Energy efficient vehicular networks in a city environment

With the advent of real-time high bandwidth multimedia services, enhancing quality of service (QoS) in communication networks has been the prime focus of researchers. With environmental awareness becoming a global concern, the need to have energy efficiency in communication networks has intensified. Moreover, as the network size and the number of users increase, the introduction of energy efficient networks has become essential. Very little work has been carried out so far in vehicular communication networks for energy efficiency, even though their size and the number of users is equivalent to that of the cellular network. Provisioning multimedia services in vehicular networks is challenging due to the dynamic nature of the environment in which they operate. Analysing the performance of such systems from both QoS and energy perspectives redefines the problem. Therefore, there is a need to introduce systems which not only maintain QoS in these environments, but also save significant amounts of energy. Vehicular networks comprise intelligent vehicles fitted with an on-board unit (OBU) with wireless communication, sensing, and computing capabilities, in addition to fixed resources. Vehicular communication will play a key role in providing safety, security, and entertainment for drivers, passengers, and pedestrians in futuristic “smart cities.” This thesis studies the performance of city vehicular communication systems in terms of QoS and energy consumption. Initially, a city vehicular mobility simulator based on a 3×3 km2 Manhattan grid is developed which includes important traffic characteristics in a typical city such as vehicular flow and speed. Next, a vehicular ad hoc network (VANET) comprising three routing protocols, namely multihop (MH) routing, position based routing with most forward within radius (PRMFR), and position based routing with nearest forward progress (PRNFP), is developed in this environment. The performance of the aforementioned routing protocols is evaluated in terms of QoS. Additionally, two energy efficient routing protocols are proposed, namely single cluster-head (SCH) and double cluster-head (DCH) based routing. The performance of the SCH and DCH routing protocols from both QoS and energy perspectives is evaluated and compared with that of the MH routing, store-carry and forward (SCF) routing, and two pure vehicle-to-roadside (V2R) routing approaches. Moreover, an energy efficient content distribution network (CDN) for this environment is proposed. A mixed integer linear programming (MILP) model that optimises the number and locations of Wi-Fi enabled caching points (CPs) and cellular basestations (BSs) is developed, with the objective of minimising the total network power consumption while serving the total traffic at each hour of the day. The performance of the proposed energy efficient CDN under different scenarios and different power management mechanisms in terms of both QoS and power efficiency is evaluated. Since there is no mechanism in the MILP model to switch a CP, once installed, into low power state, an analytical queuing model is developed for the CP, where the CP sleeps (takes vacations) during its inactivity periods to save energy while maintaining the required QoS.