Efficient dynamic roadside unit deployment method and system capable of detecting Sybil attack

Abstract

The invention discloses an efficient dynamic roadside unit deployment method and system capable of detecting a Sybil attack. The method comprises the steps of collecting geographic information, evaluating roadside unit service coverage, an overlapping range and an accident information initial diffusion speed of a deployment area, and establishing an RSU deployment constraint capable of detecting the Sybil attack; constructing an RSU deployment scheme based on particle swarm optimization and an improved memetic algorithm; obtaining transportation and reconstruction RSU overhead, and obtaining a minimum RSU adjustment cost based on a bipartite graph matching algorithm; and collecting the working load of each RSU, and constructing a heuristic RSU minimum cost adjustment scheme based on a Pareto optimal principle. Under the constraint that the Sybil attack can be detected, by constructing a heuristic RSU efficient deployment and low-cost adjustment scheme, the service effective coverage range and the accident information initial diffusion speed in RSU deployment are improved, load balancing between RSUs is achieved with the minimum cost, and the capability of detecting the Sybil attack is achieved; the problems that a traditional RSU deployment scheme is low in comprehensive performance and cannot detect Sybil attacks are solved.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent transportation technology, and in particular relates to an efficient dynamic roadside unit deployment method and system capable of detecting witch attacks. Background Art

[0002] With the rapid development of intelligent vehicles and intelligent transportation systems (ITS), roadside units (RSUs) have become an indispensable component of intelligent traffic management, location privacy protection, and emergency rescue services. Reasonable deployment of RSUs can improve traffic efficiency and reduce the risk of traffic collisions and congestion. However, large-scale deployment of RSUs has higher costs, is impractical, and has low efficiency. Moreover, the existing RSU deployment schemes still have room for improvement in some key performance indicators. Therefore, in order to improve network connectivity, it is necessary to comprehensively consider urban characteristics such as traffic, geographical environment, and signal interference to optimally deploy RSUs within limited resource constraints.

[0003] In recent years, the rapid development of ITS has also brought many threats. As an important asset of the Internet of Vehicles (IoV), RSU is crucial in areas such as multi-agent joint identification of malicious attacks and federated learning task offloading. However, many existing RSU deployment works only focus on efficiency without considering some security threats that are prone to occur in the Internet of Vehicles, increasing the operational risks of ITS. It is worth noting that Sybil attack is one of the common risks in IoV, which can seriously damage the reliability and integrity of the network. By integrating Sybil attack detection into the RSU deployment scheme and considering the necessary cooperative distance constraints of the RSU location in actual scenarios, the RSU deployment scheme can be made secure.

[0004] The number and location of RSUs significantly affect the quality of service (QoS) of ITS. However, considering the actual development of cities and ITS, long-term fixed RSU locations will not only lose the cost-effectiveness of the original deployment, but also increase security risks due to the load imbalance of RSUs. In order to ensure the service quality of the Internet of Vehicles, it is necessary to adjust the deployment location of RSUs at the lowest cost according to the changes in traffic flow caused by urban development. Unfortunately, the existing RSU adaptation and adjustment work is insufficient. Most RSU adjustment works are optimized based on randomly generated RSU sites. It is necessary to make the RSU deployment and adjustment work more continuous and practical. In addition, the existing RSU adjustment work also ignores the distance requirements between IoV agents that need to meet the security collaboration in actual scenarios, and the repeated optimization targets that also appear in the RSU deployment scheme make the performance of the traditional RSU adjustment scheme not excellent. Therefore, in the planning process of smart cities, it is crucial to establish an RSU deployment and adjustment scheme with continuous connectivity, detectable witch attacks, and superior multi-attributes. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the present invention provides an efficient dynamic roadside unit deployment method and system that can detect witch attacks, including collecting geographic information, evaluating the roadside unit service coverage, overlapping range and initial diffusion speed of accident information in the deployment area, and establishing RSU deployment constraints that can detect witch attacks; constructing an RSU deployment plan based on particle swarm optimization and improved meme algorithm; obtaining transportation and reconstruction RSU overhead, and obtaining the minimum RSU adjustment cost based on the bipartite graph matching algorithm; collecting each RSU workload, and constructing a heuristic RSU minimum cost adjustment plan based on the Pareto optimal principle. Under the constraint of detectable witch attacks, the present invention improves the effective coverage of services and the initial diffusion speed of accident information in RSU deployment by constructing a heuristic RSU efficient deployment and low-cost adjustment plan, realizes load balancing between RSUs at the lowest cost, and has the ability to detect witch attacks, solving the problem of low comprehensive performance and inability to detect witch attacks in traditional RSU deployment plans.

[0006] The technical solution adopted by the present invention to solve the technical problem is as follows:

[0007] S1: Obtain the geographic information of the deployment area, initialize the deployment matrix of the deployment area, collect all possible deployment sites of RSUs in the deployment area; obtain the RSU transmission and influence range C R , C e , the vehicle-to-vehicle communication range C V and the number of available and additionally adjusted available RSUs N r 、N e ;

[0008] S2: Randomly generate RSU deployment plan S based on the geographic information of the deployment area and the information of each entity in the area o , we get solution S o RSU service coverage and overlapping range g cover , and calculate the initial diffusion speed S of accident information in combination with the vehicle flow in the area init ; Based on the signal strength, we get the Sybil attack detection deployment constraint C S ;

[0009] S3: Design the fitness value F to measure the performance of the RSU deployment solution. S Under the constraint, according to the fitness value F, the particle swarm-two-neighborhood search meme PSO-Meme joint heuristic algorithm is used to optimize the RSU deployment scheme to obtain a near-optimal RSU deployment scheme S c ;

[0010] S4: Get RSU reconstruction cost based on the market new and transportation adjustment cost adAccording to the deployment area geographic information and RSU adjustment location, the minimum RSU adjustment cost Cost is obtained based on the bipartite graph matching algorithm Hopcroft-Karp total ;

[0011] S5: Evaluate the workload of RSUs based on the traffic volume within the effective range of each RSU in the deployment area Design a fitness value F to measure the performance of the RSU adjustment scheme ad ; in C S Under the constraints, the RSU initial deployment plan is optimized and iterated using the heuristic RSU multi-objective adaptive adjustment algorithm HRMA3 based on the Pareto optimal principle to obtain the RSU minimum cost adjustment plan S ad , while ensuring the overall efficiency of RSU deployment, load balancing between RSUs is achieved.

[0012] Furthermore, the step S2 is specifically as follows:

[0013] S201: The dynamic RSU deployment system extracts the geographic information of the deployment area and the information of each entity in the area and converts it into a deployment matrix M c And randomly generate a deployment plan S in the deployment matrix o ;

[0014] S202: The dynamic RSU deployment system is based on the entity information and deployment plan S in the area o Get the RSU service coverage and overlapping range g of the solution cover ;

[0015] The effective signal coverage area of ​​an RSU is defined as a polygon P = {P1, P2, ..., P i}, each RSU R n Each has its effective signal coverage area P i ; Assuming that Nr RSUs are available, the overlapping area of ​​the RSU effective signal is the accumulation of the overlapping areas between multiple polygons, expressed as:

[0016]

[0017] Where O(a,b) represents the function for calculating the overlapping area of ​​two polygons a and b;

[0018] Therefore, g is determined by the coverage area and overlap area of ​​the RSU effective signal. cover It is expressed as:

[0019]

[0020] Among them A T It is the total area in the city that needs to be covered by effective signals.

