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 missioncritical 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 singlehop clusters simultaneously after random deployment. Then, we prior-estimate that the number of cluster head candidates from n sensors is Θ (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.