Summary: | 碩士 === 國立中正大學 === 通訊工程研究所 === 97 === Network lifetime is an essential system parameter to wireless sensor networks (WSNs) used in the safety-critical or high-reliable applications under limited energy constraints, such as physiological sensing and nuclear surveillance. Hard lifetime WSNs require every sensor node meet the network lifetime constraint, to avoid serious circumstances such as severe health conditions or finicial disasters. Soft lifetime WSNs only need to satisfy a certain network lifetime. Therefore, methods that are able to predict the lifetime of WSNs are very important. Using cluster architectures, we may divide the sensor nodes into many groups and reduce the
average power consumption and bandwidth usage of the sensors. However, not a single cluster head selection algorithm has had the ability to predict the network lifetime of WSNs. We propose a High Energy First (HEF) cluster head selection algorithm to achieve the goals mentioned above. The idea of the High Energy First (HEF) clustering algorithm and its variants are not new but this paper is the first time to formulate a clustering algorithm in such an analytical way to answer both optimality and predictability: (1) HEF is proved to be an optimal policy of all in terms of maximizing a hard network lifetime; and (2) Within an idealized framework,
theoretical bounds on the hard network lifetime feasibility test for the HEF algorithm are provided; and (3) HEF can calculate the network survival probability for soft lifetime WSNs.
|