[0021] S203: The dynamic RSU deployment system is based on the entity information, vehicle flow and deployment plan S in the area o Get the initial diffusion speed S of the accident information init ;

[0022] There is an accident site A in the RSU deployment area. Assuming that all V2V transmissions are successful, the calculation of the average ISSAI value is expressed as:

[0023]

[0024] Where S v is the speed of vehicle v, N is the number of all vehicles involved in the accident information dissemination; T v It represents the sum of the time spent by vehicle v to transmit accident information using carrier unit Ui, expressed as:

[0025]

[0026] S204: The dynamic RSU deployment system establishes a Sybil attack detection deployment constraint C based on the signal strength and the required RSU cooperation location distance in the actual scenario S .

[0027] The necessary RSU cooperation position distance for Sybil attack detection according to the actual scenario is expressed as:

[0028]

[0029] Among them C e Indicates the impact range of RSU.

[0030] Furthermore, the step S3 is specifically as follows:

[0031] S301: The dynamic RSU deployment system allocates RSU service coverage and overlap ranges according to the priority of the deployment target. cover and the initial diffusion speed of accident information S init Combined to get the fitness value F;

[0032]

[0033] Among them, θ1 and θ2 represent the normalized weights from 0 to 1, T init Represents the short message transmission time after the accident;

[0034] S302: In C S Under the constraint, the dynamic RSU deployment system uses the particle swarm optimization algorithm to optimize the r deployment solutions S with the largest response value F according to the response value F. r , and put it into the initial solution set;

[0035] S303: The dynamic RSU deployment system sorts all solutions in the initial solution set according to the fitness value F and selects e elite solutions to form an elite class E;

[0036] E=Sort(F(s1),F(s2),…,F(s e ))e <Num(S r )

[0037] S304: In C S Under the constraints, the elite solution set E is crossover and mutation operated to obtain the offspring solution set P;

[0038] S305: The dynamic RSU deployment system performs a neighborhood search on the offspring solution set obtained after each crossover and mutation operation and selects a solution with a large fitness value to update the offspring solution set P;

[0039] Define the single solution in the elite offspring population after crossover and mutation operations as S p , whose neighborhood is N(S p ), that is, from the current solution S p By exploring the solution set generated by small step size, if make

[0040]

[0041] Then the elite offspring solution S p Update to S′ p ;

[0042] S306: If the two-neighborhood search-meme algorithm has not reached the maximum number of iterations, go to S303, otherwise output a near-optimal RSU deployment solution S c .

[0043] Furthermore, the step S4 is specifically as follows:

[0044] S401: Dynamic RSU deployment system formulates RSU reconstruction cost based on each entity and market information in the region new ;

[0045] Compared with the deployment plan Sol before adjustment, if the RSU in the adjustment plan Sol′ increases, new RSU equipment will be purchased and deployed. The cost calculation formula for the newly deployed RSU is:

[0046] Cost new =(|Sol′|-|Sol|)·(∈+δ)

[0047] Where ρ is the purchase cost and δ is the construction cost of a new RSU;

[0048] S402: The dynamic RSU deployment system formulates the RSU transportation adjustment cost based on the entities and market information in the region. ad ;

[0049] The shipping adjustment for RSU includes the removal, installation and transportation costs, so the RSU shipping adjustment overhead is calculated as follows:

[0050]

[0051] Where |R ad | is the number of RSUs that need to be adjusted, is the transportation cost per kilometer of RSU, D ad is the Euclidean distance between the RSU sites that need to be adjusted;

[0052] S403: Dynamic RSU deployment system rebuilds the cost according to RSU new Cost adjustment with RSU transportation ad Formulate RSU and calculate total cost total ;

[0053] Cost total =Cost new +Cost ad .

[0054] S404: Dynamic RSU deployment system obtains the minimum total cost of RSU adjustment based on the bipartite graph matching algorithm Hopcroft-Karp total ;

[0055] The existing RSU positions and the newly adjusted RSU positions are considered as a bipartite graph set V r and V r ′, the weight E of the edge represents the cost of moving from the existing position to the new position, which is expressed as:

[0056] G(V r ,V r ′;e),e∈E(V r ,V r ′)={D(R1,R′1),D(R1,R′2),…,D(R n ,R′ m )},

[0057] Where D(R n ,R′ m ) represents R n ∈V r and R m ′∈V r The Euclidean distance between ′.

[0058] Furthermore, the step S5 is specifically as follows:

[0059] S501: The dynamic RSU deployment system evaluates the workload of the RSU based on the traffic flow within the effective range of each RSU in the deployment area

[0060] The vehicle selects the nearest RSU to interact with, based on the RSU R n Traffic flow in the coverage area The interaction strength is Define the maximum load threshold of RSU through investigation Each RSU The workload is represented as:

[0061]

[0062] S502: The dynamic RSU deployment system designs a fitness value F to measure the performance of the RSU adjustment scheme according to the RSU service quality and the load balancing index between the RSUs. ad ;

[0063] According to the signal coverage and overlapping range of deployed RSU g cover , load balancing between RSUs and the minimum cost of adjusting RSUs, and obtain the fitness value representing the performance of the RSU adjustment scheme:

[0064] F ad =θ1·g cover -θ3·Cost total -θ4·B R ,

[0065] Among them, θ1, θ3 and θ4 represent the normalized weights from 0 to 1, B R To evaluate the standard deviation statistic of load balance between RSUs, it is expressed as:

[0066]

[0067] in is the average workload of all RSUs in the deployment area;

[0068] S503: The dynamic RSU deployment system randomly generates an RSU adjustment plan and converts the RSU deployment plan S close to the optimal one into an RSU adjustment plan. c Add to the solution set O of the adjustment optimization solution;

[0069] S504: The solutions in the solution set O are evaluated according to the fitness value F ad To sort and filter, in C SUnder the constraints, crossover and mutation operations are performed on the selected solutions to obtain an offspring population pop equal to the population size of the optimization solution set O;

[0070] S505: The dynamic RSU deployment system merges the solution set O and the offspring population pop according to the fitness value F ad Perform fast non-dominated sort;

[0071] According to the fitness value F ad Perform a quick non-dominated sort on the deployment solution population to prepare for the subsequent introduction of deployment efficiency reference points for niche selection:

[0072] Sort non-domin (F ad (s1),F ad (s2),…,F ad (s l ))

[0073] S506: The dynamic RSU deployment system adds the deployment efficiency reference point to the sorted population and performs niche selection to select an elite population set e equal to the number of the next generation pop population. ad ;

[0074] S507: Dynamic RSU deployment system performs genetic operations to generate and elite solution population sets ad The offspring with equal population size pop′;

[0075] S508: The dynamic RSU deployment system performs a neighborhood search on the offspring solution set obtained after each crossover and mutation operation and selects a solution with a large fitness value to update the offspring solution set pop′;

[0076] Define the single solution in the offspring population after crossover and mutation operations as S l , whose neighborhood is N(S l ), that is, from the current solution S l By exploring the solution set generated by small step size, if make:

[0077]

[0078] Then the elite offspring solution S l Update to S′ l ;

[0079] S509: If the HRMA3 algorithm has not reached the maximum number of iterations, proceed to S504; otherwise, output the optimal RSU minimum cost adjustment plan.

