Method for improving data transmission reliability under condition of ensuring fairness of Internet of Vehicles users

Abstract

The invention discloses a method for improving data transmission reliability under the condition of ensuring fairness of Internet of Vehicles users, which comprises the following steps of: constructing a network slice grouping rule meeting QoS (Quality of Service) requirements and load balance of Internet of Vehicles CU (Control Unit) users, and establishing an Internet of Vehicles V2V (Vehicle to Vehicle) user NOMA (Non-Orthogonal Multiple Access) clustering optimization target with constraint conditions; and analyzing the CU user network slice grouping rule, establishing a low-complexity CU user network slice grouping scheme, carrying out condition integration on the V2V user NOMA clustering optimization target with the constraint condition, designing a V2VNOMA clustering optimization method based on improved PSO as an updated optimization target according to the condition integration, and carrying out CU user network slice grouping by introducing a probability vector. Discrete mapping is carried out on the continuous variables, and finally the optimal V2VNOMA clustering scheme is obtained. According to the method, the QoS requirements of the CU users are ensured, the slicing and grouping fairness of the CU users is effectively improved, the optimal V2V user clustering fairness is effectively ensured, data collision under the NOMA protocol is avoided, and the data transmission reliability of the Internet of Vehicles is improved.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communication technology, and in particular relates to a method for improving data transmission reliability while ensuring fairness among users of an Internet of Vehicles (IoV). Background Art

[0002] Vehicle-to-infrastructure (V2I) communication links in the IoV communication system can support high-throughput communication services, including multimedia data traffic and internet data exchange. Vehicle-to-vehicle (V2V) communication links, while reusing V2I link spectrum resources, can support high-reliability, low-latency communication services, including forward collision warning and emergency braking reminders. Spectrum resource sharing between V2V and V2I links improves IoV spectrum resource utilization. With the exponential growth of vehicle users, the air interface access scale of IoV communication systems is facing challenges. Non-orthogonal multiple access (NOMA) technology, by differentiating users based on power or code domains, further improves spectrum utilization and IoV system throughput while expanding the air interface scale of IoV communication systems. To increase the flexibility of network resource allocation and effectively implement on-demand allocation, network slicing (NS), supported by software-defined networking (SDN) and network function virtualization (NFV), can provide customized spectrum resource allocation based on different service types.

[0003] However, the combination of NOMA and NS requires considering the impact of different service types' Quality of Service (QoS) on resource allocation to group network slice users and consider NOMA user clustering strategies. This has a significant impact on the throughput, communication reliability, and transmission latency of vehicle users in the IoV communication system. Therefore, how to formulate appropriate network slice grouping strategies and appropriate NOMA clustering solutions is an urgent problem to be solved in IoV communication systems. Summary of the Invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles, while effectively avoiding overload of each network slice while considering meeting the QoS of CU users.

[0005] Technical solution: The method of improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles described in the present invention specifically includes the following steps:

[0006] (1) When the base station knows the QoS requirements of each CU user in the Internet of Vehicles and the QoS requirements of each network slice within its coverage area, the BS sets rules for grouping network slices of CU users based on the network slice QoS requirements and the load of each network slice, and provides a CU user network slice grouping decision function. By calculating the grouping decision function value for the network slices that meet the CU user QoS requirements, the CU user slice grouping is completed;

[0007] (2) When the BS knows the geographical location of the V2V transmitter users within its coverage area, a V2V NOMA clustering objective function is constructed for the NOMA clustering of the V2V transmitter users;

[0008] (3) By introducing probability vectors, continuous variables are discretized and mapped, and an improved PSO algorithm is designed by introducing interference mechanism to achieve NOMA clustering of V2V transmitting users.

