Content caching with bi-level control for efficient IPTV content steaming service

With recent advance in broadband networking technologies, Internet Protocol Television (IPTV) has been the killer application of broadband Internet services. The major concern for designing IPTV service system is that it has limited content storage and network bandwidth. An efficient approach to address this concern is to deploy local content servers close to the consumers at the edge of network and to cache popular contents that are likely to be used in the near future. Content caching is a fundamental strategy for improving the performance and quality of service perceived by consumers. Frequently, the popularity distribution of video contents for video on demand service has been considered as Zipf distribution to evaluate caching strategies. But, this property is no more valid to on-demand content streaming service. Recently, many measurement studies have shown that the popularity follows Mandelbrot-Zipf distribution. In this paper, we therefore propose a prioritized dual caching algorithm which adaptively controls caches by considering the popularity nature for Mandelbrot-Zipf distribution. From trace-driven simulation experiments, we show that the proposed algorithm achieves relatively high performance improvement. In particular, the proposed algorithm is very effective when the cache size is relatively small. Finally, we evaluate how each control parameter in our proposed algorithm has an influence on the cache performance.