阅读说明：本技术 一种基于noma的车联网广播通信资源分配方法 (NOMA-based vehicle networking broadcast communication resource allocation method ) 是由 宋铁成 蒋伟 胡静 王聪 夏玮玮 燕锋 沈连丰 于 2021-07-07 设计创作，主要内容包括：本发明提供一种基于NOMA的车联网广播通信资源分配方法,引入分簇概念,基站根据V2V用户的地理位置将V2V Tx用户归并成簇,统一簇内的V2VTx用户可共享统一子信道。对V2I用户数与V2V Tx用户的簇数做适当调整后,采用Kuhn-Munkres算法获取V2I用户与V2V Tx用户子信道共享结果,并采用基于PF函数的分布式多用户粒子群算法,获取V2V Tx用户的发射功率最优解。本发明将NOMA应用于V2V广播通信下行链路场景中,在确保V2I用户通信质量的基础上,获得V2V Rx用户吞吐量与传输时延的提升,并保证了V2V Rx用户公平性,为NOMA下的V2X资源分配问题提供了新的思路。(The invention provides a NOMA-based vehicle networking broadcast communication resource allocation method, which introduces a clustering concept, a base station merges V2V Tx users into clusters according to the geographic positions of V2V users, and V2VTx users in a unified cluster can share a unified sub-channel. After the number of V2I users and the cluster number of V2V Tx users are properly adjusted, a Kuhn-Munkres algorithm is adopted to obtain a sub-channel sharing result of the V2I users and the V2V Tx users, and a PF function-based distributed multi-user particle swarm algorithm is adopted to obtain an optimal solution of the transmitting power of the V2V Tx users. The invention applies NOMA to V2V broadcast communication downlink scene, on the basis of ensuring V2I user communication quality, the improvement of V2V Rx user throughput and transmission delay is obtained, the fairness of V2V Rx users is ensured, and a new thought is provided for the problem of V2X resource allocation under NOMA.)

1. A NOMA-based vehicle networking broadcast communication resource allocation method is characterized by comprising the following steps:

step 1, in the starting stage of semi-static scheduling, a base station BS merges V2V Tx into a cluster according to the geographic position of V2V Tx users in a coverage area, the number of users in the same cluster is at most θ, and the clustering condition is as follows: the distance of the V2V Tx users in each cluster is not lower than eta and the clustering process maximizes the minimum distance of the V2V Tx in each cluster, wherein the V2V Tx users are senders in the vehicle-to-vehicle communication user type;

V2V Tx users in the same cluster may share the same subchannel and each V2I user shares the subchannel with only one V2V Tx cluster, where the V2I users are cellular vehicle users;

step 2, adjusting the Kuhn-Munkres algorithm according to the number of clusters generated in the step 1 and the number of V2I users in the coverage area of the base station, calculating a distance matrix between each V2V Tx cluster and each V2I user, and recording the distance matrix as D_{V2I_Cluster}Applying the adjusted Kuhn-Munkres algorithm to the distance matrix D_{V2I_Cluster}Negative matrix of, i.e. -D_{V2I_Cluster}To obtain the sharing result of the V2I users and each V2I Tx cluster subchannel;

step 3, in the data transmission stage of semi-static scheduling, a V2V Tx user sends broadcast information to a V2V Rx user within the range that d is not more than r, wherein r is the maximum transmission distance of the V2V Tx user, and the V2V Rx user is a receiving user in the type of vehicle-to-vehicle communication user; the PSO algorithm is applied to a distributed scene by the V2V Rx user according to SIC decoding conditions, the signal-to-interference-and-noise ratio conditions of V2V Tx reception under NOMA and the condition that V2V Tx is limited to V2I user interference, a PF function is used as a fitness function, and iteration times and accuracy are set to obtain the optimal solution of the transmission power of the V2V Tx user.

2. The method for allocating the resources of the downlink of the NOMA-based broadcast communication of claim 1, wherein the step 1 specifically comprises the steps of:

step 1.1, defining an iteration variable t, setting t to 1, establishing identifiers for all V2V Tx in the coverage area of the base station BS, and recording the identifiers as Tx₁、Tx₂、…、Tx_PThe method comprises the steps that P V2V Tx users exist in the coverage range of a base station BS, the identifiers of the P V2V Tx users are automatically filled into an empty cluster by adopting a method for generating random numbers, the maximum V2V Tx specification number of a single cluster is recorded as theta, the empty cluster is filled, the next empty cluster is filled with the identifiers of the rest (P-theta) V2V Tx users, and the identifiers of all the V2V Tx users are generated until the identifiers of all the V2V Tx users are completely filledClustering; when in useIn the case of an integer, the number of the carbon atoms,otherwiseWherein the operation result of (P/theta) is an integer quotient value, and the remainder is not counted;

step 1.2, for the clusters generated in step 1.1, calculating V2V Tx vehicles in each clusterThe distance between the vehicles is found, and the minimum value of the distance between the V2V Tx vehicles in each cluster is found and recorded as

Step 1.3, find outMinimum value of (1), is notedDefine the minimum distance between V2V Tx users within a single cluster as η, ifReturning to the step 1.1; if it isContinuing to step 1.4;

step 1.4, let t be 2, repeat steps 1.1 to 1.3, if new V2V Tx clustering result is calculatedIs greater thanSelecting the V2V Tx clustering result when t is 2 as the updated clustering result, and then t is t + 1;

step 1.5, repeat step 1.1 to step 1.4 until repeating continuous N₀And after that, the result is not updated any more, and the V2V Tx cluster corresponding to the last update is taken as the final V2V Tx clustering result.