[0009] Furthermore, the step (1) includes the following steps:

[0010] (11) There are M CU users within the coverage of the BS, and the QoS requirement of the mth CU user is Where m is the CU user count variable; the number of network slices within the BS coverage area is S, and the QoS requirement of the sth network slice is Indicates that s is a counting variable identified by the network slice. To ensure that the QoS requirements of each CU user are met, it is necessary to ensure that if the mth CU user is grouped into the sth network slice, then The number of CU users per network slice load is U s , the network slice grouping decision function WCF is designed as:

[0011]

[0012] Among them, c1 is and The weight factor of the proximity degree, c2 is the weight factor of the network slice load, c1, c2 satisfy c1>0, c2>0, and c1+c2=1;

[0013] (12) For each CU user within the BS coverage area, let m' be the CU user iteration count variable, and let m' = 1; introduce the temporary network slice identifier set NSI, and traverse and compare and If the size Then the identifier of the network slice s is added to the temporary network slice identifier set NSI, and finally the temporary network slice identifier set NSI contains identifiers of S' network slices;

[0014] (13) The i-th element in the temporary network slice identifier set NSI is NSI(i), where i is a temporary network slice identifier set element counting variable; for each network slice identifier NSI(i) in the temporary network slice identifier set NSI, calculate S' network slice grouping decision function WCF values, namely WCF(1,NSI(1)), WCF(1,NSI(2)), ..., WCF(1,NSI(S'));

[0015] (14) For the S' network slice grouping decision function WCF values, select the network slice identifier corresponding to the smallest network slice grouping decision function WCF value, and record the network slice identifier as s'*; merge the m'th user into the s'* network slice;

[0016] (15) Update the CU user iteration count variable m'=m'+1, and repeat steps (12) to (14) until all M CU users are merged into the network slice.

[0017] Furthermore, the network slice identifiers are allocated in ascending order based on the QoS requirements of each network slice, i.e.

[0018] Furthermore, the V2V NOMA clustering objective function in step (2) includes maximizing the minimum V2V transmitter spacing within each V2V NOMA cluster, minimizing the variance of the minimum V2V transmitter spacing within each V2V NOMA cluster, minimizing the difference in the number of V2V users within each V2V NOMA cluster, and ensuring that the minimum V2V transmitter spacing within each V2V NOMA cluster is greater than a threshold.

[0019] Furthermore, the implementation process of step (2) is as follows:

[0020] The number of V2V transmitter users within the BS coverage is PV, and the number of V2V NOMA clusters constructed by the BS is K. To ensure the successful NOMA decoding of V2V receiver users while taking into account the differences in the minimum distances between V2V transmitters in each V2V NOMA cluster and the fairness of the user scale among each V2V NOMA cluster, the V2V NOMA clustering objective function is expressed as:

[0021]

[0022] Among them, F is the name of the V2VNOMA clustering objective function; V is the variable that stores the V2VNOMA clustering result; V k is the V2V transmitter user identifier saved by the kth V2V NOMA cluster; |V k | is the number of V2V transmitting users in the k-th V2V NOMA cluster; d is a K-dimensional column vector, i.e., d = (d1 d2 … dK ) T , (·) T is the transpose symbol, where each element d k is the minimum distance between V2V transmitting users in the kth V2VNOMA cluster, 1≤k≤K; d max is the maximum distance between V2V transmitters within the BS coverage area; is the rounding function; φ(·) is the indicator function, Var(·) is the variance operation; ω1, ω2, and ω3 are all greater than 0 and are weight factors, satisfying: ω1+ω2+ω3=1.

[0023] Furthermore, the implementation process of step (3) is as follows:

[0024] (31) The number of particles in the improved PSO algorithm is N sw ; Define the maximum number of particle update iterations T sw ; Particle vector of the pth V2V transmitter user in the i-th particle is a K-dimensional column vector, i is the particle counting variable, p is the V2V transmitter user counting variable, (1≤i≤N sw ,1≤p≤P), Introducing Probability variables for discretization mapping is a K-dimensional column vector, Individual particle learning rate L1; global particle learning rate L2; r1 and r2 are independent random numbers that are uniformly distributed in the interval [0,1]; t p is the iterative interference constant; is the K-dimensional velocity vector of the particle vector of the p-th V2V transmitter user in the i-th particle; introduce the particle update iteration count variable t sw , and initialize t sw =0; introduce n p As the number of disturbed particles; introduce the perturbation constant ε;

[0025] (32) Initialize the K-dimensional

[0026] (33) Yes Each element in is probability mapped, and the specific mapping method is:

[0027]