[0080] An efficient dynamic roadside unit deployment system capable of detecting Sybil attacks, including a module for acquiring geographic information of the RSU deployment area and information of each entity, a module for multi-performance measurement of the RSU deployment scheme, a PSO-Meme joint heuristic RSU deployment module, and a minimum cost RSU deployment adjustment module;

[0081] The RSU deployment area geographic information and entity information acquisition module is used to obtain the RSU deployment area, including the obstacle location volume, possible RSU deployment sites near intersections and city landmark buildings, RSU entity impact and communication range, vehicle equipment information, RSU received signal strength and traffic flow data of each vehicle interaction; the traffic flow information corresponding to each historical timestamp collected by each RSU in the deployment area is cleaned and missing values ​​are filled, and each traffic flow information is normalized by wavelet transform to ensure that the data format is suitable for subsequent RSU load balancing evaluation; the basic performance attributes of each RSU and vehicle in the deployment area are collected, and after removing outliers, a weighted average is used to formulate an entity basic information database without loss of generality; the information of roads, buildings, intersections, etc. in the deployment area is projected onto a plane through a geographic information system and a traffic monitoring system to form an RSU deployment matrix, providing an accurate model basis for the subsequent scientific deployment adjustment plan of the RSU;

[0082] The RSU deployment scheme multi-performance measurement module is used to formulate Sybil attack detection constraints in RSU deployment and highly representative RSU deployment scheme measurement indicators based on the acquired RSU deployment area geographic information and entity information; through rule engine, statistical analysis, and mathematical quantification methods, measurement indicators are formulated for multiple performances of RSU, including Sybil attack detection security constraints formulated according to the necessary RSU cooperation location distance in actual scenarios, RSU signal coverage and overlap range indicators for measuring regional RSU service quality and deployment efficiency, accident information initial propagation speed indicators for measuring the system's ability to notify future vehicles after an accident occurs in the region, regional RSU standard deviation statistics for measuring load balancing between RSUs in the region, RSU minimum adjustment cost indicators for measuring the quality of RSU deployment and the quality of RSU adjustment strategies in the region, fitness values ​​for measuring the comprehensive performance of RSU deployment, and fitness values ​​for measuring the comprehensive performance of RSU adjustment schemes;

[0083] The PSO-Meme combined heuristic RSU deployment module is used to iteratively optimize the RSU deployment plan; first, the PSO's ability of fast convergence and global search is used to optimize the initial RSU deployment plan to produce a relatively good RSU deployment result, and then the relatively good RSU deployment result of PSO is iteratively optimized using a two-neighborhood search-meme algorithm with a large solution set exploration domain and a high solution set fitness value to produce the closest to the optimal RSU deployment result; the two-neighborhood search-meme algorithm performs a neighborhood search after the crossover and mutation steps on the basis of the meme algorithm to improve the fitness value of the overall solution set;

[0084] The minimum cost RSU deployment adjustment module is used to iteratively optimize the minimum cost RSU adjustment scheme based on RSU service quality and load balancing between RSUs; the calculation of the minimum cost of RSU adjustment adopts the Hopcroft-Karp algorithm in the bipartite graph to improve the efficiency of computing resources; the minimum cost RSU adjustment scheme uses fast non-dominated sorting to stratify the population, and then uses the RSU deployment efficiency reference point and niche selection based on the fitness value to perform elite screening on the population of each layer to generate a descendant population, and repeats multiple times until the maximum number of iterations is reached, and finally adjusts the RSU deployment scheme to obtain an RSU deployment scheme with the ability to detect Sybil attacks and superior multiple attributes.

[0085] The beneficial effects of the present invention are as follows:

[0086] The present invention provides an efficient dynamic roadside unit deployment method and system capable of detecting witch attacks. By establishing a Sybil attack detection constraint and iteratively optimizing the RSU deployment and adjustment scheme based on a plurality of heuristic algorithms under this constraint, the comprehensive performance of the RSU deployment scheme is improved while ensuring the ability to detect Sybil attacks, thereby solving the problem that the comprehensive performance of the RSU deployment scheme is low and the witch attack cannot be detected. The present invention first establishes the RSU service coverage, overlapping range index and initial propagation speed index of the accident information for measuring the RSU deployment performance. Under the Sybil attack detection constraint, the PSO and the two-neighborhood search-meme algorithm are combined to iteratively optimize the RSU deployment scheme, enhance the overall performance and convergence ability of the scheme, and obtain a near-optimal RSU deployment scheme. Secondly, in order to avoid the influence of traffic flow changes on the load imbalance between RSUs, in the process of adjusting the RSU deployment scheme, the present invention uses the Hopcroft-Karp bipartite graph algorithm under the Sybil attack detection constraint to effectively solve the minimum adjustment cost problem, and constructs a heuristic RSU multi-objective adaptive adjustment algorithm based on the Pareto principle to adjust the RSU deployment scheme. Compared with the traditional dynamic RSU deployment scheme, the present invention has higher detection capability for Sybil attacks; has excellent performance in RSU service coverage, overlapping range and initial propagation speed indicators of accident information; has lower adjustment cost in minimum cost adjustment of RSU deployment scheme, better RSU service coverage, overlapping range indicators and more balanced workload between RSUs. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 A simplified flow chart of an embodiment of the present invention;

[0088] Figure 2 This is a diagram illustrating the applicable scenarios of the present invention;

[0089] Figure 3 A specific flow chart for implementing efficient dynamic roadside unit deployment capable of detecting Sybil attacks in the present invention;

[0090] Figure 4 A comparison chart of the comprehensive performance between the RSU deployment solution of the present invention and the traditional RSU deployment solution;

[0091] Figure 5 A comparison chart of the RSU deployment solution in the present invention and the traditional RSU deployment solution in terms of RSU service coverage and overlap indicators;

[0092] Figure 6 This is a comparison chart of the Sybil attack detection capabilities of the RSU deployment solution in the present invention and the traditional RSU deployment solution;

[0093] Figure 7This is a comparison chart of the initial propagation speed index of the accident information between the RSU deployment scheme of the present invention and the traditional RSU deployment scheme;

[0094] Figure 8 A comparison chart of the comprehensive performance between the RSU adjustment scheme of the present invention and the traditional RSU adjustment scheme;

[0095] Fig. 9 A comparison chart of the RSU adjustment scheme of the present invention and the traditional RSU adjustment scheme in terms of RSU service coverage and overlap indicators;

[0096] Fig.10 A comparison chart of the adjustment cost between the RSU adjustment scheme of the present invention and the traditional RSU adjustment scheme;

[0097] Fig.11 A comparison diagram of the RSU adjustment scheme of the present invention and the traditional RSU adjustment scheme in terms of load balancing between RSUs;

[0098] Fig.12 The figure is a comparison chart of the Sybil attack detection capability between the RSU adjustment scheme in the present invention and the traditional RSU adjustment scheme. DETAILED DESCRIPTION

[0099] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0100] The purpose of the present invention is to provide an efficient dynamic roadside unit deployment method capable of detecting witch attacks, so as to solve the problem that the comprehensive performance of the RSU deployment scheme is low and the witch attacks cannot be detected.

[0101] according to Figure 1 As shown, the present invention provides an efficient dynamic roadside unit deployment method capable of detecting a Sybil attack, comprising the following steps:

[0102] Step 1: Collect geographic information, evaluate the roadside unit (RSU) service coverage, overlap range, and initial diffusion speed of accident information in the deployment area, and establish RSU deployment constraints that can detect Sybil attacks based on signal strength;

[0103] Obtain the geographic information of the deployment area, initialize the deployment matrix of the deployment area, collect all possible RSU deployment sites in the deployment area, obtain RSU transmission and influence range, vehicle-to-vehicle communication range and the number of available RSUs, randomly generate an RSU deployment plan based on the geographic information of the deployment area and the information of each entity in the area, obtain the RSU service coverage and overlapping range of the plan, and calculate the initial diffusion speed of the accident information in combination with the vehicle flow in the area, and obtain the deployment constraints for witch attack detection based on the signal strength.

[0104] Step 2: Build an RSU deployment solution based on particle swarm optimization and improved memetic algorithm;

[0105] Under the constraint of Sybil attack detection, the RSU deployment scheme is optimized by combining the fitness value with the advantage of particle swarm optimization and two-neighborhood search-meme algorithm to obtain a near-optimal RSU deployment scheme.