3. The method for allocating the resources of the downlink of the NOMA-based broadcast communication of claim 2, wherein the step 2 comprises the following steps:

step 2.1, calculating the distance between each V2I user and each V2V Tx cluster generated in step 1, wherein the calculation method is to calculate the minimum value of the distance between the V2I user and the V2V Tx user in each cluster, and take the minimum value as the distance between the V2I user and the V2V Tx cluster;

step 2.2, if there are M V2I users in the coverage area of the BS, and each V2I user does not share the same subchannel with other V2I users, the distance between the V2I user obtained in step 2.1 and the V2V Tx cluster isA number value, which is stored in a dimension ofIn the matrix of (2), is denoted asWherein the content of the first and second substances,the number of V2V Tx clusters generated for the clustering algorithm in claim 2;

step 2.3, ifDirectly on the distance matrix in step 2.2Negative matrix ofAcquiring a subchannel sharing result of M V2I users and a V2V Tx cluster by adopting a Kuhn-Munkres algorithm;

step 2.4, ifSupplementA virtual V2V Tx cluster, i.e. atIncrease on the right side ofAll zero columns, so that the new matrix is a dimension (M × M) square matrix, notedTo pairA negative matrix of, i.e.Acquiring a subchannel sharing result of M V2I users and a V2V Tx cluster by adopting a Kuhn-Munkres algorithm;

step 2.5, ifSupplementDistance matrix of V2I user in step 2.2Under supplement ofAll zero rows, making the new matrix dimensionIs marked asTo pairIs a negative matrix ofAnd acquiring the subchannel sharing result of the M V2I users and the V2V Tx cluster by adopting a Kuhn-Munkres algorithm.

4. The method for allocating the resources of the downlink of the NOMA-based broadcast communication over the internet of vehicles according to claim 3, wherein the step 3 comprises the following steps:

step 3.1, calculating the distance value between all the V2V Rx users in the coverage area of the base station BS and each V2V Tx user, and storing the calculated value in a matrix with dimension Q multiplied by P, wherein Q is the number of V2V Rx users in the coverage area of the base station BS, P is the number of V2V Tx users in the coverage area of the base station BS, and the matrix is

Step 3.2, establishing an auxiliary matrix with dimension of QXPFor the Wherein i is the symbol of V2V Rx, j is the symbol of V2V Tx, and the logical symbolMeans at will, ifThenOtherwiseIf it isIt means that V2V Rx i receives the broadcast data of V2V Tx j;

step 3.3, establish transmission rate calculations for receiving each V2V Tx broadcast information under NOMA for each V2V Rx user in the coverage area of the base station BS.

5. The method for allocating resources in the downlink of NOMA-based car networking broadcast communication according to claim 4, wherein step 3.3 comprises the following steps:

step 3.3.1, recording the signal-to-interference-and-noise ratio of the V2V Rx user i receiving the V2V Tx user j broadcast data to be calculated currently as

Wherein i is the identifier of the V2V Rx user, and i is more than or equal to 1 and less than or equal to Q; j, k is the ID of V2V Tx user, j is more than or equal to 1, and P is more than or equal to k; m is the identification of the V2I user, and M is more than or equal to 1 and less than or equal to M;for the transmit power of V2V Tx user j,for a transmit power of V2I user m,is the transmit power, σ, of V2V Tx user k²Is the noise power;channel gain coefficients for V2V Rx user i through V2V Tx user j,channel gain coefficients for V2V Rx users i through V2I users m,channel gain coefficients for V2V Rx users i through V2V Tx user k; beta is a_i，mIs a binary variable when beta_i，mWhen 1, V2V Rx user i shares the same subchannel with V2I user m; when beta is_i，mWhen 0, V2V Rx user i and V2I user m occupy different sub-channels; tau is_i，kIs a binary variable whenTime, τ_i，kWhen V2V Rx user i receives the broadcast information of V2V Tx user j at 1; otherwise whenTime, τ_i，kWhen V2V Rx user i is not in the coverage of V2V Tx user j, it cannot receive the broadcast information of V2V Tx user j; rho_i，kThe binary variable is specifically defined as follows:

step 3.3.2, considering the bearing capacity of the V2I user to the interference of the V2V Tx user under the co-channel, considering that for the V2I user, the signal-to-interference-and-noise ratio γ of the V2I user is guaranteed_mLower limit of (d), Pr { gamma is set_m≤γ₀}≤p₀Wherein Pr { } represents a probability distribution function, γ₀Minimum signal-to-interference-and-noise ratio, p, of acceptable V2I users₀For V2I user signal-to-interference-and-noise ratio less than gamma₀A probability threshold of time lapse; signal to interference plus noise ratio gamma of V2I user_mThe signal to interference plus noise ratio (SINR) is as follows:

whereinFor the transmit power of the base station BS for V2I user m,for the transmit power of V2V Tx user i,is the channel gain constant between base station BS and V2I user m,is the channel gain constant, μ, between V2V Tx user i and V2I user m_i，mBeing a binary variable, μ when V2V Tx user i shares the same subchannel with V2I user_i，m1, otherwise μ_i，m＝0；σ²Is the noise power;

step 3.3.3, defining slot variable slot, making slot equal to 1, and generating N_SVector of dimension P x 1Calculating a corresponding proportional fair function (PF) function value, and taking the PF function value as a fitness function value of the PSO algorithm so as to obtain an optimal PF function value of the first time slot;

step 3.3.4, let slot be slot +1, repeat step 3.3.3 until slot > N_SPSIs stopped, wherein N_SPSThe number of slots of the data transmission phase of the SPS.