[0028] Where k' is the V2V NOMA cluster count variable independent of the V2V NOMA cluster count variable k; the K-dimensional probability variable of the particle of the p-th V2V transmitting user of the i-th particle is turn up The element number corresponding to the largest element in , that is k * for The element number corresponding to the largest element in the ith particle is the pth V2V transmitter user merged into the kth * V2V NOMA clusters; all P V2V transmitter users in the i-th particle perform the same operation until all P V2V transmitter users are merged into their respective V2V NOMA clusters, and the V2V NOMA clustering result of the i-th particle is saved in the variable V i ; for all N sw The same operation is performed on each particle;

[0029] (34) Calculate the V2V NOMA clustering objective function value for each particle in (33), that is, F(V i ), 1≤i≤N sw , save the existing individual best V2V NOMA cluster The corresponding existing individual best particle And its corresponding V2V NOMA clustering objective function value Then find The maximum V2V NOMA clustering objective function value in It is the existing global optimal 2V NOMA clustering objective function value, and saves its corresponding global individual best particle and its corresponding existing global optimal V2V NOMA clustering

[0030] (35) Particle update iteration count variable t sw , t sw =t sw +1; for t sw Update the number of iterations to determine the K-dimensional velocity vector of the particle vector of the p-th V2V transmitter user in the i-th particle The determination method is:

[0031]

[0032] Update N sw Particles are updated as follows:

[0033]

[0034] Then update the K-dimensional probability variable of each particle of each V2V transmitting user Update method and update the V2V NOMA clustering results of each particle to save the variable V i ;

[0035] (36) Save the V2V NOMA clustering results of each particle updated in (35) in variable V i Calculate its V2VNOMA clustering objective function value F(V i ); for all V2V NOMA clustering objective function values ​​F(V i ),if Then update the existing individual optimal V2V NOMA clustering objective function value Save the existing individual best V2V NOMA cluster The corresponding existing individual best particle For all V2V NOMA clustering objective function values ​​F(V i ),if Then update the existing global optimal V2V NOMA clustering objective function value Update its corresponding existing global individual best particle and its corresponding existing global optimal V2V NOMA clustering

[0036] (37) Considering the current t sw Can it be iteratively disturbed by the constant t p If it is not divisible, repeat steps (35) to (37). If the current t sw Can be iteratively disturbed by constant t p Divisible by n, then randomly select n p indivual right Add the perturbation constant ε to each element in , and then repeat steps (35) to (37) until t sw Reach T sw ; Obtain the optimal V2V NOMA clustering

[0037] The present invention provides a storage medium having a computer program stored thereon. When the computer program is executed by at least one processor, the steps of the method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles are implemented as described above.

[0038] An electronic device according to the present invention includes a memory and a processor, wherein:

[0039] a memory for storing computer programs capable of running on the processor;

[0040] The processor is used to execute the steps of the method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles when running the computer program.

[0041] Beneficial effects: Compared with the prior art, the beneficial effects of the present invention are as follows: compared with the existing network slice grouping method that only groups according to the CU user transmission rate QoS indicator and the traditional greedy NOMA clustering, the present invention effectively guarantees the CU user QoS while ensuring the load balancing of each network slice; in addition, by adding an interference mechanism to the traditional PSO algorithm, it avoids falling into the local optimal solution and effectively realizes the global optimality of V2V NOMA clustering under the proposed multiple indicators; by combining NS and NOMA, the present invention groups CU users into network slices while considering the transmission rate QoS of different network slices, and effectively avoids the overload of each network slice while considering satisfying the CU user QoS. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0043] The present invention is further described in detail below with reference to the accompanying drawings.

