Research On Relay Selection And Computing Offloading Strategy For Parked Vehicle Assisted Vehicular Networks

With the rapid development of automobile industry and communication technology,vehicular networks which connects all automobiles through wireless network has emerged and gradually becomes the research hotspot.The main function of the traditional vehicular network is to ensure the information communication between vehicles.With the development of on-board applications such as automatic driving and the arrival of the era of big data,there are a lot of computing requirements which are strict on time delay in vehicular networks.The traditional vehicular networks can not meet these requirements,therefore the new vehicular networks based on mobile edge computing(MEC)has arisen.Based on the characteristics that there are a large number of parked vehicles in the city and most of them are parked on both sides of the road,some studies have considered to introduce parked vehicle resources into vehicular networks.However,there are still some disadvantages in the current research on this problem,including ignoring the departure behavior and energy-limited problem of parked vehicles,lack of energy-saving strategy while using them as communication resources,and lack of appropriate computing offload strategy while using them as computing resources.Therefore,this paper makes an in-depth analysis of the problems above and puts forward corresponding solutions,the main contributions and innovations include:(1)While introducing parked vehicles into the traditional vehicular networks as communication resources,we firstly define four modes for parked vehicles,and they can dynamically adjust their modes in order to save energy.Secondly,we propose a selection algorithm for parked vehicles as communication relay,which objective function includes vehicle departure probability,remaining power,communication coverage and other factors.Finally,we compare the algorithm with greedy algorithm,genetic algorithm and other classical algorithms.The result shows that the proposed algorithm can achieve the optimal energy-saving effect on the premise of ensuring the connectivity of the vehicular networks.(2)While introducing parked vehicles into MEC-based vehicular networks as computing resources,we propose a cooperative computing offload strategy in order to minimize the total delay of offloading.We take the computing resources of local vehicles,MEC servers,idle mobile vehicles and idle parked vehicles into consideration and design a offloading parked vehicle selection strategy.Then we propose two cooperative offloading modes according to the offloading order to fit the different real-world scenarios.A distributed cooperative routing algorithm for computing offloading between moving vehicles is also proposed to find the optimal offloading route between moving vehicles according to the task size.According to the simulation result,the proposed cooperative offloading strategy can reduce the time delay of accomplishing computation tasks effectively.