[0106] Step 3: Obtain the transportation and reconstruction RSU overhead, and obtain the minimum RSU adjustment cost based on the bipartite graph matching algorithm;

[0107] According to the market, the RSU reconstruction cost and transportation adjustment cost are obtained. According to the geographic information of the deployment area and the RSU adjustment location, the RSU workload is evaluated according to the traffic flow within the effective range of each RSU in the deployment area, and the minimum RSU adjustment cost is obtained based on the bipartite graph matching algorithm Hopcroft-Karp.

[0108] Step 4: Collect the workload of each RSU and build a heuristic RSU minimum cost adjustment plan based on the Pareto optimal principle.

[0109] The workload of RSU is evaluated according to the traffic flow within the effective range of each RSU in the deployment area, and the fitness value is designed to measure the performance of the RSU adjustment scheme. Under the constraint of Sybil attack, the RSU initial deployment scheme is optimized and iterated using the HRMA3 algorithm based on the Pareto optimal principle to obtain the RSU minimum cost adjustment scheme. While ensuring the comprehensive efficiency of RSU deployment, load balancing between RSUs is achieved with minimum cost.

[0110] according to Figure 2 As shown in the figure, effective locations such as the middle of the road, intersections and near urban landmark buildings are selected as possible deployment sites of RSU L = {L1, L2, L3, ..., L n} will be marked, each vehicle user All vehicles can freely travel in the deployment area of ​​RSU. The traffic volume will determine the workload of RSU and further affect the adjustment plan of RSU. Available RSU R={R1,R2,…,R n} will be properly placed or adjusted, each RSU connects to the IoV control network to provide services for vehicles. In addition, vehicles and RSUs are collectively referred to as carrier units The influence and communication range of RSU will affect Sybil attack detection, signal coverage and overlap indicators respectively. Accident information can be transmitted through vehicle-to-vehicle communication (V2V) and vehicle-to-infrastructure communication (V2I). When an accident occurs in a city block, nearby vehicles will broadcast the accident information, and the IoT control network will receive all IoT activity information in real time. Factors such as RSU load balancing will affect the adjustment plan of RSU. In addition, malicious vehicles that implement Sybil attacks will attack RSU by controlling virtual vehicle users. Therefore, when deploying RSU, the necessary RSU cooperation location distance constraints in actual scenarios should be considered.

[0111] In steps 2 to 4, according to Figure 3 , an efficient dynamic roadside unit deployment method that can detect Sybil attacks obtains information and calculates and iterates, and finally gives a dynamic RSU deployment result that is close to the optimal. The process is as follows:

[0112] S1, obtain the geographic information of the deployment area, initialize the deployment matrix of the deployment area, collect all possible deployment sites of RSU in the deployment area; obtain the RSU transmission and influence range C R , C e , the vehicle-to-vehicle communication range C V and the number of available RSUs N r ;

[0113] S2, randomly generate RSU deployment plan S according to the geographical information of the deployment area and the information of each entity in the area o , get the RSU service coverage and overlapping range g of the scheme cover , and calculate the initial diffusion speed S of accident information in combination with the vehicle flow in the area init ; Based on the signal strength, we get the Sybil attack detection deployment constraint C S ;

[0114] S3, design the fitness value F to measure the performance of the RSU deployment solution, in C S Under the constraint, the PSO-Meme joint heuristic algorithm is used to optimize the RSU deployment scheme according to the fitness value F to obtain a near-optimal RSU deployment scheme S c ;

[0115] S4, obtain RSU reconstruction cost based on the market new and transportation adjustment cost ad ; According to the deployment area geographic information and RSU adjustment location, the minimum RSU adjustment cost Cost is obtained based on the bipartite graph matching algorithm total ;

[0116] S5, evaluates the workload of RSU based on the traffic volume within the effective range of each RSU in the deployment area Design a fitness value F to measure the performance of the RSU adjustment scheme ad ; in C S Under the constraints, the initial deployment plan of RSU is optimized and iterated using the HRMA3 algorithm based on the Pareto optimal principle to obtain the RSU minimum cost adjustment plan S ad , while ensuring the overall efficiency of RSU deployment, load balancing between RSUs is achieved.

[0117] Specifically, in S2, an RSU deployment scheme is randomly generated according to the geographic information of the deployment area and the information of each entity in the area, and the RSU service coverage and overlapping range of the scheme are obtained. The initial diffusion speed of the accident information is calculated in combination with the vehicle flow in the area, and the deployment constraints for Sybil attack detection are obtained based on the signal strength and the necessary RSU cooperation position distance in the actual scenario. The specific steps are as follows:

[0118] S201, the dynamic RSU deployment system extracts the geographical information of the deployment area and the information of each entity in the area and converts it into a deployment matrix M c And randomly generate a deployment plan S in the deployment matrix o ;

[0119] S202, the dynamic RSU deployment system is based on the entity information and deployment plan S in the area. o Get the RSU service coverage and overlapping range g of the solution cover ;

[0120] S203, the dynamic RSU deployment system is based on the entity information, vehicle flow and deployment plan S in the area. o Get the initial diffusion speed S of the accident information init ;

[0121] S204, the dynamic RSU deployment system establishes a Sybil attack detection deployment constraint C based on the signal strength and the required RSU cooperation position distance in the actual scenario. S .

[0122] Specifically, in S3, a fitness value F is designed to measure the performance of the RSU deployment solution, and based on the fitness value F, in C S Under the constraints, the PSO-Meme joint heuristic algorithm is used to optimize the RSU deployment scheme. The specific steps are as follows:

[0123] S301, the dynamic RSU deployment system allocates RSU service coverage and overlap ranges according to the priority of the deployment target. cover and the initial diffusion speed of accident information S init Combined to get the fitness value F;

[0124] S302, in C S Under the constraint, the dynamic RSU deployment system uses the particle swarm optimization algorithm to optimize the relatively optimal deployment solution S according to the response value F. r , and put it into the initial solution set;

[0125] S303, the dynamic RSU deployment system sorts all solutions according to the fitness value F and selects the elite solution e with better performance;

[0126] S304, in C s Under the constraints, the elite solution set is crossover and mutation operations are performed to obtain the offspring solution set P;

[0127] S305, the dynamic RSU deployment system performs a neighborhood search on the offspring solution set obtained after each crossover and mutation operation and selects a solution with a large fitness value to update the offspring solution set P;

[0128] S306: If the two-neighborhood search-meme algorithm has not reached the maximum number of iterations, go to S303, otherwise output a near-optimal RSU deployment solution S c .

[0129] Specifically, in S4, the dynamic RSU deployment system first obtains the RSU reconstruction cost and transportation adjustment cost according to each physical market in the region, and then calculates the total RSU adjustment cost according to the reconstruction and transportation adjustment costs. After that, the dynamic RSU deployment system will obtain the RSU adjustment strategy and the minimum total adjustment cost based on the bipartite graph matching algorithm according to the deployment area geographic information and RSU site adjustment information. The specific steps are as follows:

[0130] S401, the dynamic RSU deployment system formulates the RSU reconstruction cost based on the entities and market information in the region new ;

[0131] S402, the dynamic RSU deployment system formulates the RSU transportation adjustment cost according to the entities and market information in the region ad ;

[0132] S403, the dynamic RSU deployment system rebuilds the cost according to the RSU new Cost adjustment with RSU transportation ad Formulate RSU and calculate total cost total ;

[0133] S404: The dynamic RSU deployment system obtains the minimum total cost of RSU adjustment based on the bipartite graph matching algorithm Hopcroft-Karp total .