[0044] like Figure 1 As shown, the method of improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles according to the present invention includes the following steps:

[0045] Step 1: When the BS knows the QoS requirements of each CU user in the Internet of Vehicles and each network slice within its coverage area, the BS sets rules for grouping network slices of CU users based on the network slice QoS requirements and the load of each network slice, and provides a CU user network slice grouping decision function. By calculating the grouping decision function value for the network slices that meet the CU user QoS requirements, the BS completes the CU user slice grouping. Specifically, the following steps are included:

[0046] Step 1.1: Define the number of CU users within the BS coverage area as M, and the QoS requirement of the mth CU user is Where m is the CU user count variable. The number of network slices within the BS coverage area is S, and the QoS requirement of the sth network slice is Represented by , where s is the network slice identification counting variable. Here, the identification order of the network slices is given based on the increasing order of the QoS requirements of each network slice, that is, In order to ensure that the QoS requirements of each CU user can be met, it is necessary to ensure that if the mth CU user is grouped into the sth network slice, then At the same time, the number of CU users of each network slice load is defined as U s , the network slice grouping decision function WCF is designed as:

[0047]

[0048] Among them, c1 and c2 are weight factors of the network slice grouping decision function WCF, and c1 is and The weight factor of the proximity degree, c2 is the weight factor of the network slice load condition, c1, c2 satisfy c1>0, c2>0, and c1+c2=1.

[0049] Step 1.2: For each CU user within the BS coverage area, denote m' as the CU user iteration count variable, and set m' = 1. Introduce the temporary network slice identifier set NSI and traverse and compare and If the size Then the identifier of the network slice s is added to the temporary network slice identifier set NSI. It is defined that there are S' network slice identifiers in the final temporary network slice identifier set NSI.

[0050] Step 1.3: For the temporary network slice identifier set NSI obtained in step 1.2, define the i-th element in the temporary network slice identifier set NSI as NSI(i), where i is the temporary network slice identifier set element counting variable. For each network slice identifier NSI(i) in the temporary network slice identifier set NSI, calculate S' network slice grouping decision function WCF values, namely WCF(1,NSI(1)), WCF(1,NSI(2)), ..., WCF(1,NSI(S')).

[0051] Step 1.4: For the S' network slice grouping decision function (WCF) values ​​in step 1.3, select the network slice identifier corresponding to the smallest WCF value and denote this network slice identifier as s'*. Merge the m'th user into the s'* network slice.

[0052] Step 1.5: Update the CU user iteration count variable m'=m'+1, and repeat steps 1.2 to 1.4 until all M CU users are merged into the network slice.

[0053] Step 2: When the BS knows the geographic location of the V2V transmitter users within its coverage area, in order to avoid the risk of NOMA decoding failure of the V2V receiving users falling in the overlapping area of ​​the communication range of the same-channel V2V transmitter users, a V2V NOMA clustering objective function is constructed for the NOMA clustering of the V2V transmitter users, which includes multiple objectives such as maximizing the minimum V2V transmitter distance within each V2V NOMA cluster, minimizing the variance of the minimum V2V transmitter distance within each V2V NOMA cluster, minimizing the difference in the number of V2V transmitter users within each V2V NOMA cluster, and ensuring that the minimum V2V transmitter distance within each V2V NOMA cluster is greater than a threshold.

[0054] Define the number of V2V transmitter users within the BS coverage area as P, and define the number of V2V NOMA clusters that the BS intends to construct as K. To ensure successful NOMA decoding of V2V receiver users while taking into account the differences in the minimum distances between V2V transmitters within each V2V NOMA cluster and the fairness of the user scale between each V2V NOMA cluster, the V2V NOMA clustering objective function is expressed as:

[0055]

[0056] Among them, F is the name of the V2VNOMA clustering objective function; V is the variable that stores the V2VNOMA clustering result; V k is the V2V transmitter user identifier saved by the kth V2V NOMA cluster; |V k | is the number of V2V transmitting users in the k-th V2V NOMA cluster; d is a K-dimensional column vector, i.e., d = (d1 d2 … d K ) T , (·) T is the transpose symbol, where each element d k (1≤k≤K) is the minimum distance between V2V transmitting users in the kth V2V NOMA cluster; d max is the maximum distance between V2V transmitters within the BS coverage area; is the rounding function; φ(·) is the indicator function, Var(·) is the variance operation; ω1, ω2, and ω3 are weight factors, satisfying: ω1+ω2+ω3=1(ω1>0,ω2>0,ω3>0).

[0057] Step 3: By introducing the probability vector, the continuous variable is discretized and mapped, and the improved PSO algorithm is designed by introducing the interference mechanism to realize the NOMA clustering of the V2V transmitting user.

