Research Of Real-time Information Service And Task Offloading Technologies In Heterogenous VANETs

In recent years,vehicular ad-hoc networks(VANETs)has been considered as one of the most promising research field to meet the market demand and industry evolution.It promotes the innovative development of the automobile and information industry and builds a novel automobile and transportation service model.Efficient information services in the heterogenous VANETs are the basis for relieving traffic jam,improving road safety and realizing green transportation.However,the information service efficiency is greatly affected due to the high dynamics and distributed architecture of VANETs.With above motivation,this work proposes three kinds of service architectures of VANET,and conducts the research on cooperative data sharing,information routing and task offloading algorithm.The critical challenges are proposed as follows.Firstly,how to improve the information sharing efficiency in a highly dynamic and distributed vehicular network,where the moving vehicles have very short time for data sharing and serious interference occurs between concurrent transmissions.Secondly,how to achieve reliable delay-constrained routing in a large-scale low-density vehicular communication environment,where the intermittent network connectivity causes frequent routing interruption and long latency.Thirdly,how to achieve high-reliability low-latency computation offloading in a high-demand heterogenous vehicular network,where the fog computation nodes have very limited communication and computation resources.Specifically,this work will deeply investigate the vehicular information services from the perspective of architecture design,problem formulation,and the algorithm proposal and analysis.The main research contributions are summarized as follows:(1)Research on cooperative data sharing algorithm in VANETs via distributed V2 V communication.Firstly,we present a novel information service architecture in bidirectional road scenarios.Then,in order to maximize the throughput of data sharing through single-hop V2 V,we propose a MAC-layer bandwidth reuse strategy based on the interference analysis between concurrent V2 V communications,a broadcast mechanism based on dynamically clustering and time slot division,as well as an election policy of sender vehicle and broadcast item based on MAC-layer backoff mechanism.Finally,we build the simulation platform to verify the complexity and the service ratio of the proposed data sharing algorithm.(2)Based on the research of single V2 V communication,we further consider the multi-hop routing problem via hybrid V2V/V2 I communication.Firstly,we propose an SDN-based heterogenous vehicular service architecture by coordinating DSRC interface and cellular interface to conduct routing process via hybrid V2 V,V2R and V2 C communication.Then,we propose the analysis model of encounter probability,expected reachability and expected delay of multi-hop routing path based on vehicles' real-time position and predicable trajectory,and the low-complexity path searching strategy based on Dijkstra algorithm,as well as the path selection mechanism based on different routing metrics.The experiment results show the characteristics and advantages of different routing mechanisms in balancing the system delivery ratio and bandwidth efficiency.(3)Based on the research of SDN-based vehicular architecture and protocol,we further investigate the fusion architecture of fog computing and vehicular network.We first present a task offloading architecture by integrating the fog computing and VANETs.Then,we solve the power allocation problem based on the Conditional-Value-at-Risk theory with the reliability guarantees,and the multi-period task allocation problem based on the branch-and-bound algorithm with the low-latency requirement,as well as an iteration algorithm to adjust the power and task allocation solution for minimizing the delay with reliability constraints.Comprehensive simulation studies are conducted to explore the efficiency of the two-step solution and the iterative optimization algorithm in terms of the power consumption,the computation delay and the communication reliability.