Dynamic Computation Offloading and Resource Allocation Over Mobile Edge Computing Networks With Energy Harvesting Capability Conference Paper uri icon

abstract

  • © 2018 IEEE. As an emerging and promising technique, the mobile edge computing (MEC) can significantly enhance the computational capability and save computing energy of mobiles, by offloading the computation-intensive tasks from the resource-constrained mobiles to the resource-rich MEC servers. However, since mobiles are generally energy constrained, mobile applications may still be interrupted when the energy of mobiles runs out. To overcome this challenge, we propose to integrate energy harvesting (EH) technique, which can enable mobiles to collect recyclable energy from ambient environments, into MEC, and develop the joint computation offloading and resource allocation scheme for the MEC system supporting multiple EH mobiles. In our considered scenario, each mobile first harvests energy from radio frequency (RF) signals emitted by a base station (BS) which is equipped with an MEC server, and then utilizes the harvested energy to execute its own task either locally at the mobile or by offloading to MEC. Moreover, our developed MEC system employs non- orthogonal multiple access (NOMA) so that multiple mobiles can utilize the same system subcarriers for task offloading to improve system performance. We first formulate the computation offloading and resource allocation problem of interest into an optimization problem, aiming to minimize the total task execution time of all mobiles under their strict timely-execution requirements. Then, we develop the joint computation offloading and resource allocation schemes, through which we can dynamically determine: 1) the energy harvesting time for mobiles; 2) the CPU clock frequencies of mobiles which intend local computing on their own; and 3) the set of mobiles which choose data offloading as well as the subcarriers and power allocations for these mobiles. Finally, we validate and evaluate the proposed offloading and resource allocation scheme through numerical analyses.

author list (cited authors)

  • Wang, F., & Zhang, X. i.

citation count

  • 10

publication date

  • May 2018

publisher