[0058] Step 3.1, define the number of particles of the improved PSO algorithm as N sw; Define the maximum number of particle update iterations T sw ; Define the particle vector of the pth V2V transmitter user in the i-th particle is a K-dimensional column vector, where i is the particle count variable, p is the V2V transmitter user count variable, Introducing Probability variables for discretization mapping is a K-dimensional column vector, Define the individual particle learning rate L1; the global particle learning rate L2; r1 and r2 are independent random numbers that are uniformly distributed in the interval [0,1]; t p is the iterative interference constant; is the K-dimensional velocity vector of the particle vector of the p-th V2V transmitter user in the i-th particle; introduce the particle update iteration count variable t sw , and initialize t sw =0; introduce n p As the number of disturbed particles; introduce the perturbation constant ε.

[0059] Step 3.2: Initialize the K-dimensional space by using random numbers uniformly distributed in the interval [0,1].

[0060] Step 3.3, Each element in is probability mapped, and the specific mapping method is:

[0061]

[0062] Where k' is the V2V NOMA cluster count variable independent of the V2V NOMA cluster count variable k; the K-dimensional probability variable of the particle of the p-th V2V transmitting user of the i-th particle is turn up The element number corresponding to the largest element in , that is where k * for The element number corresponding to the largest element in the ith particle is the pth V2V transmitter user merged into the kth * V2VNOMA clusters; all P V2V transmitter users in the i-th particle perform the same operation until all P V2V transmitter users are merged into their respective V2V NOMA clusters, and the V2VNOMA clustering result of the i-th particle is saved in the variable V i For all N sw Perform the same operation on each particle.

[0063] Step 3.4: Calculate the V2V NOMA clustering objective function value for each particle in step 3.3, that is, F(V i )(1≤i≤N sw ), save the existing individual best V2V NOMA cluster The corresponding existing individual best particle And its corresponding V2V NOMA clustering objective function value Then find The maximum V2VNOMA clustering objective function value in It is the existing global optimal 2V NOMA clustering objective function value, and saves its corresponding global individual best particle and its corresponding existing global optimal V2V NOMA clustering

[0064] Step 3.5: Particle update iteration count variable t sw , t sw =t sw +1; for t sw Update the number of iterations to determine the K-dimensional velocity vector of the particle vector of the p-th V2V transmitter user in the i-th particle The determination method is:

[0065]

[0066] Update N sw Particles are updated as follows:

[0067]

[0068] Then update the K-dimensional probability variable of each particle of each V2V transmitting user Its update method and update of each particle's V2V NOMA clustering result save variable V i Repeat step 3.3 as described above.

[0069] Step 3.6: Save the V2V NOMA clustering results of each particle updated in step 3.5 into variable V i Calculate its V2V NOMA clustering objective function value F(V i )(1≤i≤N sw ); for all V2V NOMA clustering objective function values ​​F(V i )(1≤i≤N sw ),if Then update the existing individual optimal V2V NOMA clustering objective function value Save the existing individual best V2V NOMA cluster The corresponding existing individual best particle For all V2V NOMA clustering objective function values ​​F(V i )(1≤i≤N sw ),if Then update the existing global optimal V2V NOMA clustering objective function value Update its corresponding existing global individual best particle and its corresponding existing global optimal V2V NOMA clustering

[0070] Step 3.7: Consider the current t sw Can it be iteratively disturbed by the constant t p If it is not divisible, repeat steps 3.5 to 3.7. If the current t sw Can be iteratively disturbed by constant t p Divisible by n, then randomly select n p indivual right Add the perturbation constant ε to each element in , and then repeat steps 3.5 to 3.7 until t sw Reach T sw Finally, the best V2VNOMA clustering is obtained.

[0071] The present invention also provides a storage medium having a computer program stored thereon. When the computer program is executed by at least one processor, the steps of the method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles are implemented as described above.

[0072] The present invention also provides an electronic device comprising a memory and a processor, wherein the memory is used to store a computer program that can be run on the processor; and the processor is used to execute the steps of the method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles when running the computer program.

