eduzhai > Applied Sciences > Engineering >

Minimum Overhead Beamforming and Resource Allocation in D2D Edge Networks

  • king
  • (0) Download
  • 20210505
  • Save

... pages left unread,continue reading

Document pages: 15 pages

Abstract: Device-to-device (D2D) communications is expected to be a critical enabler ofdistributed computing in edge networks at scale. A key challenge in providingthis capability is the requirement for judicious management of theheterogeneous communication and computation resources that exist at the edge tomeet processing needs. In this paper, we develop an optimization methodologythat considers topology configuration jointly with device and network resourceallocation to minimize total D2D overhead, which we quantify in terms of timeand energy required for task processing. Variables in our model include taskassignment, CPU allocation, subchannel selection, and beamforming design formultiple input multiple output (MIMO) wireless devices. We propose two methodsto solve the resulting non-convex mixed integer program: semi-exhaustive searchoptimization, which represents a "best-effort " at obtaining the optimalsolution, and efficient alternate optimization, which is more computationallyefficient. As a component of these two methods, we develop a coordinatedbeamforming algorithm which we show obtains the optimal beamformer for a commonreceiver characteristic. Through numerical experiments, we find that ourmethodology yields substantial improvements in network overhead compared withlocal computation and partially optimized methods, which validates our jointoptimization approach. Further, we find that the efficient alternateoptimization scales well with the number of nodes, and thus can be a practicalsolution for D2D computing in large networks.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×