Author(s): Debraj De, Shaojie Tang, WenZhan Song, Diane Cook, Sajal Das
In this paper we present ActSee, an activity-aware radio duty-cycling protocol that utilizes the learned activity pattern information to intelligently adjust radio duty-cycles. The goal is to minimize data delivery latency and maximize throughput, while still conserving energy in the network. Based on the collected intelligence about activity transition patterns in form of Activity Transition Probability Graph (AT P G), ActSee solves a linear programming problem to select the duty cycles for a set of nodes. The non-uniform and dynamic duty cycle assignments are distributed by the active node(s) in a defined neighborhood. As a case study we have evaluated ActSee using the activity pattern information from real Smart Home testbed CASAS . The activity events are generated based on that pattern, in the form of AT P G. Experimental results from both real sensor network testbed and sensor network simulator TOSSIM validate that our proposed activity-aware radio duty cycling protocol ActSee outperforms the existing designs (traditional MAC protocols, uniform duty cycling, reactive duty cycling) in terms of: data delivery latency, throughput and network lifetime.