Efficient Implementation of Data Aggregation in WSNs by Mobile Agent Paradigm

In many sensor network applications viz., environmental monitoring, spatial exploration and battlefield surveillance, sensed data is aggregated and transmitted to the sinks for analysis. Thus, in-network data aggregation becomes an important technique in wireless sensor networks and has been well studied in recent years. In general, any sensory network suffers with two problems i.e., i) the time latencies of the existing scheduling algorithms are still high, and ii) the other one is that all the existing algorithms are centralized, which are inherently inefficient. Hence, in this paper, we present a distributed scheduling algorithm generating collision-free schedules with latency bound, based on TBID, which employs multiple Mobile Agents (MAs) for completing the aggregation task on WSNs. The algorithm not only determines the proper number of MAs for minimizing the total aggregation cost but also constructs low-cost itineraries for each of them. In this paper, we adopt TBID that improves upon NOID by following a more direct approach to the problem of determining low-cost MA itineraries. Specifically, based on an accurate formula for the total energy expended during MA migration, we follow a greedy-like approach for distributing SNs in multiple MA itineraries, and algorithm determines a spanning forest of trees in the network, calculates efficient tree traversal orders (itineraries), and eventually, assigns these itineraries to individual MAs.)