Dynamic tasks scheduling based on weighted bi-graph in Mobile Cloud Computing

Abstract

The development of mobile computing and rapid growth of mobile applications have promoted the emergence of several edge-based computing paradigms, i.e. Fog computing, edge Computing, which can serve as the middle layer of end-user devices and the powerful cloud. As an implementation subtype of Fog computing, Mobile Cloud Computing (MCC) aims at leveraging limited resources available at network edge in order to enrich mobile applications and promote end-users’ experience. Thus, an efficient resource allocation or scheduling scheme is vital for ensuring the effectiveness of MCC. In order to improve the performance of scheduling algorithm and promote its applications in practice, this paper proposes the Dynamic Tasks Scheduling algorithm based on Weighted Bi-graph model (DTSWB), which takes the dynamics of both tasks and providers into consideration. Specially, the scheduling problem is translated to be a maximum weighted bi-graph matching problem and an integer programming model is formulated. Then, the matching problem is solved by DTSWB, which mainly consists of four steps: state information collection of offloaded tasks and service providers, mapping relationship establishment, profit matrix determination and optimal matching based on Kuhn Munkras (KM). At last, the effectiveness and validity of the proposed algorithm are verified by a series of simulations and the simulation results show that DTSWB achieves better performance than existing scheduling algorithms.