[0073] The above content is an example of the method of the present invention. Any modifications or similar substitutions made by technicians in this technical field to the described specific embodiments shall fall within the scope of protection of the present invention as long as they do not deviate from the method of the present invention or exceed the scope defined by the claims.

Claims

1. A method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles, characterized by: The following steps are involved: (1) When the base station knows the QoS requirements of each CU user in the Internet of Vehicles and the QoS requirements of each network slice within its coverage area, the BS sets rules for grouping network slices of CU users based on the network slice QoS requirements and the load of each network slice, and provides a CU user network slice grouping decision function. By calculating the grouping decision function value for the network slices that meet the CU user QoS requirements, the CU user slice grouping is completed; (2) When the BS knows the geographical location of the V2V transmitter users within its coverage area, a V2V NOMA clustering objective function is constructed for the NOMA clustering of the V2V transmitter users; (3) By introducing probability vectors, continuous variables are discretized and mapped, and an improved PSO algorithm is designed by introducing interference mechanism to achieve NOMA clustering of V2V transmitting users.

2. The method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles according to claim 1, characterized in that: The step (1) comprises the following steps: (11) There are M CU users within the coverage of the BS, and the QoS requirement of the mth CU user is Where m is the CU user count variable; the number of network slices within the BS coverage area is S, and the QoS requirement of the sth network slice is Indicates that s is a counting variable identified by the network slice. To ensure that the QoS requirements of each CU user are met, it is necessary to ensure that if the mth CU user is grouped into the sth network slice, then The number of CU users per network slice load is U s , the network slice grouping decision function WCF is designed as: Among them, c1 is and The weight factor of the proximity degree, c2 is the weight factor of the network slice load, c1, c2 satisfy c1>0, c2>0, and c1+c2=1; (12) For each CU user within the BS coverage area, let m' be the CU user iteration count variable, and let m' = 1; introduce the temporary network slice identifier set NSI, and traverse and compare and If the size Then the identifier of the network slice s is added to the temporary network slice identifier set NSI, and finally the temporary network slice identifier set NSI contains identifiers of S' network slices; (13) The i-th element in the temporary network slice identifier set NSI is NSI(i), where i is a temporary network slice identifier set element counting variable; for each network slice identifier NSI(i) in the temporary network slice identifier set NSI, calculate S' network slice grouping decision function WCF values, namely WCF(1,NSI(1)), WCF(1,NSI(2)), ..., WCF(1,NSI(S')); (14) For the S' network slice grouping decision function WCF values, select the network slice identifier corresponding to the smallest network slice grouping decision function WCF value, and record the network slice identifier as s'*; merge the m'th user into the s'* network slice; (15) Update the CU user iteration count variable m'=m'+1, and repeat steps (12) to (14) until all M CU users are merged into the network slice.

3. The method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles according to claim 2, characterized in that: The network slice identifiers are assigned based on the QoS requirements of each network slice in ascending order, i.e.

4. The method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles according to claim 1, characterized in that: The V2V NOMA clustering objective function in step (2) includes maximizing the minimum V2V transmitter spacing within each V2V NOMA cluster, minimizing the variance of the minimum V2V transmitter spacing within each V2V NOMA cluster, minimizing the difference in the number of V2V users within each V2V NOMA cluster, and ensuring that the minimum V2V transmitter spacing within each V2V NOMA cluster is greater than a threshold.

5. The method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles according to claim 1, characterized in that: The implementation process of step (2) is as follows: The number of V2V transmitter users within the BS coverage is PV, and the number of V2V NOMA clusters constructed by the BS is K. To ensure the successful NOMA decoding of V2V receiver users while taking into account the differences in the minimum distances between V2V transmitters in each V2V NOMA cluster and the fairness of the user scale among each V2V NOMA cluster, the V2V NOMA clustering objective function is expressed as: Among them, F is the name of the V2V NOMA clustering objective function; V is the variable that stores the V2V NOMA clustering result; V k is the V2V transmitter user identifier saved by the kth V2VNOMA cluster; |V k | is the number of V2V transmitting users in the kth V2V NOMA cluster; d is a K-dimensional column vector, i.e., d = (d1 d2…d K ) T , (·) T is the transpose symbol, where each element d k is the minimum distance between V2V transmitting users in the kth V2VNOMA cluster, 1≤k≤K; d max is the maximum distance between V2V transmitters within the BS coverage area; is the rounding function; φ(·) is the indicator function, Var(·) is the variance operation; ω1, ω2, and ω3 are all greater than 0 and are weight factors, satisfying: ω1+ω2+ω3=1.

