eduzhai > Applied Sciences > Engineering >

Deep Learning for Wireless Coded Caching with Unknown and Time-Variant Content Popularity

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 12 pages

Abstract: Coded caching is effective in leveraging the accumulated storage size inwireless networks by distributing different coded segments of each file inmultiple cache nodes. This paper aims to find a wireless coded caching policyto minimize the total discounted network cost, which involves both transmissiondelay and cache replacement cost, using tools from deep learning. The problemis known to be challenging due to the unknown, time-variant content popularityas well as the continuous, high-dimensional action space. We first propose aclustering based long short-term memory (C-LTSM) approach to predict the numberof content requests using historical request information. This approachexploits the correlation of the historical request information betweendifferent files through clustering. Based on the predicted results, we thenpropose a supervised deep deterministic policy gradient (SDDPG) approach. Thisapproach, on one hand, can learn the caching policy in continuous action spaceby using the actor-critic architecture. On the other hand, it accelerates thelearning process by pre-training the actor network based on the solution of anapproximate problem that minimizes the per-slot cost. Real-world trace-basednumerical results show that the proposed prediction and caching policy usingdeep learning outperform the considered existing methods.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×