| Wireless Sensor Networks (WSNs), which are made by the convergence of sensor, micro-electro-mechanism system and network technologies, is a novel technology about acquiring and processing information. WSNs usually operate in unattended environment with limited energy. That is, the battery can not be recharged. Therefore, minimizing energy consumption, balancing the energy consumption of nodes, guarantying energy-efficiency and prolonging the network lifetime are key and most important requirements in the design of sensor network protocols and algorithms.Any practical clustering protocol in a Wireless Sensor Network (WSN) should maintain long connectivity, low and balanced energy consumption. In the existing clustering algorithms, the distance between regular node and cluster-head (CH) is normally used as a unique metric for regular nodes to choose CH to join. However, they ignore the impact of the location of the base station on network lifetime. With regard to the energy consumption, in this paper we study not only the impact of distance between regular node and CH but also the impact of distance between the CHs and the base station . In order to maintain long connectivity, low and balanced energy consumption of WSNs, a novel matrix Time is introduced and is used as a new criterion for optimally assigning the regular nodes into clusters. By using this metric and solving a Min/Min problem in the static scenario and a Max problem in the dynamic scenario, we are able to evaluate the network lifetime for each scenario. For the two scenarios, we formulate the calculation of network lifetime into an optimizing problem; we then give procedures on how to achieve its solution. Based on this, we present a distributed approach for network configuration in WSNs with the optimal energy-efficiency, which is termed as DOCE. Simulation results demonstrate that DOCE significantly outperforms LEACH and Ad hoc network design algorithm (ANDA). |
PDF Full Text Request