS throughput is maximized to achieve efficient resource utilization and optimize user experience. The joint optimization problem of beamforming and power allocation is established as follows: Where P is the total transmission power of the base station; constraint c 1a is the high-speed user QoS and perfect SIC constraints, R0 is the minimum data rate required to meet the high-speed user QoS; constraint c 1b represents the base station beamforming power constraint; constraint c 1c is the total system power constraint.

6. The NOMA-OTFS-based vehicle networking downlink communication system power allocation and beamforming method according to claim 5 is characterized in that: The solution is solved by continuous convex approximation and semi-positive programming algorithm, and the optimal beamforming and power allocation vectors are output; the power allocation-beamforming joint optimization algorithm specifically includes: The number of NOMA users is 2, and the beamforming vectors w0,w are fixed. j ,j=1,2, solve the power vector γ as follows: First, the power allocation optimization sub-problem is transformed into: Convert the subproblem into a concave difference function, then: Using continuous convex approximation technology to further transform the function Transformed into a convex function, transformed into and Secondly, by introducing auxiliary variables The original objective function is transformed into And further by fixing the variables Transform the objective function into Continue to introduce auxiliary variables Transform the original objective function into The objective function is used when optimizing the beamforming vectors. The power allocation vector γ and the beamforming vectors w1 and w2 of the low-speed users are fixed. The optimization sub-problem for w0 is: The problem is transformed into a convex problem using the following transformation, that is, And use Tr(W0)≤1,Rank(W0=1,W 0,n,n ≥0,n=1,...,V rewrites constraint c 4b , where Tr(·) is the trace function, W 0,n,n are the main diagonal elements of the W0 matrix; by relaxing the constraints, the original optimization problem is transformed into: Tr(W0)≤1(c 5b ) W 0,n,n ≥0,n=1,...,V(c 5c ) The Gaussian randomization technique is used to obtain the suboptimal solution. The power allocation vector γ and the beamforming vectors w0 and w2 are fixed, and the subproblem of optimizing w1 is as follows: s.t.log2(1+SINR i I )≥R0,i=0,...,K(c 6a ) Let the matrix And after relaxing the rank 1 matrix constraint, we have: Tr(W1)≤1(c 7d ) W 1,n,n ≥0,n=1,...,V(c 7e ) Among them, (c 7e ) 1,n,n are the main diagonal elements of the W1 matrix; Gaussian randomization technique is used to recover the suboptimal solution of the problem.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the NOMA-OTFS-based vehicle network downlink communication system power allocation and beamforming method described in any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, the steps of the power allocation and beamforming method of the NOMA-OTFS-based vehicle network downlink communication system described in any one of claims 1-6 are implemented.

Images