Task Offloading Technology For Mobile Blockchain In Edge Computing

Blockchain technique has been applied in many fields such as electronic currency,smart grid and Internet of Things,as its security characteristics of decentralization,traceability,non-tamperability and etc.However,Proof-of-Work(Po W),the most widely used consensus mechanism in blockchain applications,is still being disputed because of its high computing resource requirements.Meanwhile,this high computing consuming characteristic has become an obstacle to the flourish o f blockchain on mobile devices.In order to solve this problem,mobile edge computing technique is utilized by researchers to offload the computation tasks of Po W,and maintain the blockchain applications in mobile environment.But most of the existing works deviate from the reality as they assume that a single mobile user can purchase all resources from the edge server,and ignore the incentive strategy for the participants.With regard to the problems mentioned above,a three-stage auction task offloading model is proposed for mobile blockchain in this thesis.Specifically,we model the block generation process in blockchain following the Poisson distribution,and formalize the expected benefits for each mobile user.Then the problem of maximizing social w elfare is formalized in mathematics,and social benefits are introduced to quantify it.The proposed method eases the shortage of computing resources by offloading the computation-intensive tasks of mobile users to the edge servers.Edge server is incentiv ized to share resources into the groups of mobile users instead of one user by implementing the auction and group-buying mechanism,in which ensured that mobile users can afford resource costs.In order to match mobile users with edge servers,a potential winner determination algorithm is designed in the first stage.The algorithm divides mobile users into several groups,and then selects different mobile users from each group according to the preference of users.In addition,a winner potential payment determination algorithm based on the Vickrey-Clarke-Groves auction is designed to determine the potential payment of mobile users and the bids of base stations for the edge servers.In the second stage,a matching algorithm between base station and edge server is designed.We match base station with edge server based on the bids and resource information,and calculate the actual payment of the base station.In the third stage,the cost settlement is completed based on the matching of the first two stages,and then the edge server is assigned to the corresponding group.Finally,the mobile user offloads the task to the edge server for execution.In this thesis,theoretical analysis shows that the proposed scheme is qualified for the three economic properties of truthfulness,individual rationality,and budget balance.Meanwhile,it also shows that the proposed algorithm is efficient,and it is an approximation algorithm with approximation ratio of 1-?.Finally,simulation experiments are conducted to investigate the impact of blockchain parameters on social welfare.Experimental results show that the proposed mechanism outperforms other three existing works by 33.78%,21.84% and 19.69% on average,respectively,for 1000 mobile users in terms of social welfare.