A Study Of Cluster-based Resource Allocation Algorithms In Vehicular Network Communications

In the near future,Intelligent Traffic Systems(ITS)will be a major component of cities around the world,improving road safety and providing recreational services,but currently ITS suffer from a lack of bandwidth and low utilization.In order to meet different service demands and improve spectrum utilization,Vehicle to Vehicle(V2V)links are allowed to multiplex the spectrum resources of Vehicle to Infrastructure(V2I)links,but spectrum multiplexing can cause co-channel interference,the presence of which leads to attenuation of the transmission rate of vehicles and thus affects the reliability of data transmission,so it is necessary to develop an effective resource allocation strategy to reduce the interference between V2 V and V2 I links.Furthermore,in Vehicle Edge Computing(VEC)scenarios with limited wireless and computational resources,stringent latency requirements are challenging problems and only joint optimization of resource allocation and task offloading can reduce system latency.The main research elements of the thesis are as follows.First,considering a V2 V communication-enabled vehicular network communication scenario,in which multiple V2 V links can multiplex the spectrum resources of the same V2 I link,the thesis proposes a vehicular network resource allocation algorithm based on graph coloring and three-dimensional matching.The algorithm first uses graph coloring to cluster V2 V links,then optimally solves the transmit power of V2 I links and V2 V links,and finally uses a three-dimensional matching algorithm to optimally allocate channel resources to V2 I links,V2 V clusters and resource blocks,so as to reduce the interference between links using the same resource.From the theoretical analysis and experimental results,it can be seen that the method effectively improves the sum rate of V2 I links and converges to a suboptimal solution with a relatively small number of iterations,while effectively solving the problem of low spectrum utilization in vehicular network communications.This thesis further extends the vehicular network resource allocation problem to the VEC scenario and proposes to optimally solution the resource allocation and task offloading problem in vehicular network communication.As computational and spectrum resources are also limited in the VEC scenario,it may not be possible to guarantee the best experience for all users,so this thesis performs a joint optimization solution for spectrum and computational resources.To reduce system latency,joint resource allocation and task offloading management are proposed,in which V2 V clustering,transmission power control,sub-channel allocation,optimal offloading ratio and computational resource allocation schemes are given.Due to the heterogeneous nature of the offloading task,the optimization problem is transformed into two phases.In the first stage,the K-Means algorithm and the 3D matching algorithm are used for resource allocation;in the second stage,the Lagrange multiplier method is used to allocate computational resources and the optimal offloading ratio.Finally,the theoretical analysis and numerical results show that the proposed scheme can effectively reduce the system latency with lower system energy consumption compared to other algorithms.