Intelligent Routing for Opportunistic Networks Based on Distributed Learning Automata
Chamran Asgari, Ahmad Zareie, Rohollah Rahmati Torkashvand · 2013
Opportunistic Networks are indeed typical MANET networks. There are mobile nodes in these networks; moreover, there is no available information about network topology; however, network topology is highly dynamic. This type of networks is highly discrete. There may be no direct connection - or even multi-step connection - between source and destination nodes when a message is sent to destination node from source node. In these networks, although there is no connection route between destination and source, network should be able to transmit messages from source to destination. In fact, routing is done in a dynamic way. In this article, we present a novel approach for intelligent routing and transmitting messages from sources to various destinations based on distributed learning automata. This approach aims to find the best carriers which transmit messages. A good carrier should considerably reduce consumption resources, messages duplication and communication latency. Simulation results by NS2 software and comparing them with HIBOP andEPIDEMIC algorithms implied that this approach is more efficient than other presented ones.