Adaptive power management for mobile agent-based information retrieval

Mobile agent technology has been used in wireless distributed information retrieval for many years. However, the energy-efficiency issue in such systems remains unexplored. In this paper, we proposed an adaptive application-driven power management (AADPM) protocol for mobile agent-based applications within the IEEE 802.11b infrastructure WLAN environment. Our goal is to minimize energy while achieving low round trip time (RTT) delay. We discuss the protocol and evaluate its effectiveness using the network simulator NS2. We also draw horizontal comparisons among a variety of PM methods reported in the literature. Experimental results show that, compared to the power save mode supported by 802.11b, AADPM reduces the network interface card energy consumption by 65% while only introducing 3% RTT delay.