6. The method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles according to claim 1, characterized in that: The implementation process of step (3) is as follows: (31) The number of particles in the improved PSO algorithm is N sw ; Define the maximum number of particle update iterations T sw ; Particle vector of the pth V2V transmitter user in the i-th particle is a K-dimensional column vector, i is the particle counting variable, p is the V2V transmitter user counting variable, (1≤i≤N sw ,1≤p≤P), Introducing Probability variables for discretization mapping is a K-dimensional column vector, Individual particle learning rate L1; global particle learning rate L2; r1 and r2 are independent random numbers that are uniformly distributed in the interval [0,1]; t p is the iterative interference constant; is the K-dimensional velocity vector of the particle vector of the p-th V2V transmitter user in the i-th particle; introduce the particle update iteration count variable t sw , and initialize t sw =0; introduce n p As the number of disturbed particles; introduce the perturbation constant ε; (32) Initialize the K-dimensional (33) Yes Each element in is probability mapped, and the specific mapping method is: Where k' is the V2V NOMA cluster count variable independent of the V2V NOMA cluster count variable k; the K-dimensional probability variable of the particle of the p-th V2V transmitting user of the i-th particle is turn up The element number corresponding to the largest element in , that is k * for The element number corresponding to the largest element in the ith particle is the pth V2V transmitter user merged into the kth * V2V NOMA clusters; all P V2V transmitter users in the i-th particle perform the same operation until all P V2V transmitter users are merged into their respective V2V NOMA clusters, and the V2V NOMA clustering result of the i-th particle is saved in the variable V i ; for all N sw The same operation is performed on each particle; (34) Calculate the V2V NOMA clustering objective function value for each particle in (33), that is, F(V i ), 1≤i≤N sw , save the existing individual best V2V NOMA cluster The corresponding existing individual best particle And its corresponding V2V NOMA clustering objective function value Then find The maximum V2V NOMA clustering objective function value in It is the existing global optimal 2V NOMA clustering objective function value, and saves its corresponding global individual best particle and its corresponding existing global optimal V2V NOMA clustering (35) Particle update iteration count variable t sw , t sw =t sw +1; for t sw Update the number of iterations to determine the K-dimensional velocity vector of the particle vector of the p-th V2V transmitter user in the i-th particle The determination method is: Update N sw Particles are updated as follows: Then update the K-dimensional probability variable of each particle of each V2V transmitting user Update method and update the V2V NOMA clustering results of each particle to save the variable V i ; (36) Save the V2V NOMA clustering results of each particle updated in (35) in variable V i Calculate its V2V NOMA clustering objective function value F(V i ); for all V2V NOMA clustering objective function values ​​F(V i ),if Then update the existing individual optimal V2V NOMA clustering objective function value Save the existing individual best V2V NOMA cluster The corresponding existing individual best particle For all V2V NOMA clustering objective function values ​​F(V i ),if Then update the existing global optimal V2V NOMA clustering objective function value Update its corresponding existing global individual best particle and its corresponding existing global optimal V2V NOMA clustering (37) Considering the current t sw Can it be iteratively disturbed by the constant t p If it is not divisible, repeat steps (35) to (37). If the current t sw Can be iteratively disturbed by constant t p Divisible by n, then randomly select n p indivual right Add the perturbation constant ε to each element in , and then repeat steps (35) to (37) until t sw Reach T sw ; Obtain the optimal V2V NOMA clustering 7. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by at least one processor, implements the steps of the method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles as described in any one of claims 1 to 6.

8. An electronic device, characterized in that: comprising a memory and a processor, wherein: a memory for storing computer programs capable of running on the processor; A processor is configured to execute the steps of the method for improving data transmission reliability while ensuring fairness among users of the Internet of Vehicles as described in any one of claims 1 to 6 when running the computer program.

Images