Performance Evaluation And Optimization Of Video Content Distribution System For Telecom Operators

Intelligent industries develop fast. After the popularization of smart phones and tablets, smart TVs have begun to flourish. The original core function of TVs is to provide video broadcast service. And due to this feature and user habits, video business will be one of the most important businesses of smart TVs. In order to ensure the quality of service, most smart TV video businesses adopt Content Delivery Network (CDN) architecture. Compared to general CDNs, Telecom Operators CDNs do not have to lease bandwidth and can couple tightly with the underlying network, so they have advantages over general CDNs in both cost and performance. Thus Telecom Operators CDNs may become the main-stream video CDNs for smart TVs. But with the fast development of smart TVs, the content distribution systems expanding rapidly and the deployment and operating costs multiply, which extremely restricts the continuing development of Telecom Operators CDNs. Thus it is of great practical significance to study the cost control of Telecom Operators video CDNs.To meet the practical requirements, this paper evaluates and optimizes the performance of Telecom Operators video CDNs from the perspective of cost. It proposes a statistic-based distribution performance evaluation scheme. The scheme firstly analyzes actual conditions of Telecom Operators video CDNs, including working out the real network topology and building user behavior model; secondly proposes metrics and related parameters which operators really care about to measure the performance of Telecom Operators video content distribution; thirdly builds a theoretical model of content distribution strategies to qualitatively describe the relationships between performance metrics and related parameters; and then implements a large-scale simulation system of Telecom Operators video CDNs and conducts a large number of simulation experiments based on real topology and user behaviors. Experimental results show the quantitative relationships between performance metrics and related parameters, and get the optimized distribution strategy which minimizes the cost. This paper contributes to the Telecom operators’ decision-making in choosing distribution strategies and setting parameters.