6. A method of resource allocation for the downlink of NOMA-based broadcast communication over the Internet of vehicles according to claim 5, wherein γ in step 3.3.2_mThe signal to interference plus noise ratio formula is approximately operated, namely:

wherein the corner mark i^*An identifier indicating a V2V Tx user closest to the V2I user m; because of the SINR QoS condition of V2I user m, i.e., Pr { gamma }_m≤γ₀}≤p₀It was transformed as follows:

note the bookWherein alpha is_Bm、Is a large scale fading coefficient, g_Bm、Is a small scale fading coefficient, further converting gamma_m≤γ₀Wherein, γ₀For the minimum signal-to-interference-and-noise ratio experienced by the V2I user, we obtain:

therefore, it isSo Pr { gamma_m≤γ₀Denotes the double integral:

thus, Pr { γ }_m≤γ₀}≤p₀Expressed as:

V2V Tx user i to be closest to V2I user m^*The transmission power is converted into:

7. method for resource allocation in the downlink for NOMA-based radio over internet protocol communications according to claim 6, wherein said step 3.3.3 comprises in particular the steps of:

step 3.3.3.1, V2V Tx user i closest to V2I user m and sharing the same sub-channel in step 3.3.2^*Initializing the transmission power of all V2V Tx users under the constraint of transmission power, namely generating N_SPower value vector with dimension P x 1The generation method comprises the following steps: for V2V Rx users receiving V2V Tx user broadcast information, the calculation amount of a PSO algorithm is shared, and power distribution and data receiving conditions of the V2V Tx users in a cross region are exchanged among different V2V Rx users;

at step 3.3.3.2, it is set that the data transmission phase in an SPS period has N_SPSDefining iterative variable itera in PSO algorithm, calculating transmission rate of each V2V Rx user when itera is equal to 1, and recording all transmission rate values in RateRxTxTxTxTx with dimension Q × p_Q×PIn the matrix, RateRxTx (i, j) represents the transmission rate at which V2V Rx user i receives V2V Tx user j broadcast information:

wherein E [.]Is a desired operator;indicating that NOMA is applied to the V2V user scenario, the signal to interference plus noise ratio of the broadcast information received by V2V Rx user i to V2V Tx user j is given in step 3.3.1; the transmission rate of all V2V Tx broadcast information that V2V Rx user i can receive is accumulated, namely RateRxTxTx_Q×PSumming the rows of the matrix, and storing the total transmission rate of V2V Rx in the dimension of Q × N_SPSIn matrix ofIn (1), initializeIs an all-zero matrix, i.e.:

wherein P is the total number of V2V Tx users in the coverage area of the base station BS;

step 3.3.3.3, calculating a fitness function fitness value at the time of the current slot, wherein the fitness function fitness is defined as follows:

when slot is 1, the initial value of the weight vector with dimension 1 × Q is weight (1, 1., 1)

fitness＝weight·RateRxSlot(：，slot)

＝(1，1，...，1)(RateRxSlot(1，slot)，...，RateRxSlot(Q，slot))^T；

When the slot is more than 1 and less than or equal to N_SPSIn time, the ith component of weight is noted as weight_iWherein i is more than or equal to 1 and less than or equal to Q, and the calculation method comprises the following steps:

fitness＝weight·RateRxSlot(：，slot)

＝(weight₁，...，weight_Q)(RateRxSlot(1，slot)，...，RateRxSlot(Q，slot))^T

step 3.3.3.4, for N_SPower value vector with dimension P x 1Respectively calculate corresponding N_SA fitness value; corresponding power value vector of maximum fixness valueIs marked asCorresponding the current maximum fitness value of each power vector to a power value vectorIs marked asGenerate eachUpdate gradient vector ofWherein:

when itera is equal to 1,

v_tran (1 × P), wherein v_tFor the t power value vectorA corresponding update gradient vector;

when itera > 1, the first time,

wherein the content of the first and second substances,the t power value vector after the last PSO iteration updating is obtained;

wherein rand generates random numbers between (0, 1), c₁For updating the speed coefficient of each power vector itself, c₂Updating the speed coefficient for the global of each power vector;

step 3.3.3.5 Using the updated gradient vector of step 3.3.3.4To N_SPower value vector with dimension P x 1The updating is carried out, and the updating is carried out,

step 3.3.3.6, repeat 3.3.3.2 to 3.3.3.5 until itera is maxgen, where maxgen is the maximum iteration number, and finally obtain the optimal fitness value of maxgen iteration and the corresponding optimal power value vector

Technical Field

The invention belongs to the technical field of wireless communication, and particularly relates to a NOMA-based vehicle networking broadcast communication resource allocation method.

Background

The resource allocation of the Internet of vehicles is an important factor for improving the performance of the Internet of vehicles, and has an important role in optimizing the Internet of vehicles. The centralized resource allocation of the base station BS is established in such a way that the base station can acquire Channel State Information (CSI) of all users in the coverage area, whereas the vehicle channels in the car networking scenario are fast time-varying channels, and the centralized resource allocation method is no longer applicable. In addition, due to the limited wireless spectrum resources and power resources, the network interface size in the car networking is limited under the traditional OMA. The internet of vehicles resource allocation is therefore challenging in many ways.

The resource allocation of the Internet of vehicles has optimization requirements on frequency spectrum resources and power resources, and under the condition that vehicle users increase day by day, the network access requirements of the users of the Internet of vehicles gradually increase, so that the security of the Internet of vehicles is challenged. Therefore, a method for allocating broadcast communication resources of the internet of vehicles based on the NOMA is needed to be developed. Most researches on centralized resource allocation of a base station BS are common, and a D2D-based vehicle networking resource allocation method, a graph theory-based V2V resource allocation method, and the like are common. Centralized resource allocation of a base station BS is difficult to implement in an internet of vehicles communication scenario due to difficulty in acquiring a fast time-varying channel CSI. In addition to the above reasons, the vehicle networking resource allocation method based on D2D still occupies a large amount of calculation overhead at the BS end of the base station, and is not suitable for the V2V communication scenario, and most of the existing V2V resource allocation method documents based on graph theory have limitations on the scale of the graph and are difficult to adapt to the large connection demand scenario of the vehicle networking. In addition, the existing vehicle networking resource allocation scheme mostly focuses on the improvement of the throughput of the vehicle networking system, and the research on the fairness of the vehicle networking users is relatively deficient.

