Abstract
Reliability and energy efficiency are two important requirements of the data gathering process in wireless sensor networks. Accordingly, we propose a novel data gathering algorithm which meets these requirements. The proposed scheme categorizes the sensed data into valuable and normal data and handles each type of data based on its demands. The main requirement of valuable data is reliability. Thus, the adopted strategy to gather this type of data is to send several copies of data packets toward the sink. The rise of energy exhaustion in this scheme is tolerable. This is due to that, the valuable data is generated at a low rate. On the other hand, our main concern in gathering normal data is energy efficiency. As most of the sensed data is normal, an energy-efficient approach to gather normal data results in considerable energy conserving. Thus, we exploit clustering technique for normal data gathering. We also propose a lightweight intrusion detection system to detect malicious nodes. Simulation results and theoretical analysis confirm that our proposed algorithm provides reliability and energy efficiency at an acceptable level.