Achieving Energy Efficiency for Bursty Traffic in Wireless Sensor Networks

Media Access Control (MAC) layer has been regarded as the crucial component for optimizing energy performance in Low Power Devices and Wireless Sensor Networks (WSN) operating in Internet of Things (IoT) environment. As evident from the literature, Dynamic Duty-Cycling (DDC) and Dynamic Channel Polling (DCP) are major schemes deployed at the MAC layer targeted at achieving energy efficiency. This paper compares the energy and delay performance of these two schemes for the bursty traffic applications. The protocols Adaptive and Dynamic Polling-MAC (ADP-MAC) and Traffic Auto-ADaptive mechanism, T-AAD have been selected from each category of DDC and DCP respectively. Mica2 sensor motes and Avrora emulator have been used for conducting experiments. Effect of varying burst generation intervals and wake up duration of nodes have been studied over the energy and delay performance. ADP-MAC has shown to outperform T-AAD due to having efficient match between the channel polling and packet generation instants. Although, T-AAD adjusts the duty-cycle in accordance with the traffic pattern, the preamble transmission cost and delay still remains significantly higher as compared to ADP-MAC. Hence, it has been revealed that DDC technique outperforms DCP both in terms of energy and delay, when operating under bursty traffic.