Routing Research And Node Design For Opportunistic Networks

Opportunistic network is a kind of self organizing network with delay tolerance. It does not need the complete path between the source node and the destination node, and only rely on the chance of meeting between nodes to realize the transmission of message. Because of its convenient, flexible, low cost and other characteristics, the opportunistic network would be a important supplement for the traditional wireless network in some situations. In order to meet the demand of the adaptive routing networking opportunities, some research and design are made which including the opportunistic network routing method based on ant colonyalgorithm (ant colony optimization, ACO) and neural network algorithm. Nodes with random behavior using ant colony algorithm while using neural network algorithm to deal with nodes with regularity behavior.The main research work of this paper is as follows:1. Research on how to combine ant colony algorithm and neural network algorithm with opportunity network routing, and design the overall process of the routing algorithm.2. Improve the traditional ant colony algorithm based on the characteristics of the opportunity network. A way to combined heuristic algorithm with pheromone is proposed to solve the contradiction between the convergence speed and the avoidance of local preferential solution. In the way of pheromone updating, it is proposed that using routing message with backtrace to carry on the back-track pheromone. In the aspect of node routing construction, a new method of node routing based on opportunistic network routing is proposed.3. In the neural network modeling, it is put forward that using the node’s historical transmission quality information instead of a single connected state information, to better adapt to the characteristics of the opportunistic network. At the same time, based on the middleware technology, a kind of load balancing technology in opportunistic node is proposed.4. Construct the corresponding simulation system based on the designed scheme. Build simulation framework, write the corresponding algorithm, routing model, data cache module and collect the corresponding experimental dataset. Simulations in four different scenarios of data set are made. By comparing the performance results with other routing algorithm like probabilistic routing algorithm and spray focus routing algorithm, algorithm in this paper has certain advantages in delivery success rate and node resources consumption while running on the experimental datasets with community characteristics.