Disclosure of Invention

The invention aims to provide a NOMA-based vehicle networking broadcast communication resource allocation method to solve the technical problem.

In order to solve the technical problems, the specific technical scheme of the invention is as follows:

a NOMA-based vehicle networking broadcast communication resource allocation method comprises the following steps:

step 1, in the starting stage of semi-persistent scheduling (SPS), a base station BS merges V2V Tx into clusters according to the geographic position of a user of a sender (V2V Tx) in the types of vehicle-to-vehicle communication users in a coverage area, the number of users in the same cluster is at most theta, and the clustering condition is as follows: the distance of the V2V Tx users in each cluster is not lower than eta and the clustering process maximizes the minimum spacing of V2V Tx in each cluster;

V2V Tx users within the same cluster may share the same subchannel, and each cellular vehicle user V2I user shares the subchannel with only one V2V Tx cluster:

step 2, adjusting the Kuhn-Munkres algorithm according to the number of clusters generated in the step 1 and the number of V2I users in the coverage area of the base station, calculating a distance matrix between each V2V Tx cluster and each V2I user, and recording the distance matrix as D_{V2I_Cluster}Applying the adjusted Kuhn-Munkres algorithm to the distance matrix D_{V2I_Cluster}Negative matrix of, i.e. -D_{V2I_Cluster}To obtain the sharing result of the V2I users and each V2I Tx cluster subchannel;

and 3, in the data transmission stage of the SPS, the V2V Tx user sends broadcast information to V2V Rx users within the range that d is not more than r (wherein r is the maximum transmission distance of the V2V Tx user), the V2V Rx users apply the PSO algorithm to a distributed scene according to SIC decoding conditions, the signal-to-interference-and-noise ratio conditions of receiving each V2V Tx under NOMA and the condition that the interference of the V2V Tx to the V2I user is limited, and the PF function is used as a fitness function to set iteration times and accuracy to obtain the optimal solution of the transmission power of the V2V Tx user.

Further, step 1 specifically includes the following steps:

step 1.1, defining an iteration variable t, setting t to 1, establishing identifiers for all V2V Tx in the coverage area of the base station BS, and recording the identifiers as Tx₁、Tx₂、…、Tx_PThe method comprises the steps that P V2V Tx users exist in the coverage range of a base station BS, the identifiers of the P V2V Tx users are automatically filled into an empty cluster by adopting a method for generating random numbers, the maximum V2V Tx specification number of a single cluster is recorded as theta, the empty cluster is filled, the next empty cluster is filled with the identifiers of the rest (P-theta) V2V Tx users, and the like until the identifiers of all the V2V Tx users are completely filled, and the identifiers are generatedClustering; when in useIn the case of an integer, the number of the carbon atoms,otherwiseWherein the operation result of (P/theta) is an integer quotient value, and the remainder is not counted;

step 1.2, for the clusters generated in step 1.1, calculating the distance between the V2V Tx vehicles in each cluster, and finding out the minimum value of the distance between the V2V Tx vehicles in each cluster, and recording the minimum value as the distance between the V2V Tx vehicles in each cluster

Step 1.3, find outMinimum value of (1), is notedDefine the minimum acceptable distance between V2V Tx users within a single cluster as η, ifReturning to the step 1.1; if it isContinuing to step 1.4;

step 1.4, let t be 2, repeat steps 1.1 to 1.3, if new V2V Tx clustering result is calculatedIs greater thanSelecting the V2V Tx clustering result when t is 2 as the updated clustering result, and then t is t + 1;

and step 1.5, repeating the steps 1.1 to 1.4 until the result is not updated after repeating the continuous N0 times, and taking the V2V Tx cluster corresponding to the last update as the final V2V Tx clustering result.

Further, the step 2 specifically comprises the following steps:

step 2.1, calculating the distance between each V2I user and each V2V Tx cluster generated in step 1, wherein the calculation method is to calculate the minimum value of the distance between the V2I user and the V2V Tx user in each cluster, and take the minimum value as the distance between the V2I user and the V2V Tx cluster;

step 2.2, if there are M V2I users in the coverage area of the BS, and each V2I user does not share the same subchannel with other V2I users, the distance between the V2I user obtained in step 2.1 and the V2V Tx cluster is(wherein,number of V2V Tx clusters generated for the clustering algorithm in claim 2) and storing these distance values in the dimension ofIn the matrix of (2), is denoted as

Step 2.3, ifDirectly on the distance matrix in step 2.2A negative matrix of, i.e.Acquiring a subchannel sharing result of M V2I users and a V2V Tx cluster by adopting a Kuhn-Munkres algorithm;

step 2.4, ifSupplementA virtual V2V Tx cluster, i.e. atIncrease on the right side ofAll zero columns, so that the new matrix is a dimension (M × M) square matrix, notedTo pairA negative matrix of, i.e.Acquiring a subchannel sharing result of M V2I users and a V2V Tx cluster by adopting a Kuhn-Munkres algorithm;

step 2.5, ifSupplementDistance matrix of V2I user in step 2.2Under supplement ofAll zero rows, making the new matrix dimensionIs marked asTo pairA negative matrix of, i.e.And acquiring the subchannel sharing result of the M V2I users and the V2V Tx cluster by adopting a Kuhn-Munkres algorithm.

Further, step 3 specifically includes the following steps:

step 3.1, calculating the distance value between all the V2V Rx users in the coverage area of the base station BS and each V2VTx user, and storing the calculated value in a matrix with dimension Q multiplied by P, wherein Q is the number of V2V Rx users in the coverage area of the base station BS, and P is the coverage area of the base station BSThe V2VTx user number in the matrix is

Step 3.2, establishing an auxiliary matrix with dimension of QXPFor the (where i is the designation of V2V Rx, j is the designation of V2V Tx, logical notationDenotes "arbitrary") ifThenOtherwiseIf it isIt means that V2V Rx i receives the broadcast data of V2V Tx j;

step 3.3, establish transmission rate calculations for receiving each V2V Tx broadcast information under NOMA for each V2V Rx user in the coverage area of the base station BS.