[0134] Specifically, in S5, the workload of RSU is evaluated according to the traffic flow within the effective range of each RSU in the deployment area; the fitness value for measuring the performance of the RSU adjustment scheme is designed; under the Sybil attack detection deployment constraint, the RSU initial deployment scheme is optimized and iterated using the HRMA3 algorithm to obtain an efficient and secure RSU minimum cost adjustment scheme. The specific steps are as follows:

[0135] S501, the dynamic RSU deployment system evaluates the workload of the RSU according to the traffic flow within the effective range of each RSU in the deployment area

[0136] S502, the dynamic RSU deployment system designs a fitness value F to measure the performance of the RSU adjustment scheme according to the RSU service quality and the load balancing index between the RSUs. ad ;

[0137] S503, the dynamic RSU deployment system randomly generates an RSU adjustment plan and converts the RSU deployment plan S close to the optimal one into an RSU adjustment plan. c Add to the solution set O of the adjustment optimization solution;

[0138] S504: the solutions in the solution set O are evaluated according to the fitness value F. ad To sort and filter, in C S Under the constraints, crossover and mutation operations are performed on the selected solutions to obtain an offspring population pop equal to the population size of the optimization solution set O;

[0139] S505, the dynamic RSU deployment system merges the solution set O and the offspring population pop, according to the fitness value F ad Perform fast non-dominated sort;

[0140] S506, the dynamic RSU deployment system adds the deployment efficiency reference point to the sorted population and performs niche selection, selecting an elite population set e equal to the number of the next generation pop population ad ;

[0141] S507, Dynamic RSU deployment system performs genetic operation to generate and elite solution population set e ad The offspring with equal population size pop′;

[0142] S508, the dynamic RSU deployment system performs a neighborhood search on the offspring solution set obtained after each crossover and mutation operation and selects a solution with a large fitness value to update the offspring solution set pop′;

[0143] S509: If the HRMA3 algorithm has not reached the maximum number of iterations, proceed to S504; otherwise, output the optimal RSU minimum cost adjustment solution.

[0144] In particular, in S202, due to the influence of obstacles such as buildings on the RSU signal, the signal strength will be interfered to a certain extent, affecting the communication quality. The effective signal coverage area of ​​an RSU is in square meters and is defined as a polygon P = {P1, P2, ..., P i}, each RSU R n Each has its effective signal coverage area P i Assuming that Nr RSUs are available, the overlapping area of ​​the effective RSU signals is the accumulation of the overlapping areas between multiple polygons, which can be expressed as:

[0145]

[0146] Where O(a,b) represents the function for calculating the overlapping area of ​​two polygons a and b. Therefore, g is determined by the coverage area and overlapping area of ​​the RSU effective signal. cover It can be expressed as:

[0147]

[0148] Among them A T It is the total area in the city that needs to be covered by effective signals.

[0149] In particular, in S203, there is an accident site A in the RSU deployment area, and it is assumed that all V2V transmissions are successful. The calculation of the average ISSAI value can be expressed as:

[0150]

[0151] Where S v is the speed of vehicle v, and N is the number of all vehicles involved in the propagation of accident information. v It represents the sum of the time spent by vehicle v to transmit accident information using carrier unit Ui, which can be expressed as:

[0152]

[0153] In particular, in S204, the necessary RSU cooperation position distance for Sybil attack detection according to the actual scenario can be expressed as:

[0154]

[0155] Among them C e Indicates the impact range of RSU.

[0156] In particular, in S301, the RSU service coverage and overlap range g are allocated according to the priority of the deployment target. cover and the initial diffusion speed of accident information S init Combined to get the fitness value F:

[0157]

[0158] Among them, θ1 and θ2 represent the normalized weights from 0 to 1, T init Represents the short message propagation time after the incident.

[0159] In particular, in S303, all solutions are sorted according to the fitness value F and elite solutions with better performance are selected:

[0160] Sort(F(s1),F(s2),…,F(s e ))e <Num(S r )

[0161] In particular, in S305, the single solution in the elite offspring population after crossover and mutation operations is defined as S p , whose neighborhood is N(S p ), that is, from the current solution S p By exploring the solution set generated by small step size, if make:

[0162]

[0163] Then the elite offspring solution S p Update to S′ p .

[0164] In particular, in S401, if the RSU in the adjusted solution Sol′ increases compared to the deployment solution Sol before adjustment, new RSU equipment will be purchased and deployed, and the cost calculation formula for the newly deployed RSU is:

[0165] Cost new =(|Sol′|-|Sol|)·(∈+δ)

[0166] Where ∈ is the purchase cost and δ is the construction cost of a new RSU.

[0167] In particular, in S402, the transportation adjustment of the RSU includes the disassembly, installation and transportation costs, so the RSU transportation adjustment cost is calculated as follows:

[0168]

[0169] Where |R ad | is the number of RSUs that need to be adjusted, is the transportation cost per kilometer of RSU, D ad is the Euclidean distance between the RSU sites that need to be adjusted.

[0170] In particular, in S403, based on the RSU transportation adjustment cost and the reconstruction cost obtained above, the total cost of calculating the RSU adjustment can be expressed as:

[0171] Cost total =Cost new +Cost ad

[0172] In particular, in S404, the existing RSU positions and the newly adjusted RSU positions are regarded as a bipartite graph set V r and V r ′, the weight E of the edge represents the cost of moving from the current position to the new position, so the bipartite graph can be expressed as:

[0173] G(V r ,V r ′;e),e∈E(V r ,V r ′)={D(R1,R′1),D(R1,R′2),…,D(R n ,R′ m )}

[0174] Where D(R n ,R′ m ) represents R n ∈V r and R m ′∈V r The Euclidean distance between ′.

[0175] Specifically, in S501, the vehicle selects the nearest RSU to interact with, according to the RSU R n Traffic flow in the coverage area The interaction strength is Define the maximum load threshold of RSU through investigation Each RSU The workload can be expressed as:

[0176]

[0177] In particular, in S502, according to the signal coverage and overlapping range g of the deployed RSUs, cover , load balancing between RSUs and the minimum cost of adjusting RSUs, and obtain the fitness value representing the performance of the RSU adjustment scheme:

[0178] F ad =θ1·g cover -θ3·Cost total -θ4·B R

[0179] Among them, θ1, θ3 and θ4 represent the normalized weights from 0 to 1, B R To evaluate the standard deviation statistic of load balance between RSUs, it is expressed as:

[0180]

[0181] in is the average workload of all RSUs in the deployment area.

[0182] In particular, in S505, according to the fitness value F ad Perform a quick non-dominated sort on the deployment solution population to prepare for the subsequent introduction of deployment efficiency reference points for niche selection:

[0183] Sort non-domin (F ad (s1),F ad (s2),…,F ad (s l ))

[0184] In particular, in S508, a single solution in the offspring population after crossover and mutation operations is defined as S l , whose neighborhood is N(S l ), that is, from the current solution S l By exploring the solution set generated by small step size, if make:

[0185]

[0186] Then the elite offspring solution S l Update to S′ l .

[0187] according to Figure 4 As shown, Figure 4 A comparison chart of the comprehensive performance of the RSU deployment scheme PSO-Meme proposed in this invention and the traditional RSU deployment scheme is given. Figure 4 In the above figure, a higher fitness value represents a better overall performance of RSU deployment. The scheme PSO-Meme shows excellent performance, with an average fitness value of 0.9431. Although the optimal condition (using enumeration method, ignoring obstacles, and 100% information transmission success rate) has the highest fitness value, it is obviously not suitable for the RSU deployment field because its space complexity and time complexity are impractical.

[0188] according to Figure 5 As shown, Figure 5A comparison chart of the RSU service coverage and overlap indicators between the RSU deployment scheme PSO-Meme proposed in the present invention and the traditional RSU deployment scheme is given. Obviously, the PSO-Meme scheme has the highest RSU service coverage and overlap values ​​under different RSU deployment numbers, which means that the PSO-Meme scheme can maximize the effective use of RSU resources.

