Network Coded Compression- based Caching for Device-to-Device Communications
Christoforos Vlachos, Giorgos Chochlidakis, Janus Heide, Vasilis Friderikos · Research Portal (King's College London) · 2016
Content caching at the network level is attracting increasing interest as a technique that not only eases the data traffic generated to the base station (BS), but also reduces the access latency and the overall power consumption. By caching popular content along the network, user requests can be then satisfied by closely located mobile clients without the need to retrieve them from the source, which is assumed to be, in the best case scenario, the BS. However, due to the devices’ capacity restrictions, the involved users cannot cache a large number of files but only a subset of these. To this end, various cache management policies are applied in order to achieve effective content delivery among the network terminals. This work provides a network coded, cooperative cache management method that considers the compression of the existing files in a user’s cache to store more contents of interest. This technique is then compared with a number of popular cache orchestration methods where its performance in terms of traffic offloading from the BS is illustrated. Numerical results show that not only most of the content/data requests are being satisfied by local Device-to-Device (D2D) communications in the long run but also the compression-based technique can effectively alleviate the BS traffic in a shorter time window.