Energy efficiency in ad-hoc wireless networks

In ad-hoc wireless networks, nodes are typically battery-powered, therefore energy limitations are among the critical constraints in ad-hoc wireless networks' development. The approaches investigated in this thesis to achieve energy efficient performance in wireless networks can be grouped into three main categories. 1. Each wireless network node has four energy consumption states: transmitting, receiving, listening and sleeping states. The power consumed in the listening state is less than the power consumed in the transmitting and receiving states, but significantly greater than that in the sleeping state. Energy efficiency is achieved if as many nodes as possible are put into the sleeping states. 2) Since energy is consumed for transmission nonlinearly in terms of the transmission range, transmission range adjustment is another energy saving approach. In this work, the optimal transmission range is derived and applied to achieve energy efficient performance in a number of scenerios. 3) Since energy can be saved properly arranging the communication algorithms, network topology management or network routing is the third approach which can be utilised in combination with the above two approaches. In this work, Geographical Adaptive Fidelity (GAF) algorithms, clustering algorithms and Geographic Routing (GR) algorithms are all utilised to reduce the energy consumption of wireless networks, such as Sensor Networks and Vehicular Networks. These three approaches are used in this work to reduce the energy consumption of wireless networks. With the GAF algorithm. We derived the optimal transmission range and optimal grid size in both linear and rectangular networks and as a result we show how the network energy consumptions can be reduced and how the network lifetime can be prolonged. With Geographic Routing algorithms the author proposed the Optimal Range Forward (ORF) algorithm and Optimal Forward with Energy Balance (OFEB) algorithm to reduce the energy consumption and to prolong the network lifetime. The results show that compared to the traditional GR algorithms (Most Forward within Radius, Nearest Forward Progress), the network lifetime is prolonged. Other approaches have also been considered to improve the networks's energy efficient operation utilising Genetic Algorithms to find the optimal size of the grid or cluster. Furthermore realistic physical layer models, Rayleigh fading and LogNormal fading, are considered in evaluating energy efficiency in a realistic network environment.