Auction Based Secure Computation Offloading in Vehicular Networks

The cloud computing and edge computing technologies have provided new opportunities for vehicular task offloading in vehicular networks. However, the computing task of a vehicle can be offloaded to multiple edge servers. Therefore, how to select the optimal edge server to complete the offloading service is a challenge. In addition, some malicious edge servers may declare unreasonable prices to execute the offloading service. A market mechanism is thus required to constrain the bids of edge servers in the networks. Besides, the edge servers may provide the offloading services with low quality which will decrease the quality of experience (QoE) of the service requesters. In order to solve the above mentioned challenges, this chapter proposes an auction based task offloading framework to ensure that the computing tasks requested by vehicles can be safely offloaded and executed. First, to constrain the bids of the edge servers in the networks, a task offloading scheme based on the first price sealed auction is proposed for the edge servers which intend to join in the task offloading process. Then, considering the service quality of each edge server, the cloud server is used to evaluate the quality of the edge servers. By designing the security evaluation and prediction algorithm for edge servers based on transductive support vector machine (TSVM), the optimal edge server can be selected for the vehicle to execute the offloading task. With the edge-cloud networks, we evaluate the secure offloading strategy of the vehicle, where the simulation results show that the task offloading scheme proposed in this chapter has a higher efficiency than the conventional schemes.