Further, step 3.3 specifically includes the following steps:

step 3.3.1, recording the signal-to-interference-and-noise ratio of the V2V Rx user i receiving the V2V Tx user j broadcast data to be calculated currently as

Wherein i is the identifier of the V2V Rx user, and i is more than or equal to 1 and less than or equal to Q; j, k is the ID of V2V Tx user, j is more than or equal to 1, and P is more than or equal to k; m is the identification of the V2I user, and M is more than or equal to 1 and less than or equal to M;for the transmit power of V2V Tx user j,for a transmit power of V2I user m,is the transmit power, σ, of V2V Tx user k²Is the noise power;channel gain coefficients for V2V Rx user i through V2V Tx user j,channel gain coefficients for V2V Rx users i through V2I users m,channel gain coefficients for V2V Rx users i through V2V Tx user k; beta is a_i，mIs a binary variable when beta_i，mWhen 1, V2V Rx user i shares the same subchannel with V2I user m; when beta is_i，mWhen 0, V2V Rx user i and V2I user m occupy different sub-channels; tau is_i，kIs a binary variable whenTime, τ_i，kWhen V2V Rx user i receives the broadcast information of V2V Tx user j at 1; otherwise whenTime, τ_i，k＝0，At this time, the V2V Rx user i is not in the coverage of the V2V Tx user j, and cannot receive the broadcast information of the V2V Tx user j; rho_i，kThe binary variable is specifically defined as follows:

step 3.3.2, considering the bearing capacity of the V2I user to the interference of the V2V Tx user under the co-channel, considering that for the V2I user, the signal-to-interference-and-noise ratio γ of the V2I user is guaranteed_mLower limit of (d), Pr { gamma is set_m≤γ₀}≤p₀(wherein Pr {. denotes a probability distribution function, γ)₀Minimum signal-to-interference-and-noise ratio, p, of acceptable V2I users₀For V2I user signal-to-interference-and-noise ratio less than gamma₀A probability threshold of affordability); wherein, the signal-to-interference-and-noise ratio gamma of the V2I user_mThe signal to interference plus noise ratio (SINR) is as follows:

whereinFor the transmit power of the base station BS for V2I user m,for the transmit power of V2V Tx user i,is the channel gain constant between base station BS and V2I user m,is the channel gain constant, μ, between V2VTx user i and V2I user m_i，mBeing a binary variable, μ when V2V Tx user i shares the same subchannel with V2I user_i，m1, otherwise μ_i，m＝0；σ²Is the noise power;

step 3.3.3, defining slot variable slot, making slot equal to 1, and generating N_SVector of dimension P x 1Calculating a corresponding proportional fair function (PF) function value, and taking the PF function value as a fitness function value of a PS0 algorithm to obtain an optimal PF function value of a first time slot;

step 3.3.4, let slot be slot +1, repeat step 3.3.3 until slot > N_SPS(wherein, N is_SPSThe number of slots that are the data transmission phase of the SPS).

Further, the impact of V2V Tx users on V2I users in said step 3.3.2 is mainly the impact of the transmit power of V2V Tx users who are closest to V2I users and share the same sub-channel with V2I users; in the case where NOMA is applied to V2V Tx, the channel difference between V2V Tx and V2V Rx is large, and the transmission power attenuates fast with distance, the distribution between vehicles satisfies the poisson distribution, and thus the above-described γ can be matched_mThe signal to interference plus noise ratio formula is approximately operated, namely:

wherein the corner mark i^*An identifier indicating a V2V Tx user closest to the V2I user m; considering the SINR QoS condition of V2I user m, i.e. Pr { gamma }_m≤γ₀}≤p₀It can be transformed as follows:

note the bookWherein alpha is_Bm、Is a large scale fading coefficient, g_Bm、Is a small scale fading coefficient, then gamma can be further converted_m≤γ₀(wherein, γ)₀The minimum signal to interference plus noise ratio that can be tolerated for the V2I user) yields:

therefore, it isSo Pr { gamma_m≤γ₀It can be expressed as a double integral:

thus, Pr { γ }_m≤γ₀}≤p₀Expressed as:

V2V Tx user i to be closest to V2I user m^*The transmission power is converted into:

further, the step 3.3.3 specifically includes the following steps:

step 3.3.3.1, V2V Tx user i closest to V2I user m and sharing the same sub-channel in step 3.3.2^*Initializing the transmission power of all V2V Tx users under the constraint of transmission power, namely generating N_SPower value vector with dimension P x 1The generation method comprises the following steps: for V2V Rx users receiving V2V Tx user broadcast information, the calculation amount of a PSO algorithm is shared, and power distribution and data receiving conditions of the V2V Tx users in a cross region are exchanged among different V2V Rx users;

at step 3.3.3.2, it is set that the data transmission phase in an SPS period has N_SPSDefining iterative variable itera in PSO algorithm, calculating transmission rate of each V2V Rx user when itera is equal to 1, and recording all transmission rate values in RateRxTxTxTxTx with dimension Q multiplied by P_Q×PIn the matrix, RateRxTx (i, j) represents the transmission rate at which V2V Rx user i receives V2V Tx user j broadcast information:

wherein E [.]Is a desired operator;indicating that NOMA is applied to the V2V user scenario, the signal to interference plus noise ratio of the broadcast information received by V2V Rx user i to V2V Tx user j is given in step 3.3.1; the transmission rate of all V2V Tx broadcast information that V2V Rx user i can receive is accumulated, namely RateRxTxTx_Q×PSumming the rows of the matrix, and storing the total transmission rate of V2V Rx in the dimension of Q × N_SPSIn matrix ofIn (1), initializeIs an all-zero matrix, i.e.:

wherein P is the total number of V2V Tx users in the coverage area of the base station BS;

step 3.3.3.3, calculating a fitness function fitness value at the time of the slot at present, wherein the fitness function fitness is defined as follows:

when slot is 1, the initial value of the weight vector with dimension 1 × Q is weight (1, 1., 1)

fitness＝weight·RateRxSlot(：，slot)

