Task Offloading Of Internet Of Vehicles By Optimizing Delay And Resource Utilization

In order to enhance the increasingly congested traffic conditions,the technology of the IoV has come out.In the technology of the IoV,smart vehicle applications has spawned many tasks with large computational tasks and delay-sensitive tasks in a rapid growth,such as autonomous driving and real-time road conditions.The terminal aim is to facilitate our lives and become speedier.However,those promising applications usually need to process data,which has large quantities and require high computing power.However,the computing power of each vehicle has a circumscribe bound,it is tough for the vehicle itself to complete the requirements of the task fully.To cope with this challenge,existing papers propose that vehicle-to-all(V2X)communication is a promising technology that can support edge computing and transmission tasks across vehicles.Combined the mobile edge computing with the IoV,they make full use of their respective advantages to shorten the distance between the server and the vehicle,and provide data processing more effectively.By using V2 X communication technology,spectrum resources can used more effectively,energy consumption can abated markedly,and the quality of calculations can enhanced.The common vehicle-to-all(V2X)communication technologies in the vehicle-mounted MEC environment for task sharing are V2 V and V2 I.The existing mainstream technology is to offload vehicle-mounted assignments to edge base stations through V2 I technology for processing.Although edge base stations can handle tasks sent by vehicles,vehicle-to-infrastructure(V2I)communication has great limitations.To ensure communication quality,the scene is usually set up on city roads,and the base stations are evenly distributed on the roads usually.If the vehicle is not within range of the target base station,the delay will increase.The base station representing the edge suffers from low flexibility and low mobility.Currently,some scholars use vehicle-to-vehicle(V2V)communication technology in the unloading process,i.e.vehicles are multi-hop in a vehicle-to-vehicle manner.In order to optimize the offload problem,it transmits to a base station suitable for offload,but for remote locations and steep terrain,this method still has limitations.Considered to optimize the above problems better,and there are fewer base stations or remote areas,this paper proposes V2 V as the main technology and V2 I as a supplement technology.The method combines the two technology together which means that we can offload the tasks to the vehicle that has enough computing ability.If we cannot find the current vehicle for the V2 V to offload,then we can use V2 I communication technology to offload the task to the edge node.To select the best task offloading decision,this paper deal with the question we mentioned by put it into three steps: First,this paper write an algorithm for preliminary screening of the target vehicle according to the coordinates and speed of the vehicle.Then this paper uses the permutation and combination strategy based on the Cartesian product algorithm to obtain a collection of preliminary selection strategies.In the second step,this paper make a definition of a multi-target joint optimization problem for the computational offloading task,which is to minimizes delay and maximizes resource utilization.NSGA-Ⅲ is effective in solving jointly optimizing multiple target problems,so NSGA-Ⅲ can used to find the global optimal solution for the computational offloading method.First,our paper encodes the offloading strategy.Then in order to provide constraints for our paper,this paper needs to give a fitness function.After that,this paper uses the NSGA-Ⅲ genetic algorithm to perform crossover and mutation operations on our initial strategy to create brand new uninstall solution.After many iterations,the most desirable solution in the population is selected.The third step,this paper use the method SAW on the basis of the second step,referring to the two indicators of delay and resource utilization,weighting the offloading solution of the second step,after using the MCDM method,we normalize calculation results to get better downloads.Finally,the validity of simulation steps has confirmed.