Complex Communication Network Structure And Load Balancing Of Smart Distribution Grid

High-speed, bi-directional, integrated communication system is the foundation of smart distribution and utilization grid to achieve the dynamic real-time information and power interactive. The weakness of smart distribution grid is the access network. There is no one single network can meet the industrial demand of the service quality for this part communication business. Therefore, in order to ensure the safe and reliable of operation smart distribution grid, ensure the real-time reliable transmission of distribution business, there is a need for a comprehensive, integrated smart distribution grid.Based on the needs and current situation of the smart distribution grid, the existing communication technology is analyzed, a complex communication network structure is given according to the service need of smart distribution and utilization grid. This structure is based on the industrial Ethernet access and the heterogeneous wireless network. The power line carrier communication and wireless sensor network are adopted as necessary supplements to achieve a smart distribution grid. It can satisfy the various stages demands of the development of smart distribution grid for communication network and provide network infrastructure for the further development and implementation of the smart distribution part.The basic principles and characteristics of load balancing in heterogeneous wireless network under the smart distribution grid environment is studied in this paper; strengths and weaknesses of the existing load balancing algorithms are analyzed; and the improvement strategy for the load balancing is discussed. In order to reduce the average blocking probability of the network and balance the service load between cells, this paper introduced the support vector machine theory in load balancing process. To solve the problem of existing load balancing algorithms for heterogeneous wireless networks that failed to take full advantage of the resources in the entire system, and that the judgment indicators are too simple, a load balancing algorithm based on decision hyperplane in multi-dimensional decision space is proposed. Firstly a decision hyperplane is trained using the load indicators, then load levels are distinguished according to the distance between the points and the hyperplane, thus control the access of new business in order to achieve overall system load balancing. Simulation results show that the algorithm can decrease the system blocking probability while ensuring the resource utilization of the system, meanwhile, load balance of every cell of the heterogeneous wireless network is effectively realized.