Mobile Edge Offloading Mechanism For Proof Of Work In Blockchain

With the advances of Internet technology,mobile devices have become an indispensable tool for our work and life,thus ensuring data security of mobile applications has become an urgent problem to be solved.As an effective security technology,blockchain has become a feasible solution.Due to hardware limitations,mobile devices cannot afford the computing resources required for proof of work in blockchain.The mobile edge computing is proposed to solve this problem.Therefore,the thesis mainly focuses on designing an effective offloading mechanism which can allocate heterogeneous edge computing resources,encourage edge service providers and mobile devices to participate in resources transactions.The optimization objective of this mechanism is to maximize the total social welfare of participants.This thesis proposes a combinatorial double auction model for edge offloading to allocate resources based on group buying.Group buying mechanism attracts more mobile devices to buy edge computing resources to solve the proof of work puzzle,then the utilization of edge servers can also be improved.In a given wireless network,the combinatorial double auction model for allocating resources based on group buying is as follows.First,miners and edge servers submit their bids to the intermediate auctioneer.Second,the intermediate auctioneer allocates edge computing resources for miners in turn based on the bids of edge servers and miners under resource constraints,and calculates the discount price according to the group buying rules.Third,the edge server provides computing resources for miners to support mining task after the miners pay the edge server for the computing service.This thesis designs two greedy algorithms based on different group buying rules to allocate resources to ensure that miners and edge service providers with higher bids will be allocated preferentially.The first greedy allocation algorithm is based on a step function,in which only if a miner joining the group-buying can make the group size reach the step value,the miner can trigger a discount upgrade.The second greedy allocation algorithm is based on a piecewise continuous function,in which any miner who joins the group-buying can trigger a discount upgrade.This paper also designs a pricing algorithm based on the Vickrey-ClarkeGroves auction to determine the winner's charges and payments.Furthermore,a customed Tabu search algorithm has been devised to optimize the greedy algorithm based on a step function.Theoretical analysis proves that the designed algorithms are computationally efficient and satisfies the three economic properties which are budget balance,individual rationality and truthfulness respectively.The simulation compares the performance of the algorithms with existing benchmark algorithms.In addition,the experiments also explore the influences of neighborhood size and the number of iterations for Tabu algorithm on the total social utilities.The experimental results show that for the case of 500 miners in the auction,the total utility increases by 6 times,and utilization ratio of edge servers increases by 1.5 times,in comparison to the existing benchmark approaches accordingly.