Wireless Self-organizing Network Traffic Load Balancing Algorithm,

Wireless self-organized network, so called Ad hoc network, is a kind of special mobile communication network, while all nodes in network are peer to each other and with no need of central control node. The network traffic load balance technique, is a kind of method to split the communication flow into several available paths equally, to improve the network throughput and utilize the network efficiency.The task of this paper is to implement a load balance algorithm in the leveled architecture tree topology wireless self-organized network. This special network structure consists of many self-organized sub-network. Different with traffic control strategy inside the network, the technique studied by this paper focuses on the administration of inter-network traffic, which balances the inter-network load on the path connecting different sub-networks. The technique dynamically modifies the distribution of traffic on the available paths, according to the load status of those paths in real time.The main task of this paper including:1. Introduce wireless self-organized network, especially the characteristic of leveled architecture tree topology wireless self-organized network, analyze routing protocol under this kind of network, particularly the principle of popular load balance techniques, and the suitable network environment of these techniques, and their shortages.2. Improve the load balance technique while taking the characteristic of the network into consideration, take a deep look of the load balance technique with sense of the load on the path, after analysis of the popular load sense technique, a detection packet load sense algorithm will be devised which is fitting the specific network architecture.3. Analyze the Linux platform's support of the load balance technique, and have a comprehensive discussion of the Linux system's routing subsystem, make a choice of suitable load balance algorithm, and the complete load balance with load sense technique will be presented.4. Implement the algorithm in the Linux platform, and validate the algorithm under the testing environment, analyze the effect to the network throughput and the network efficiency brought by the algorithm combining the testing result, and summarize the whole paper in the end.