| The hybrid architecture streaming media content delivery system which integrated with the content delivery network and peer-to-peer technology is recently proposed as an advanced solution in content delivery area, it is designed for optimize both the topology and distribution stradegy,which is increased the overall system capacity and extensibility effectively.Because of the huge economic benefits it can be brought in by distributing the network bandwidth of the edge streaming media VoD server of the system reasonablely,thus it is a hot topic in the industry and academic study fields since the bandwidth allocation issues has been proposed.Based on the integer planning model of server bandwidth distribution, the issue that how to maximize the overall system service capacity under the constraint of streaming data channel and global bandwidth in hybrid architectrue streaming media content delivery system is studied in this dissertation, and the Genetic Algorithm based Bandwidth Distribution Algorithmis(GA-BDA)is proposed. The server bandwidth distribution model that proposed in this dissertation is on the basis of theoretical analysis on VoD characteristics of different popularity during the CDN and P2P collaboration delivery process, take the streaming data channel which was occupied for providing VoD service as the server bandwidth allocation units,bandwidth allocation problem is transformed into integer programming problems, and use genetic algorithms to solve it.The GA-BDA is designed adopting the integer coding mechanism, as well as the roulette selection, mixed-crossover and adjacent mutation operation, the objective is to improve the overall system service accommodate ratio.In the simulations, the overall system service capacity and service accommodate ratio is adopted as the performance indicators for the analysis, simulation results show that the GA-BDA reached its designed purpose, the convergence speed and solution quality have reached a good performance. |
PDF Full Text Request