＝(1，1，...，1)(RateRxSlot(1，slot)，...，RateRxSlot(Q，slot))^T；

When the slot is more than 1 and less than or equal to N_SPSIn time, the ith component of weight is noted as weight_iWherein i is more than or equal to 1 and less than or equal to Q, and the calculation method comprises the following steps:

fitness＝weight·RateRxSlot(：，slot)

＝(weight₁，...，weight_Q)(RateRxSlot(1，slot)，...，RateRxSlot(Q，slot))^T

step 3.3.3.4, for N_SPower value vector with dimension P x 1Respectively calculate corresponding N_SA fitness value; corresponding power value vector of maximum fixness valueIs marked asCorresponding the current maximum fitness value of each power vector to a power value vectorIs marked asGenerate eachUpdate gradient vector ofWherein:

when itera is equal to 1,

v_trand (1 × P); (wherein, v_tFor the t power value vectorCorresponding update gradient vector)

When itera > 1, the first time,

(wherein,for the t power value vector after the last PSO iteration update)

Wherein rand generates random numbers between (0, 1), c₁For updating the speed coefficient of each power vector itself, c₂Updating the speed coefficient for the global of each power vector;

step 3.3.3.5 Using the updated gradient vector of step 3.3.3.4To N_SPower value vector with dimension P x 1The updating is carried out, and the updating is carried out,

step 3.3.3.6, repeat 3.3.3.2 to 3.3.3.5 until itera is maxgen, where maxgen is the maximum iteration number, and finally obtain the optimal fitness value of maxgen iteration and the corresponding optimal power value vector

The invention discloses a NOMA-based vehicle networking broadcast communication resource allocation method, which has the following advantages:

the invention applies NOMA to V2V broadcast communication downlink scene, on the basis that a base station BS reserves sub-channels for V2V Tx users, the problem of sub-channel sharing between V2V users and V2I users is solved based on a Kuhn-Munkres method, and then the problems of V2V Rx user throughput, transmission delay and user fairness are solved by using a PF function-based distributed multi-user particle swarm algorithm.

Drawings

FIG. 1 is a flat road Internet of vehicles broadcast communication downlink scenario diagram of the present invention;

FIG. 2 is a flow chart of a NOMA-based vehicle networking broadcast communication resource allocation method of the present invention;

Detailed Description

In order to better understand the purpose, structure and function of the present invention, the following describes a method for allocating broadcast communication resources in car networking based on NOMA in further detail with reference to the accompanying drawings.

As shown in fig. 1, where BS is a base station, V2V Tx is a sender in a vehicle-to-vehicle communication user type, V2V Rx is a receiver in a vehicle-to-vehicle communication user type, V2I is a cellular vehicle user, R is a base station service radius, and d0 is a distance from the base station to a straight road.

A method for allocating downlink resources for broadcast communication over internet of vehicles based on NOMA, as shown in fig. 1 and 2, comprising the following steps:

step 1, in the starting stage of semi-persistent scheduling (SPS), a base station BS merges V2V Tx into clusters according to the geographic position of a user of a sender (V2V Tx) in the types of vehicle-to-vehicle communication users in a coverage area, the number of users in the same cluster is at most theta, and the clustering condition is as follows: 1. the distance of the V2V Tx users in each cluster is not lower than eta; 2. the clustering process maximizes the minimum spacing of V2V Tx within each cluster. The V2V Tx users in the same cluster can share the same subchannel, and each V2I user only shares the subchannel with one V2V Tx cluster, and the specific steps are as follows:

step 1.1, setting t to 1, establishing identifiers for all V2V Tx in the coverage area of the base station BS, and recording the identifiers as Tx₁、Tx₂、…、Tx_P(assuming that there are P V2V Tx users in the coverage area of the base station BS), adopting the method of generating random number, willThe identifications of P V2V Tx users are automatically filled into an empty cluster, the maximum V2V Tx specification number of a single cluster is recorded as theta, the empty cluster is filled, the next empty cluster is filled with the identifications of the rest (P-theta) V2V Tx users, and the like is carried out until the identifications of all V2V Tx users are completely filled, and the generation of the empty cluster is realizedAnd (4) clustering. When in useIn the case of an integer, the number of the carbon atoms,otherwiseWherein the operation result of (P/theta) is an integer quotient value, and the remainder is not counted;

step 1.2, for the clusters generated in step 1.1, calculating the distance between the V2V Tx vehicles in each cluster, and finding out the minimum value of the distance between the V2V Tx vehicles in each cluster, and recording the minimum value as the distance between the V2V Tx vehicles in each cluster

Step 1.3, find outMinimum value of (1), is notedDefine the minimum acceptable distance between V2V Tx users within a single cluster as η, ifReturning to the step 1.1; if it isContinuing to step 1.4;

step 1.4, let t be 2, repeat steps 1.1 to 1.3 if new V2V Tx clustering result is calculatedIs greater thanSelecting the V2V Tx clustering result when t is 2 as the updated clustering result, and then t is t + 1;

and step 1.5, repeating the steps 1.1 to 1.4 until the result is not updated after repeating the continuous N0 times, and taking the V2V Tx cluster corresponding to the last update as the final V2V Tx clustering result.