[0189] according to Figure 6 As shown, Figure 6 A comparison chart of the Sybil attack detection capabilities of the RSU deployment scheme PSO-Meme proposed in this invention and the traditional RSU deployment scheme is given. As the number of RSUs increases, the probability of detecting a Sybil attack will also increase. Please note that the detection rate here does not include the false alarm rate. Since the scheme PSO-Meme has Sybil security constraints, when the number of RSUs reaches 6, the detection rate can reach 88.92. This also proves that PSO-Meme can provide higher security for Sybil detection.

[0190] according to Figure 7 As shown, Figure 7 A comparison chart of the initial propagation speed of accident information between the RSU deployment scheme PSO-Meme proposed in the present invention and the traditional RSU deployment scheme is given. The higher the initial propagation speed of accident information, the faster the initial propagation speed of the accident, which also means that the accident information reaches the RSU faster. It can be seen that the PSO-Meme scheme has the highest initial propagation speed value of the accident in environments with different numbers of vehicles, so that the accident vehicle can convey the accident information through the RSU in the shortest time at any site.

[0191] according to Figure 8 As shown, Figure 8 The comparison chart of the comprehensive performance between the RSU adjustment scheme HRMA3 proposed by the present invention and the traditional RSU adjustment scheme is given. Figure 4 same, Figure 8 Among them, the HRMA3 scheme has the average fitness value closest to the optimal condition (using the enumeration method, ignoring obstacles, and the information transmission success rate is 100%). This result shows that the comprehensive performance of the HRMA3 scheme is better than other traditional schemes.

[0192] according to Fig. 9 As shown, Fig. 9 A comparison chart of the RSU service coverage and overlap indicators between the RSU adjustment scheme proposed in the present invention and the traditional RSU adjustment scheme is given. It can be seen that under different additional adjustment available RSU numbers, HRMA3 has the highest RSU service coverage and overlap, indicating that it has the best RSU service quality even after RSU adjustment.

[0193] according to Fig.10 As shown, Fig.10 A comparison chart of the adjustment cost between the RSU adjustment scheme HRMA3 proposed in the present invention and the traditional RSU adjustment scheme is given. HRMA3 can make the least changes to the existing RSU deployment scheme, so it spends the least on the cost of RSU adjustment, indicating that the HRMA3 scheme has better adjustment capabilities in different scenarios of additional adjustment of the number of available RSUs.

[0194] according to Fig.11 As shown, Fig.11 A comparison chart of the RSU adjustment scheme HRMA3 proposed by the present invention and the traditional RSU adjustment scheme in terms of load balancing between RSUs is given. It can be intuitively seen that compared with other traditional schemes, the HRMA3 scheme has the lowest load balancing standard deviation under different additional adjustment available RSU numbers, indicating that HRMA3 is most capable of achieving workload balancing between each RSU.

[0195] according to Fig.12 As shown, Fig.12 A comparison chart of the Sybil attack detection capabilities of the RSU adjustment scheme HRMA3 proposed in the present invention and the traditional RSU adjustment scheme is given. Although when the number of additional available RSUs is 0, the Sybil attack detection rate of the HRMA3 scheme is slightly lower than the original RSU deployment before adjustment due to the optimization target priority, the magnitude is very small, about 2.3%. Compared with other traditional RSU adjustment schemes, the Sybil attack detection rate of the HRMA3 scheme is the highest under different additional adjustment numbers of available RSUs.

[0196] Example:

[0197] In order to ensure the reliability of the results without loss of generality, this example assigns a simulation system for an efficient dynamic roadside unit deployment scheme that can detect witch attacks. Each group of situations in the experiment was simulated 20 times, and the average value was used as the final result. The experiment randomly selected a city block in California, USA, which is 994×1011m 2 , AT is 66.41% of the total block area. The number of available RSUs N rBetween 3 and 6 vehicles, the number of vehicles V is between 20 and 100. Assuming the channel bandwidth is 20MHz, the signal strength within the effective signal range of the RSU should be greater than -65dBm, the RSU influence range radius is 250m, the RSU transmission range radius is 200m, and the V2V communication radius is 25m. The physical CPU has from 700MHz to 2000MHz, the accident information size is 10Mb, and 30 CPU instructions are required for each Mb of information. The population size in the PSO algorithm is 16 and the number of iterations is 10 rounds. In the two-neighborhood search-meme algorithm, the population size is 8, the number of offspring is 10, the number of elites is 3, and the maximum number of iterations is 20. In addition, in the HRMA3 algorithm, the maximum number of RSUs for relocation is 3, the number of additional available RSUs is between 0 and 2, the maximum number of iterations is 60, and the number of offspring populations is 42.

[0198] In summary, the present invention provides an efficient dynamic roadside unit deployment method that can detect witch attacks. By establishing Sybil attack detection constraints and iteratively optimizing RSU deployment and adjustment schemes based on multiple heuristic algorithms under these constraints, the comprehensive performance of the RSU deployment scheme is improved while ensuring the ability to detect Sybil attacks, solving the problem that the comprehensive performance of the RSU deployment scheme is low and the witch attack cannot be detected. The present invention first establishes RSU service coverage, overlapping range indicators and accident information initial propagation speed indicators to measure the RSU deployment performance. Under the Sybil attack detection constraint, the PSO and two-neighborhood search-meme algorithm are combined to iteratively optimize the RSU deployment scheme, enhance the overall performance and convergence ability of the scheme, and obtain a near-optimal RSU deployment scheme. Secondly, in order to solve the problem that changes in vehicle flow will cause unbalanced loads between RSUs, in the process of adjusting the RSU deployment scheme, the present invention uses the Hopcroft-Karp bipartite graph algorithm under the Sybil attack detection constraint to effectively solve the minimum adjustment cost problem, and constructs a heuristic RSU multi-objective adaptive adjustment algorithm based on the Pareto principle to adjust the RSU deployment scheme. Compared with the traditional dynamic RSU deployment scheme, the present invention has higher detection capability for Sybil attacks; has excellent performance in RSU service coverage, overlapping range and initial propagation speed indicators of accident information; has lower adjustment cost in minimum cost adjustment of RSU deployment scheme, better RSU service coverage, overlapping range indicators and more balanced workload between RSUs.

[0199] In one embodiment of the present invention, a highly efficient dynamic roadside unit deployment system capable of detecting a Sybil attack is provided, which can be used to implement the highly efficient dynamic roadside unit deployment method capable of detecting a Sybil attack. Specifically, the system includes:

[0200] The module for acquiring geographic information and entity information of the RSU deployment area is used to obtain data such as the volume of obstacles, possible RSU deployment sites near intersections and city landmarks, RSU entity impact and communication range, vehicle equipment information, signal strength and traffic volume received by RSU from each vehicle interaction, etc. The traffic volume information corresponding to each historical timestamp collected by each RSU in the deployment area is cleaned and missing values ​​are filled. Each traffic volume information is normalized by wavelet transform to ensure that the data format is suitable for subsequent RSU load balancing evaluation; the basic performance attributes of each RSU, vehicle, etc. in the deployment area are collected, and after removing outliers, a weighted average is used to formulate a basic entity information database without loss of generality; the information of roads, buildings, intersections, etc. in the deployment area is projected onto a plane through the geographic information system and traffic monitoring system to form an RSU deployment matrix, providing an accurate model basis for the subsequent scientific deployment adjustment plan of RSU.

[0201] The multi-performance measurement module of the RSU deployment scheme is used to formulate Sybil attack detection constraints in RSU deployment and highly representative RSU deployment scheme measurement indicators based on the acquired geographic information of the RSU deployment area and the information of each entity. Through the rule engine, statistical analysis, mathematical quantification and other methods, measurement indicators are formulated for various performances of RSU, including Sybil attack detection security constraints formulated according to the necessary RSU cooperation location distance in the actual scenario, RSU signal coverage and overlap range indicators to measure the regional RSU service quality and deployment efficiency, accident information initial propagation speed indicators to measure the system's ability to notify future vehicles after an accident occurs in the region, regional RSU standard deviation statistics to measure the load balancing between RSUs in the region, RSU minimum adjustment cost indicators to measure the quality of RSU deployment and the fitness value to measure the comprehensive performance of RSU adjustment schemes.

