Mr Nickmohammad Balouchzehi is going to defense his Phd thesis on “Optimal Infrastructure Design and Infrastructure-based Service Discovery in Vehicular Networks” on Saturday January 02, 2016. The session will be held in Phd defense hall, Computer Engineering Department, Iran University of Science and Technology, Tehran, Iran.
Daily growth in number of cars causes some challenges in different fields including traffic jam, fuel consumption and environmental pollutions. Intelligent transfer systems (ITS) have been proposed as a solution to these challenges. One of the most important technologies in these systems is vehicular communication which tries to solve these challenges through providing a variety of safety, infotainment and traffic services. However, the efficiency of the vehicular networks in order to provide such services depends on the effectiveness of service discovery and announcement mechanisms. The efficiency of these mechanisms is strongly affected by vehicular networks partitioning in low penetration rate and low network capacity in congested zones. Hence, it is necessary to use infrastructure nodes in vehicular environments to improve the efficiency of these solutions. However, due to the high cost of RSU’s installation and maintenance, the minimum number of them should be placed in vehicular environment as they can optimized the placement and maintenance costs.
In this thesis, first, a Binary Integer Programming (BIP) model has been presented which took in consideration the characteristics and constraints of the urban environment and vehicular networks. In this study, the density parameter, out of all traffic parameters, has been considered in RSU’s placement. The other effective traffic parameter is traffic flow between the zones. Due to the some restrictions in applying this parameter to the BIP model, a graph-based model has been presented for the RSU placement which took in consideration the traffic parameters and other environmental and network parameters as well as QoS requirements of applications. Due to the high time complexity of the BIP solutions and for the purpose of using the presented models in the urban environments and in a large scale, some heuristics have been presented to solve the models with polynomial time complexities.
In the following, peer to peer systems and deployed infrastructure nodes are used to location discovery and efficient traffic information advertisement and discovery in urban scenario. In location discovery mechanisms, a structured P2P system is suited on deployed infrastructure nodes. Moreover, location information is cached in mobile nodes and used by them in location discovery. In traffic information discovery and advertisement, vehicular congestion can be avoided through choosing various routes and proper vehicle distribution. Also, the proper distribution of traffic packets among the areas with a good density, can avoid packet’s losing via middle nodes or their lost due to the partitioning. Finally due to the high costs of RSU installation and maintenance, a hybrid system has been presented to place minimum number of RSUs in the most optimal places for navigation of vehicles and distribution of traffic packets, also the vehicle nodes use as peer to peer layer nodes in the other key and effective zones.
Realistic simulation has been used in order to evaluate applied models for RSU placements as well as assessing the presented solutions for the discovery and announcement of the traffic services in the urban areas. This kind of simulation is based on simultaneous use of three simulators including traffic simulator SUMO, network simulator OMNeT and vehicular networks simulator Veins. These three simulators are connected to each other through the TCP sockets. Simulation has been done in four different urban maps and one highway map and it has been done based on real traffic traces and generated traffic which has been produced by Traffic Modeler tool. The simulation results show the proper efficiency of presented solutions in the various urban and highway maps.
Keywords: Optimal RSU Placement, Traffic Services Discovery, Peer to Peer Systems, Traffic Information Systems, Vehicular Networks
Phd cadidate: Nickmohammad Balouchzehi
Supervisor: Professor Dr.Fathy, Dr.Akbari .
Jury Committee: Professors Dr.Yzdani, Dr. Dehghan, Dr. JahedMotlagh,Dr. Berangi, Dr. Vahid Azhari
Time: Location: 09:00 AM, Saturday January 02, 2016, Phd defense hall, Computer Engineering Department, Iran University of Science and Technology, Tehran, Iran