Step 2, adjusting the Kuhn-Munkres algorithm according to the number of clusters generated in the step 1 and the number of V2I users in the coverage area of the base station, calculating a distance matrix between each V2V Tx cluster and each V2I user, and recording the distance matrix as D_{V2I_Cluster}Applying the adjusted Kuhn-Munkres algorithm to the distance matrix D_{V2I_Cluster}Negative matrix of, i.e. -D_{V2I_Cluster}To obtain the sharing result of the V2I user and each V2ITx cluster subchannel, the specific steps are as follows:

step 2.1, calculating the distance between each V2I user and each V2V Tx cluster generated in step 1, wherein the calculation method is to calculate the minimum value of the distance between the V2I user and the V2V Tx user in each cluster, and take the minimum value as the distance between the V2I user and the V2V Tx cluster;

step 2.2, if there are M V2I users in the coverage of the BS (assuming that each V2I user does not share the same subchannel with other V2I users), the distance between the V2I user obtained in step 2.1 and the V2V Tx cluster isThe numerical values, wherein,the number of V2V Tx clusters generated for the clustering algorithm in claim 2, these distance values being stored in the dimension ofIn the matrix of (2), is denoted as

Step 2-3, ifDirectly on the distance matrix in step 2.2A negative matrix of, i.e.Acquiring a subchannel sharing result of M V2I users and a V2V Tx cluster by adopting a Kuhn-Munkres algorithm;

step 2.4, ifSupplementA virtual V2V Tx cluster, i.e. atIncrease on the right side ofAll zero columns, so that the new matrix is a dimension (M × M) square matrix, notedTo pairA negative matrix of, i.e.Acquiring a subchannel sharing result of M V2I users and a V2V Tx cluster by adopting a Kuhn-Munkres algorithm;

step 2.5, ifSupplementDistance matrix of V2I user in step 2.2Under supplement ofAll zero rows, making the new matrix dimensionIs marked asTo pairA negative matrix of, i.e.Acquiring a subchannel sharing result of M V2I users and a V2V Tx cluster by adopting a Kuhn-Munkres algorithm;

step 3, in the data transmission stage of semi-persistent scheduling (SPS), a V2V Tx user sends broadcast information to a V2V Rx user within the range that d is not more than r, wherein r is the maximum sending distance of the V2V Tx user, and the V2V Rx user is a receiving user in the type of vehicle-to-vehicle communication user; the method comprises the following steps that the V2V Rx user applies a PSO algorithm to a distributed scene according to SIC decoding conditions, the conditions of receiving signal-to-interference-and-noise ratios of V2V Tx under NOMA and the conditions that V2V Tx is limited to interference of the V2I user, a PF function is used as a fitness function, iteration times and accuracy are set, and the optimal solution of the transmitting power of the V2V Tx user is obtained, and the method specifically comprises the following steps:

step 3.1, for all V2V Rx users in the coverage area of the base station BS, calculating the distance between the user and each V2VTx userSeparating the value and storing the calculated value in a matrix with the dimension of Q multiplied by P, wherein Q is the number of V2V Rx users in the coverage area of the base station BS, and P is the number of V2VTx users in the coverage area of the base station BS, and the matrix is recorded as

Step 3.2, establishing an auxiliary matrix with dimension of QXPFor the (where i is the designation of V2V Rx, j is the designation of V2V Tx, logical notationDenotes "arbitrary") ifThenOtherwiseIf it isIt means that V2V Rx i receives the broadcast data of V2V Tx j.

Step 3.3, establishing transmission rate calculation for receiving each V2V Tx broadcast information under NOMA for each V2V Rx user in the coverage area of the base station BS, which comprises the following specific steps:

step 3.3.1, recording the signal-to-interference-and-noise ratio of the V2V Rx user i receiving the V2V Tx user j broadcast data to be calculated currently as

Wherein i is the ID of V2V Rx user (i is more than or equal to 1 and less than or equal to Q), j, k is the ID of V2V Tx user (j is more than or equal to 1 and k is less than or equal to P), and M is the ID of V2I user (M is more than or equal to 1 and less than or equal to M).For the transmit power of V2V Tx user j,for a transmit power of V2I user m,is the transmit power, σ, of V2V Tx user k²Is the noise power.Channel gain coefficients for V2V Rx user i through V2V Tx user j,channel gain coefficients for V2V Rx users i through V2I users m,channel gain coefficients for V2V Rx users i through V2V Tx users k. Beta is a_i，mIs a binary variable when beta_i，mWhen 1, V2V Rx user i shares the same subchannel with V2I user m; when beta is_i，mWhen 0, V2V Rx user i and V2I user m occupy different sub-channels. Tau is_i，kIs a binary variable whenTime, τ_i，kAt this time, V2V Rx user i may receive the broadcast information of V2V Tx user j; otherwise whenTo, tau_i，kAt this time, V2V Rx user i is not in the coverage of V2V Tx user j, and cannot receive the broadcast information of V2V Tx user j. Rho_i，kThe binary variable is specifically defined as follows:

step 3.3.2, considering the bearing capacity of the V2I user to the interference of the V2V Tx user under the co-channel, considering that for the V2I user, the signal-to-interference-and-noise ratio γ of the V2I user is guaranteed_mLower limit of (d), Pr { gamma is set_m≤γ₀}≤p₀(wherein Pr {. denotes a probability distribution function, γ)₀Minimum signal-to-interference-and-noise ratio, p, of acceptable V2I users₀For V2I user signal-to-interference-and-noise ratio less than gamma₀A probability threshold of affordability); wherein, the signal-to-interference-and-noise ratio gamma of the V2I user_mThe signal to interference plus noise ratio (SINR) is as follows:

whereinFor the transmit power of the base station BS for V2I user m,for the transmit power of V2V Tx user i,is the channel gain constant between base station BS and V2I user m,is the channel gain constant, μ, between V2VTx user i and V2I user m_i，mBeing a binary variable, μ when V2V Tx user i shares the same subchannel with V2I user_i，m1, otherwise μ_i，m＝0；σ²Is the noise power.

