Intelligent Routing Algorithms from Vehicle to Vehicle Vehicle to Infrastructure and Hybrid Communications in Vanets

Abstract

Vehicular Ad hoc Networks (VANETs) are Mobile Ad hoc Networks (MANETs) where newlinevehicles are nodes. Some characteristics differentiate VANETs from MANETs, such as newlinedynamic varying node mobility, periodic network partitioning, roadway restrictions, etc. newlineThe aim of VANETs is to provide people travelling on the road with ubiquitous newlinecommunication, to collect and relay information from all nearby vehicles in the event of newlineemergencies, such as severe traffic delays, collisions and weather disasters. Vehicles in newlineVANET have smart devices, which enable an Intelligent Transportation System (ITS). newlineThe goal of the ITS is to ensure driving ease, road safety and the dissemination of newlineupdated road information. newlineIt is possible to categorize VANET architecture as Vehicle-to-Vehicle (V2V), newlineVehicle-to-Infrastructure (V2I) and Hybrid. VANET raises several research issues such newlineas routing, network congestion, network management, information management, privacy newlineand security. The design of effective routing algorithms for V2V, V2I and hybrid newlineVANET scenarios is one particular field that still faces important problems in VANET. newlineThis dissertation includes the design of VANET routing algorithms for V2V, V2I newlineand hybrid scenario with emphasis on realistic environments such as variable vehicle newlinedensity, mobility, real time traffic scenario etc., using intelligent techniques i.e., cognitive newlineagents, machine learning algorithm and knowledge based representation. newlineIn the first objective, routing protocol for V2V communication in VANETs is newlineproposed. V2V communication routing protocol is based on Belief-Desire-Intention newline(BDI) framework of cognitive agent approach. An agency is located at each vehicle. newlineSimulation results show the enhancement in network lifetime, packet delivery ratio and newlinedepletion in control overhead. newlineThe second objective is to design and develop efficient routing protocol for V2I newlinecommunication in VANETs using machine learning and multi-agent systems. Decision newlinetree algorithm is applied at vehicle to identify the event as c