A Collaborative Computation Offloading Scheme for 5G Heterogeneous Networks

Mobile Edge Computing enabled Macro Base Stations (MEC-MBSs) can supplement the ever-increasing functionality requirements of the Mobile Devices (MDs). However, unlike Remote Cloud (RC) centers, MEC-MBSs have limited resources and become overloaded if too many MDs offload their tasks simultaneously. Instead of deteriorating the users' experience by postponing or refusing the excessive offloaded requests, a cooperation strategy has been studied and a Collaborative Utility-Maximization Scheme (CUMS) is proposed in this paper. CUMS exploits the available free computational resources of the neighboring MEC-MBSs in the first tier and RC as a second tier to process the extra workloads. CUMS optimizes both task placement policy and radio resource allocation process concurrently without violating the task completion deadlines. Overall system utility maximization problem is formulated to minimize the time and energy consumption for all the MDs and solved by CUMS in three phases: categorization of MDs, preferential resource assignment and iterative resources allocation respectively. In addition, a thorough simulation study has also been conducted and the superiority of the proposed scheme is verified in different system configurations.