Research On Distributed Networking Technology Of Self-Organizing Networks

The concept of self-organization was proposed by Ashby in 1947[1], and its basic idea is to rely on local interactions to achieve global order and structure [2]. The self-organization phenomenon exists widely in many kinds of subjects such as physics, chemistry, biology, sociology and so on. The most prominent feature of self-organization is as followings: (1) the distributed control: the self-organization network has no centralized control, and the individual doesn't know the global information; (2) the emerging features: the individual's simple and local behavior can produce complex global phenomenon [3]. Although there is still no commonly accepted exact definition of a self-organizing system, we refer to it as a set of entities that achieves a global system behavior via local interactions between its entities without centralized control [4].Therefore, in recent years, people tried to solve some problems in the field of technology and sociology through establishing the self-organization system model.With the rapid development of network technology, the network behavior has displayed the complexity and unpredictability. In this situation, the introduction of self-organizing is an inevitable trend. With the introduction of self-organizing to information networks, we can design the nodes' simple behavior, and use the interactions among many of nodes to obtain the complex and ideal network performance characteristics [5]. The model of self-organizing networks has been widely accepted, but at present, the networking technology of self-organizing networks is an unsolved and key problem, and we urgently need a kind of innovative ways to study the networking technology of self-organization networks.In this paper, we proposed two kinds of networking model in self-organization networks from the angle of distributed networking technology: 1, the networking model of self-organizing networks based on bouncing and clustering. This model uses the reflection theorem of light and the strategy of cluster merging to form a network in a distributed manner. This model has obvious advantages compared with the traditional mobility model in the aspect of networking speed; 2, the networking model of self-organizing networks based on cluster expansion. The nodes in this model spontaneously adjust their moving dirctions and positions according to the networking needs to make the radius of the cluster increase, so that more and more isolated nodes join in the network or this cluster and the other cluster merge into a new cluster, and finally achieve the distributed networking.This model has obvious advantages compared with the traditional mobility model in the aspect of networking speed and coverage area. We adopts the analogy method of cellular automaton to complete the simulation experiment, then analyze and evaluate the experiment results in the aspect of networking speed and coverage area.