The Key Technology Research, Video-on-demand System

This paper focuses on the research of Key technologies used in Video-On-Demand. Through investigations of existent research productions and documents, the development process and the mutuality of related important algorithms in this field is tided up and presented in this paper, also with the algorithm performance analyses together. Following algorithm analyzing, integrating and improving work is accomplished: Base on the research of classical Interval Caching algorithm, a novel Cache Management algorithm, called Data Pre-fetching Based on Read-Write Semaphore, is presented in this paper; To improve the performance of Stream Tapping algorithm, an new algorithm, called Partitioned Stream Tapping, is presented. And the simulation result show that the performance scale improved is almost 50%; By integrating two type of VOD server-side Multicast algorithms: FCFS Batching Multicast algorithm and Patching Multicast algorithm, an new algorithm, called FCFS Dynamic Optimized Patching Multicast algorithm is presented; In the same time, a VOD server-side Broadcast algorithm, called Cascade Broadcast algorithm is presented in order to overcome the robust shortcoming of Skyscraper Broadcast algorithm. For these two important algorithms, simulation results are showed in this paper. Additionally, A scheme of applying RTP in QoS network is also presented in this paper. And at last, an overall architecture and functional module design of a VOD prototype system is presented. In the prototype system, Data Pre-fetching Based on Read-Write Semaphore is used in cache management; Cascade Broadcast is used in service for demanding requests of hot movies; FCFS Dynamic Optimized Patching Multicast is used in service for lukewarm movies.