A dynamic popularity-based partial caching scheme for video on demand service in IPTV networks

Caching video objects closer to the users in delivery of on-demand video services in IPTV networks reduces the load on the network and improves the latency in video delivery. Partial caching of the video objects is attractive due to the space constraints of the cache and also due to the fact that some parts of the video might be more popular than the others. However, fixed segment-based caching of videos does not take into account the changing popularity of the segments and the changes in the viewing patterns of the users. In this work, we propose a partial caching strategy that considers the changes in the popularity of the segments over time and the access patterns of the users to compute the utility of the objects in the cache. We also propose to partition the cache to avoid the eviction of the popular objects (those not accessed frequently) by the unpopular ones which are accessed with higher frequency. We measured the popularity distribution and ageing of popularity from two online datasets and use the parameters in simulations. Our simulation results show that the proposed caching scheme improves the byte hit ratio when compared to the LRU caching scheme both for static and dynamic object pools and ageing of popularity.