eduzhai > Applied Sciences > Engineering >

Computation Offloading in Multi-Access Edge Computing Networks A Multi-Task Learning Approach

  • Save

... pages left unread,continue reading

Document pages: 13 pages

Abstract: Multi-access edge computing (MEC) has already shown the potential in enablingmobile devices to bear the computation-intensive applications by offloadingsome tasks to a nearby access point (AP) integrated with a MEC server (MES).However, due to the varying network conditions and limited computationresources of the MES, the offloading decisions taken by a mobile device and thecomputational resources allocated by the MES may not be efficiently achievedwith the lowest cost. In this paper, we propose a dynamic offloading frameworkfor the MEC network, in which the uplink non-orthogonal multiple access (NOMA)is used to enable multiple devices to upload their tasks via the same frequencyband. We formulate the offloading decision problem as a multiclassclassification problem and formulate the MES computational resource allocationproblem as a regression problem. Then a multi-task learning based feedforwardneural network (MTFNN) model is designed to jointly optimize the offloadingdecision and computational resource allocation. Numerical results illustratethat the proposed MTFNN outperforms the conventional optimization method interms of inference accuracy and computation complexity.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×