In a practical scenario, the impact of V2V Tx users on V2I users is mainly the impact of the transmit power of the V2V Tx users who are closest to the V2I user and share the same sub-channel with the V2I user. In the case where NOMA is applied to V2V Tx, the channel difference between V2V Tx and V2V Rx is large, and the transmission power attenuates fast with distance, the distribution between vehicles satisfies the poisson distribution, and thus the above-described γ can be matched_mThe signal to interference plus noise ratio formula is approximately operated, namely:

wherein the corner mark i^*Indicates the identity of the V2V Tx user closest to V2I user m. Considering the SINR QoS condition of V2I user m, i.e. Pr { gamma }_m≤γ₀}≤p₀. It can be transformed as follows:

note the bookWherein alpha is_Bm、Is a large scale fading coefficient, g_Bm、Is a small scale fading coefficient. Then gamma can be further converted_m≤γ₀(wherein, γ)₀The minimum signal to interference plus noise ratio that can be tolerated for the V2I user) yields:

therefore, it isSo Pr { gamma_m≤γ₀It can be expressed as a double integral:

thus, Pr { γ }_m≤γ₀}≤p₀Expressed as:

at this time, the V2V Tx user i closest to the V2I user m may be selected^*The transmission power is converted into:

step 3.3.3, defining slot variable slot, making slot equal to 1, and generating N_SVector of dimension P x 1And calculating a corresponding proportional fair function (PF) function value, and taking the PF function value as a fitness function value of the PSO algorithm, thereby obtaining an optimal PF function value of the first time slot. The method comprises the following specific steps:

step 3.3.3.1, V2V Tx user i closest to V2I user m and sharing the same sub-channel in step 3.3.2^*Initializing the transmission power of all V2V Tx users under the constraint of transmission power, namely generating Ns power value vectors with dimension P multiplied by 1The generation method comprises the following steps: for V2V Rx users receiving V2V Tx user broadcast information, the PSO algorithm calculation amount is shared, and power allocation and data reception conditions of different V2V Rx users are exchanged through V2V Tx users in a cross region.

At step 3.3.3.2, it is set that the data transmission phase in an SPS period has N_SPSDefining iterative variable itera in PSO algorithm, and calculating each V2 when itera is equal to 1V Rx user transmission rate, all transmission rate values are recorded in RateRxTx with dimension Q x P_Q×PIn the matrix, RateRxTx (i, j) represents the transmission rate at which V2V Rx user i receives V2V Tx user j broadcast information:

wherein E [.]Is a desired operator;indicating that NOMA is applied to the V2V user scenario, the signal to interference plus noise ratio of the broadcast information received by V2V Rx user i to V2V Tx user j is given in step 3.3.1; the transmission rate of all V2V Tx broadcast information that V2V Rx user i can receive is accumulated, namely RateRxTxTx_Q×PSumming the rows of the matrix, and storing the total transmission rate of V2V Rx in the dimension of Q × N_SPSIn matrix ofIn (1), initializeIs an all-zero matrix, i.e.:

where P is the total number of V2V Tx users within the coverage of the base station BS.

Step 3.3.3.3, calculating a fitness function fitness value at the time of the slot at present, wherein the fitness function fitness is defined as follows:

when slot is 1, the initial value of the weight vector with dimension 1 × Q is weight (1, 1., 1)

fitness＝weight·RateRxSlot(：，slot)

＝(1，1，...，1)(RateRxSlot(1，slot)，...，RateRxSlot(Q，slot))^T；

When the slot is more than 1 and less than or equal to N_sPsIn time, the ith component of weight is noted as weight_iWherein i is more than or equal to 1 and less than or equal to Q, and the calculation method comprises the following steps:

fitness＝weight·RateRxSlot(：，slot)

＝(weight₁，...，weight_Q)(RateRxSlot(1，slot)，...，RateRxSlot(Q，slot))^T

step 3.3.3.4, for N_SPower value vector with dimension P x 1Respectively calculate corresponding N_SA fitness value; corresponding power value vector of maximum fixness valueIs marked asCorresponding the current maximum fitness value of each power vector to a power value vectorIs marked asGenerate eachUpdate gradient vector ofWherein:

when itera is equal to 1,

v_trand (1 × P); (wherein, v_tFor the t power value vectorCorresponding update gradient vector)

When itera > 1, the first time,

(wherein,for the t power value vector after the last PSO iteration update)

Wherein rand generates random numbers between (0, 1), c₁For updating the speed coefficient of each power vector itself, c₂Updating the speed coefficient for the global of each power vector;

step 3.3.3.5 Using the updated gradient vector of step 3.3.3.4To N_SPower value vector with dimension P x 1The updating is carried out, and the updating is carried out,

step 3.3.3.6, repeat 3.3.3.2 to 3.3.3.5 until itera is maxgen, where maxgen is the maximum iteration number, and finally obtain the optimal fitness value of maxgen iteration and the corresponding optimal power value vector

Step 3-3-4, make slot ═ slot +1, repeat step 3.3.3, until slot > N_SPS(wherein, N is_SPSThe number of slots that are the data transmission phase of the SPS). And finally, the power distribution of each time slot and the corresponding optimal PF function value are given.

The application characteristics of NOMA in a V2V broadcast communication downlink are comprehensively used, the problem that vehicle user CSI is difficult to obtain to solve resource allocation in a vehicle networking fast time-varying channel scene of a base station BS is solved, the fairness of a V2V Rx user is ensured, an algorithm can effectively converge to an optimal solution, the frequency spectrum utilization rate is effectively improved, and the network interface scale is enlarged.

It is to be understood that the present invention has been described with reference to certain embodiments, and that various changes in the features and embodiments, or equivalent substitutions may be made therein by those skilled in the art without departing from the spirit and scope of the invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from the essential scope thereof. Therefore, it is intended that the invention not be limited to the particular embodiment disclosed, but that the invention will include all embodiments falling within the scope of the appended claims.