Computer Science Department Thesis Defense - Yashar Ghaemi

Please join the Computer Science Department for the upcoming thesis defense:

Presenter: Yashar Ghaemi

Thesis title: Efficient Routing Algorithms in Vehicular Ad Hoc Networks

Abstract: During the last decade, the research on Intelligent Transportation System (ITS) has improved exponentially in real-life scenarios to provide optimized transport network performance. It is a matter of importance that the emergency messages being delivered in a timely manner to prevent vehicular traffic problems. ITS system per se could be a part of a vehicular ad hoc network (VANET) which is an extension of the wireless network. In all sorts of wireless ad hoc networks, the network topology is subjected to change due to the mobility of network nodes; therefore, an existing explored route between two nodes could be demolished in a minor fraction of time. When it turns to the VANETs, the topology likely changes due to the high velocity of nodes. On the other hand, time is a crucial factor playing an important role in message handling between the network's nodes. In this regard, we propose a centralized ITS, and it uses a Multi-Path Routing Protocol (MPRP) that effectively identifies an optimized path for packet delivery to the destination vehicle with a minimal time delay. Our algorithm gives a higher priority to the alert messages compared to normal messages. As a result, our algorithm would realize two goals. Firstly, speed up the data transmission rate and deliver data packets, particularly warning messages, to the destination vehicle promptly and therefore avoids vehicular problems such as car accidents. Secondly, the MPRP algorithm reduces the data traffic load, particularly of the normal messages to alleviate the pressure on the network and therefore avoids network congestion and data collisions. This, in turn, lessens the packets' retransmissions. To demonstrate the effectiveness of the proposed protocol, MPRP has been compared with other protocols such as AOMDV, FF-AOMDV, EGSR, and QMR. The performance evaluation of MPRP took place on parameters such as simulation time, the number of nodes, maximum allowance mobility speed, variant number of application sinks (number of nodes sending packets), and variable packet sizes to calculate different QoS metrics like throughput, end-to-end delay, packet delivery ratio, packet loss ratio, and routing overhead. Simulation results demonstrate that our proposed protocol proves its excellent performance compared to other protocols. Moreover, during this study, we discovered several metrics impacting the performance of the network. The vehicular mobility speed has a crucial impact on the performance of the data communication through the VANETs. However, in VANETs, nodes' mobility speed could be higher than other forms of ad hoc network. The mobility of nodes is more predictable due to the roads' structure and constraints; therefore, mobility speed is not the only reason behind the topological changes. Nodes density has another influential factor on the network topology and scalability so that a highly dense environment is susceptible to having data collision and congestion. On the contrary, a lowly dense environment is vulnerable to having low nodes connectivity. Having network congestion itself is an essential factor in network performance. In this regard, we introduce a new fitness function (FFn) as the optimization base for the genetic algorithm (GA) and hence propose two mechanisms which are MPRP-FFn and MPRP-GA. The result of the proposed protocols has been compared with two recent protocols of TA-AOMDV and EHO-AOMDV as well as MPRP. We evaluate our protocols through QoS metrics, including throughput, packet loss ratio (PLR), end-to-end delay, routing overhead and energy consumption over various network configurations, including the simulation time, random loss ratio, mobility speed and the number of nodes. Our protocols demonstrate superiority compared to other methods.

Committee Members:

Dr. Hosam El-Ocla (supervisor, committee chair), Dr. Thiago E Alves de Oliveira, Dr. Dariush Ebrahimi (Thompson Rivers University)

Please contact grad.compsci@lakeheadu.ca for the Zoom link. Everyone is welcome.