Computational and communication resource allocation for mobile cooperative cloudlet computing systems

Abstract

Mobile users expect mobile device to work as static computers. However, the limited computational resource of mobile devices can not satisfy users' requirements. To solve this problem, a novel mobile cooperative cloudlet computing (M3C) model is studied in this paper which utilizes computational resource distributed in nearby cloudlet access points (APs) to shorten users' waiting time. A partial critical path (PCP) based workflow scheduling algorithm is proposed to assign tasks of mobile application workflow to APs, a dynamic programming based algorithm is proposed to select placements for APs, and a channel allocation algorithm is proposed to allocate wireless communication resource to radio links between APs. Simulation results show that the M3C model can achieve a shorter users' waiting time compared with the independent cloudlet computing (Icloudlet) model.