[0202] The PSO-Meme joint heuristic RSU deployment module is used to iteratively optimize the RSU deployment plan. First, the PSO's ability of fast convergence and global search is used to optimize the initial RSU deployment plan to produce a relatively better RSU deployment result. Then, the two-neighborhood search-meme algorithm with a larger solution set exploration domain and a higher solution set fitness value is used to iteratively optimize the relatively better RSU deployment result of PSO to produce the closest to the optimal RSU deployment result. Based on the traditional meme algorithm, the two-neighborhood search-meme algorithm performs a neighborhood search after the crossover and mutation steps to improve the fitness value of the overall solution set;

[0203] The minimum cost RSU deployment adjustment module is used to iteratively optimize the minimum cost RSU adjustment scheme based on RSU service quality and load balancing between RSUs. The calculation of the minimum cost of RSU adjustment uses the Hopcroft-Karp algorithm in the bipartite graph to improve the efficiency of computing resources; the minimum cost RSU adjustment scheme uses fast non-dominated sorting to stratify the population, and then uses the RSU deployment efficiency reference point and niche selection based on the fitness value to perform elite screening on the population of each layer to generate the offspring population, and repeats multiple times until the maximum number of iterations is reached. Finally, the RSU deployment scheme is adjusted to obtain an RSU deployment scheme with the ability to detect Sybil attacks and superior multi-attributes.

[0204] The division of modules in the embodiments of the present invention is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional module in each embodiment of the present invention may be integrated into one processor, or may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of software functional modules.

[0205] In one embodiment of the present invention, a computer device is provided, the computer device comprising a processor and a memory, the memory being used to store a computer program, the computer program comprising program instructions, and the processor being used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, which are suitable for implementing one or more instructions, and are specifically suitable for loading and executing one or more instructions in a computer storage medium to implement a corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of an efficient dynamic roadside unit deployment method that can detect witch attacks.

[0206] In one embodiment of the present invention, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It can be understood that the computer-readable storage medium here can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the efficient dynamic roadside unit deployment method that can detect witch attacks in the above embodiment.

[0207] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0208] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0209] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0210] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0211] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An efficient dynamic roadside unit deployment method capable of detecting Sybil attacks, characterized in that: The steps include: S1: Obtain the geographic information of the deployment area, initialize the deployment matrix of the deployment area, collect all possible deployment sites of RSUs in the deployment area; obtain the RSU transmission and influence range C R , C e , the vehicle-to-vehicle communication range C V and the number of available and additionally adjusted available RSUs N r 、N e ; S2: Randomly generate RSU deployment plan S based on the geographic information of the deployment area and the information of each entity in the area o , we get solution S o RSU service coverage and overlapping range g cover , and calculate the initial diffusion speed S of accident information in combination with the vehicle flow in the area init ; Based on the signal strength, we get the Sybil attack detection deployment constraint C S ; S3: Design the fitness value F to measure the performance of the RSU deployment solution. S Under the constraint, according to the fitness value F, the particle swarm-two neighborhood search meme PSO-Mem e The combined heuristic algorithm optimizes the RSU deployment scheme to obtain a near-optimal RSU deployment scheme S c ; S4: Get RSU reconstruction cost based on the market new and transportation adjustment cost ad According to the deployment area geographic information and RSU adjustment location, the minimum RSU adjustment cost Cost is obtained based on the bipartite graph matching algorithm Hopcroft-Karp total ; S5: Evaluate the workload of RSUs based on the traffic volume within the effective range of each RSU in the deployment area Design a fitness value F to measure the performance of the RSU adjustment scheme ad ; in C S Under the constraints, the RSU initial deployment plan is optimized and iterated using the heuristic RSU multi-objective adaptive adjustment algorithm HRMA3 based on the Pareto optimal principle to obtain the RSU minimum cost adjustment plan S ad , while ensuring the overall efficiency of RSU deployment, load balancing between RSUs is achieved.

2. According to the method of claim 1, the efficient dynamic roadside unit deployment method capable of detecting Sybil attacks is characterized in that: The step S2 is specifically as follows: S201: The dynamic RSU deployment system extracts the geographic information of the deployment area and the information of each entity in the area and converts it into a deployment matrix M c And randomly generate a deployment plan S in the deployment matrix o ; S202: The dynamic RSU deployment system is based on the entity information and deployment plan S in the area o Get the RSU service coverage and overlapping range g of the solution cover ; The effective signal coverage area of ​​an RSU is defined as a polygon P = {P1, P2, ..., P i }, each RSUR n Each has its effective signal coverage area P i ; Assuming that Nr RSUs are available, the overlapping area of ​​the RSU effective signal is the accumulation of the overlapping areas between multiple polygons, expressed as: Where O(a, b) represents the function for calculating the overlapping area of ​​two polygons a and b; Therefore, g is determined by the coverage area and overlap area of ​​the RSU effective signal. cover It is expressed as: Among them A T The total area in the city that needs to be covered by effective signals; S203: The dynamic RSU deployment system is based on the entity information, vehicle flow and deployment plan S in the area o Get the initial diffusion speed S of the accident information init ; There is an accident site A in the RSU deployment area. Assuming that all V2V transmissions are successful, the calculation of the average ISSAI value is expressed as: Where S v is the speed of vehicle v, N is the number of all vehicles involved in the accident information dissemination; T v It represents the sum of the time spent by vehicle v to transmit accident information using carrier unit Ui, expressed as: S204: The dynamic RSU deployment system establishes a Sybil attack detection deployment constraint C based on the signal strength and the required RSU cooperation location distance in the actual scenario S ; The necessary RSU cooperation position distance for Sybil attack detection according to the actual scenario is expressed as: Among them C e Indicates the impact range of RSU.

3. The method for an efficient dynamic roadside unit deployment capable of detecting Sybil attacks according to claim 2, characterized in that: The step S3 is specifically as follows: S301: The dynamic RSU deployment system allocates RSU service coverage and overlap ranges according to the priority of the deployment target. cover and the initial diffusion speed of accident information S init Combined to get the fitness value F; Among them, θ1 and θ2 represent the normalized weights from 0 to 1, T init Represents the short message transmission time after the accident; S302: In C S Under the constraint, the dynamic RSU deployment system uses the particle swarm optimization algorithm to optimize the r deployment solutions S with the largest response value F according to the response value F. r , and put it into the initial solution set; S303: The dynamic RSU deployment system sorts all solutions in the initial solution set according to the fitness value F and selects e elite solutions to form an elite class E; E=Sort(F(s1),F(s2),…,F(s e ))e<Num(S r ) S304: In C S Under the constraints, the elite solution set E is crossover and mutation operated to obtain the offspring solution set P; S305: The dynamic RSU deployment system performs a neighborhood search on the offspring solution set obtained after each crossover and mutation operation and selects a solution with a large fitness value to update the offspring solution set P; Define the single solution in the elite offspring population after crossover and mutation operations as S p , whose neighborhood is N(S p ), that is, from the current solution S p By exploring the solution set generated by small step size, if make Then the elite offspring solution S p Update to S′ p ; S306: If the two-neighborhood search-meme algorithm has not reached the maximum number of iterations, go to S303, otherwise output a near-optimal RSU deployment solution S c .

