Energy Efficiency and Routing in Sensor Networks
Bilge Kartal Çetin · VBN Forskningsportal (Aalborg Universitet) · 2011
Wireless sensor networks (WSNs) consist of tiny devices, which have a battery, a sensor, a microprocessor and a radio transmitter component. Due to a large range of application area, performance metrics in sensor network are strictly application-specic. However, 'unattended operation of the network for long time' or 'long network lifetime' can be determined as a common performance requirement for the most of the applications. Energy depletion of the nodes can interrupt communication and, in a worse case, it could cause network partitioning which leads the interruption of monitoring. Unlike ad-hoc networks, recharging or replacing of the sen- sors battery may be inconvenient, or even impossible in some monitoring environments. Therefore, the key challenge in the design of wireless sen- sor network protocols is how to maximize the network lifetime, which is limited by battery energy in sensor nodes, while providing the application requirement. In sensor networks, there are two important energy consuming pro- cesses, the rst is transmission-reception phase and the second is listening the radio for any possible event. Therefore, there are two strategies for en- ergy saving. The rst is reducing the networking activities, which means less packet exchanging between the nodes or balanced load allocation. The second is avoiding the idle waiting time for packet arrivals. The rst strategy covers routing techniques for resource constraints networks and the second covers duty-cycling techniques. In this thesis, rstly mathematical characterization of networking in sensor network is studied by considering network dynamics and duty cy- cling of sensor nodes. Then energy eciency techniques in sensor network are investigated. Secondly, a mathematical tool in which maximum life- time routing can be investigated for dierent network parameters is de- veloped by considering a duty-cycling mechanism in the network. Upper bound on network lifetime is sought by considering idle and sleep mode energy consumption as well as energy consumption in transmission and reception for sensor networks. The solution of the developed optimization problems gives an analytical benchmark for designing of maximum lifetime routing algorithms by giving the most energy balanced trac allocation between the possible routes in the network by considering a duty-cycling mechanism. Lastly, an energy ecient routing protocol is proposed and evaluated by using the developed mathematical tool.