Efficient Information Delivery Using Network Coding

Description

With the popularity of powerful mobile devices such as smartphones, mobile users can now enjoy a wide range of information services such as content distribution and file download, which become key drivers of the mobile market. For instance, it is common to find people around to download and store information such as digital newspapers into their smartphones for online or offline reading. It is not hard to envisage a growing demand for efficient information dissemination provided by these data services in many advanced mobile and wireless applications. The strength of data broadcast for information dissemination is its ability to answer multiple queries for the same data in a single transmission through a shared communication channel. Recently, many effective data scheduling algorithms have been proposed in the literature, each excels in achieving its targeted application-specific performance objective. Thereafter, no significant breakthrough has been achieved because the performance is bounded by the fact that only clients requesting the same data can be served at a time by transmitting the requested data on the shared channel. In this research, we propose a generalized encoding framework with caching strategies to incorporate network coding into data broadcast systems. With network coding, clients requesting different data can be served simultaneously by mixing multiple data together. This not only significantly reduces bandwidth consumption, but also improves system performance. On the server side, the encoding framework integrates network coding with existing scheduling algorithms to exploit the synergy between data scheduling and network coding in order to maximize resource efficiency without jeopardizing the original performance criteria stipulated by the application requirements. On the client side, a network coding-cognizant caching strategy can improve data availability by increasing the chance for clients to decode data of interest. Therefore, requests can be served locally without sending explicit requests to the server via the slow communication channel. This improves the request response time and reduces workload at the server. More importantly, this saves bandwidth and the limited client resources. We speculate a non-incremental improvement in terms of both system performance and resource efficiency in data broadcast systems with our proposed encoding framework with caching strategies. By enabling efficient information dissemination, many emerging wireless and mobile applications can provide better quality of services guarantee to the ever-increasing number of sophisticated users, which is their key to success.