Toggle Main Menu Toggle Search

Open Access padlockePrints

Jointly Optimized Energy-minimal Resource Allocation in Cache-enhanced Mobile Edge Computing Systems

Lookup NU author(s): Peng Liu


Full text for this publication is not currently held within this repository. Alternative links are provided below where available.


OAPA Mobile edge computing (MEC) has attracted extensive researches recently due to its ability to augment computational capabilities of mobile devices. This paper considers a cache-enhanced multiuser MEC system where the task can be cached in the MEC servers to avoid the transmission of duplicate data. To further improve the energy efficiency and satisfy the users’ requirement on delay, we jointly optimize caching, computation and communication resources in this system. The formulated problem is a mixed integer non-convex optimization problem that is very challenging to solve. We thus propose an efficient iterative algorithm by jointly applying the block coordinate descent and convex optimization techniques, which is guaranteed to converge to at least a suboptimal solution. Specifically, the formulated joint optimization problem is decomposed into two subproblems to respectively optimize caching policy and resource allocation, which are alternately optimized by convex optimization in each iteration. To further speed up the algorithm convergence, an efficient initialization scheme based on linear weighted method is proposed for caching policy. Extensive simulation results are provided to demonstrate that proposed jointly optimizing caching,computation and communication method can improve the energy efficiency with lower time cost as compared to other benchmark methods.

Publication metadata

Author(s): Liu P, Xu G, Yang K, Wang K, Meng X

Publication type: Article

Publication status: Published

Journal: IEEE Access

Year: 2018

Volume: 7

Pages: 3336-3347

Online publication date: 25/12/2018

Acceptance date: 13/12/2018

ISSN (electronic): 2169-3536

Publisher: IEEE


DOI: 10.1109/ACCESS.2018.2889815


Altmetrics provided by Altmetric