The Application Of Cultural Algorithm In Load Balancing Of Hybrid P2P Network

Cultural algorithm is an evolutionary algorithm which is a simulation of biological evolution rules for solving the optimal approximate solution, and it is generated based on genetic algorithms. But the genetic algorithm population is only one storage space for the individual, while cultural algorithm adds a storage space named belief space on top of the genetic algorithm, this belief space stores the spatial knowledge of individuals which guides the evolution of individuals. With the guidance of knowledge, the individual in spatial quickly and accurately produce the optimal solution.Currently cultural algorithm has been applied to resource scheduling, function optimization, image processing, data mining, fault diagnosis and other fields.P2P technology is the abbreviation of the "peer" technology (Peer-to-Peer) and "peer to peer" technology (Point-to-Point). And compared with the traditional C/S structure, P2P structure will allocate the resources to each client to storing, effectively reduce the burden on the system server.With the P2P technology applied to the instant messaging education platform, you can resolve a lot of problems in non-P2P platforms such as real-time differential, the system response time longer, the server become a bottleneck because so many tasks.Because the P2P technology is in widespread using.Real-time communication education platform based on P2P technology uses the hybrid P2P structure that benefits both to prevent too large server load and guarantee stability of the system.Multimedia Technology Laboratory of North China University of Technology used the genetic algorithm to do the load balance scheduling in the platform and achieved certain results.However, because genetic in the evolution process has certain blindness, it caused the problem of slow convergence and falling into local optimum. And the result will be that the load of some servers is too large, and the other is be idled. With the growing number of users on the platform, this problem exposed more and more obviously. Firstly, this paper studies the culture algorithm in-depth, and analyzes the existing issues of the platform. Cultural algorithm is applied to the instant communication education platform.Finally, do the functional stress test on the platform and the results show that the server load value to become balance faster when cultural algorithm is applied to the platform.