4. The method for an efficient dynamic roadside unit deployment capable of detecting Sybil attacks according to claim 3 is characterized in that: The step S4 is specifically as follows: S401: Dynamic RSU deployment system formulates RSU reconstruction cost based on each entity and market information in the region new ; Compared with the deployment plan Sol before adjustment, if the RSU in the adjustment plan Sol′ increases, new RSU equipment will be purchased and deployed. The cost calculation formula for the newly deployed RSU is: Cost new =(|Sol′|-|Sol|)·(∈+δ) Where ∈ is the purchase cost and δ is the construction cost of a new RSU; S402: The dynamic RSU deployment system formulates the RSU transportation adjustment cost based on the entities and market information in the region. ad ; The shipping adjustment for RSU includes the removal, installation and transportation costs, so the RSU shipping adjustment overhead is calculated as follows: Where |R ad | is the number of RSUs that need to be adjusted, is the transportation cost per kilometer of RSU, D ad is the Euclidean distance between the RSU sites that need to be adjusted; S403: Dynamic RSU deployment system rebuilds the cost according to RSU new Cost adjustment with RSU transportation ad Formulate RSU and calculate total cost total ; Cost total =Cost new +Cost ad S404: Dynamic RSU deployment system obtains the minimum total cost of RSU adjustment based on the bipartite graph matching algorithm Hopcroft-Karp total ; The existing RSU positions and the newly adjusted RSU positions are considered as a bipartite graph set V r and V r ′, the weight E of the edge represents the cost of moving from the existing position to the new position, which is expressed as: G(V r ,V′ r ;e), e∈E(V r ,V′ r )={D(R1,R′1),D(R1,R′2),…,D(R n ,R′ m )}, Where D(R n , R′ m ) represents R n ∈V r and R m ′∈V r The Euclidean distance between ′.

5. The method for an efficient dynamic roadside unit deployment capable of detecting Sybil attacks according to claim 4, characterized in that: The step S5 is specifically as follows: S501: The dynamic RSU deployment system evaluates the workload of the RSU based on the traffic flow within the effective range of each RSU in the deployment area The vehicle selects the nearest RSU to interact with, based on the RSU R n Traffic flow in the coverage area The interaction strength is Define the maximum load threshold of RSU through investigation Each RSU The workload is represented as: S502: The dynamic RSU deployment system designs a fitness value F to measure the performance of the RSU adjustment scheme according to the RSU service quality and the load balancing index between the RSUs. ad ; According to the signal coverage and overlapping range of deployed RSU g cover , load balancing between RSUs and the minimum cost of adjusting RSUs, and obtain the fitness value representing the performance of the RSU adjustment scheme: F ad =θ1·g couer -θ3·Cost total -θ4·B R , Among them, θ1, θ3 and θ4 represent the normalized weights from 0 to 1, B R To evaluate the standard deviation statistic of load balance between RSUs, it is expressed as: in is the average workload of all RSUs in the deployment area; S503: The dynamic RSU deployment system randomly generates an RSU adjustment plan and converts the RSU deployment plan S close to the optimal one into an RSU adjustment plan. c Add to the solution set O of the adjustment optimization solution; S504: The solutions in the solution set O are evaluated according to the fitness value F ad To sort and filter, in C S Under the constraints, crossover and mutation operations are performed on the selected solutions to obtain an offspring population pop equal to the population size of the optimization solution set O; S505: The dynamic RSU deployment system merges the solution set O and the offspring population pop according to the fitness value F ad Perform fast non-dominated sort; According to the fitness value F ad Perform a quick non-dominated sort on the deployment solution population to prepare for the subsequent introduction of deployment efficiency reference points for niche selection: Sort non-domin (F ad (s1),F ad (s2),…,F ad (s l )) S506: The dynamic RSU deployment system adds the deployment efficiency reference point to the sorted population and performs niche selection to select an elite population set e that is equal to the number of the next generation pop population. ad ; S507: Dynamic RSU deployment system performs genetic operations to generate and elite solution population sets ad The offspring with equal population size pop′; S508: The dynamic RSU deployment system performs a neighborhood search on the offspring solution set obtained after each crossover and mutation operation and selects a solution with a large fitness value to update the offspring solution set pop′; Define the single solution in the offspring population after crossover and mutation operations as S l , whose neighborhood is N(S l ), that is, from the current solution S l By exploring the solution set generated by small step size, if make: Then the elite offspring solution S l Update to S′ l ; S509: If the HRMA3 algorithm has not reached the maximum number of iterations, proceed to S504; otherwise, output the optimal RSU minimum cost adjustment plan.

6. A roadside unit deployment system using the efficient dynamic roadside unit deployment method as claimed in claim 1, comprising an RSU deployment area geographic information and entity information acquisition module, an RSU deployment scheme multi-performance measurement module, a PSO-Mem e Combined heuristic RSU deployment module and minimum cost RSU deployment adjustment module; The RSU deployment area geographic information and entity information acquisition module is used to obtain the RSU deployment area, including the obstacle location volume, possible RSU deployment sites near intersections and city landmark buildings, RSU entity impact and communication range, vehicle equipment information, and signal strength and traffic flow data received by the RSU for each vehicle interaction; the traffic flow information corresponding to each historical timestamp collected by each RSU in the deployment area is cleaned and missing values ​​are filled, and each traffic flow information is normalized by wavelet transform to ensure that the data format is suitable for subsequent RSU load balancing evaluation; The basic performance attributes of each RSU and vehicle in the deployment area are collected, and after removing outliers, a weighted average is used to develop a basic entity information database without loss of generality; through the geographic information system and traffic monitoring system, the roads, buildings, intersections and other information in the deployment area are projected onto a plane to form an RSU deployment matrix, providing an accurate model basis for the subsequent scientific deployment adjustment plan of RSU; The RSU deployment scheme multi-performance measurement module is used to formulate Sybil attack detection constraints in RSU deployment and highly representative RSU deployment scheme measurement indicators based on the acquired RSU deployment area geographic information and entity information; through rule engine, statistical analysis, and mathematical quantification methods, measurement indicators are formulated for multiple performances of RSU, including Sybil attack detection security constraints formulated according to the necessary RSU cooperation location distance in actual scenarios, RSU signal coverage and overlap range indicators for measuring regional RSU service quality and deployment efficiency, accident information initial propagation speed indicators for measuring the system's ability to notify future vehicles after an accident occurs in the region, regional RSU standard deviation statistics for measuring load balancing between RSUs in the region, RSU minimum adjustment cost indicators for measuring the quality of RSU deployment and the quality of RSU adjustment strategies in the region, fitness values ​​for measuring the comprehensive performance of RSU deployment, and fitness values ​​for measuring the comprehensive performance of RSU adjustment schemes; The PSO-Meme combined heuristic RSU deployment module is used to iteratively optimize the RSU deployment plan; first, the PSO's ability of fast convergence and global search is used to optimize the initial RSU deployment plan to produce a relatively good RSU deployment result, and then the relatively good RSU deployment result of PSO is iteratively optimized using a two-neighborhood search-meme algorithm with a large solution set exploration domain and a high solution set fitness value to produce the closest to the optimal RSU deployment result; the two-neighborhood search-meme algorithm performs a neighborhood search after the crossover and mutation steps on the basis of the meme algorithm to improve the fitness value of the overall solution set; The minimum cost RSU deployment adjustment module is used to iteratively optimize the minimum cost RSU adjustment scheme based on RSU service quality and load balancing between RSUs; the calculation of the minimum cost of RSU adjustment adopts the Hopcroft-Karp algorithm in the bipartite graph to improve the efficiency of computing resources; the minimum cost RSU adjustment scheme uses fast non-dominated sorting to stratify the population, and then uses the RSU deployment efficiency reference point and niche selection based on the fitness value to perform elite screening on the population of each layer to generate a descendant population, and repeats multiple times until the maximum number of iterations is reached, and finally adjusts the RSU deployment scheme to obtain an RSU deployment scheme with the ability to detect Sybil attacks and superior multiple attributes.

Images