Computation Offloading Strategy Based on Joint Allocation in Mobile Device Cloud

Abstract

The contradiction between mobile terminal limited energy and users’ high requirements is more and more intensified, to provide satisfactory computation performance for real time applications, offloading workloads to Mobile Device Cloud (MDC) will be adopted. In this paper, we consider the problem of communication competition in the network, and put forward the offloading strategy based on the Joint Allocation Scheme (JAS) for lower latency. In the physical layer, the channel bandwidth allocation mode of OFDMA is used to make transmission rate realistic, combined with the task assignment, a delay model belongs to the NP problem can be established. We use the step-by-step solution, firstly consider the allocation of continuous task, and then use the simulated annealing algorithm (SAA) to get the optimal solution.