Accelerating Initialization for Sensor Networks

Initialization is an indispensable process for wireless sensor networks since newly deployed sensors lack a reliable infrastructure for communication. Many efforts have been made to address the problem on initialization. However, most of them fall into the category of energy efficiency. In mission-critical scenarios, it is vital to reduce the initialization time in order to gather information as early as possible. In this paper, we study a new problem of accelerating initialization. First, we provide a novel parallel initialization mechanism that selects cluster heads before deployment and organizes single-hop clusters simultaneously after random deployment. Then, we prior-estimate that the number of cluster head candidates from n sensors is ¿(log n). Finally, simulation results show that these cluster head candidates can achieve the parallel initialization bounded by 16 time slots even for a large n.