Joint Computation Offloading and Radio Resource Allocations in Small-Cell Wireless Cellular Networks

• Published in: IEEE transactions on green communications and networking Vol. 4; no. 3; pp. 745 - 758
• Main Authors: Chen, Hong, Zhao, Dongmei, Chen, Qianbin, Chai, Rong
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.09.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Base stations; Cellular communication; cloudlets; Computation offloading; Decisions; Delays; Electronic devices; Energy conservation; Energy consumption; mobile computation offloading; Mobile computing; Network latency; Nonlinear programming; Radio equipment; radio resource allocations; Resource allocation; Resource management; Servers; Small cell cellular networks; Task analysis; Wireless communication; Wireless networks
• ISSN: 2473-2400; 2473-2400
• DOI: 10.1109/TGCN.2020.2976932

Mobile computation offloading (MCO) is a technique that can help reduce energy consumption of mobile devices (MDs) by offloading their tasks to more powerful devices for execution. In MCO, the offloading decision for a task depends on availability of both communication and computation resource. In small cell cellular networks, cloudlet servers are usually co-located with the small base stations (SBSs). As a result, offloading decisions of the MDs are coupled with SBS associations, while strong overlapping coverage between the SBSs can result in complicated interference conditions in wireless transmissions that affect the offloading performance. In this paper, offloading decisions and SBS associations are jointly optimized with transmission power and channel assignments in a small cell cellular network. The objective is to minimize the total energy consumption of all MDs, subject to task's latency constraints. The problem is first formulated as a mixed binary nonlinear programming problem, then transformed and solved using the general bender decomposition (GBD). A heuristic solution is proposed that recursively allows more MDs to make offloading decisions based on the current transmission conditions. Compared to using GBD, this solution results in much lower worst-case complexity, while achieving good energy performance for a wide range of system parameters.