Building a case for FIPA Compliant Multiagent Approaches for Wireless Sensor Networks

Abstract

There have been some efforts to apply agent based approaches in Wireless Sensor Networks (WSNs) for different application scenarios in recent past. These efforts have met with varying levels of success so far. However, the efforts to assess the suitability of Foundation for Intelligent Physical Agents (FIPA) compliant multiagent solutions for WSNs are almost non-existent. In this paper we present a scenario from an industrial domain and asses the suitability of a FIPA compliant multiagent based solution for it. The scenario is based on a WSN deployment in an underground mine where it could be put to use for several purposes like hazardous gas plume detection and tracking, mine lighting control, Proximate Environment Monitoring (PEM), structural health monitoring of mine structure, to name a few. However, we only consider the PEM usage scenario in this paper.

Towards a FIPA compliant multiagent based middleware architecture for sensor networks

“…These approaches normally develop a semantic routing tree rooted at the base station that is used to disseminate queries in the network as well as to collect the result back at the root. Considering the mobile base station scenario like in an underground mine, the cost of developing a new semantic routing tree each time rooted at the current location of the user and associated communication costs, both in developing the routing tree and subsequently servicing the user's query, would be prohibitively high [3].…”

“…Similarly, in a chemical processing plant, workers should be aware of the environmental conditions around them and should immediately be informed in case any chemical spills happen around their workplace. Any WSN deployment in such industrial setups should meet this Proximate Environment Monitoring (PEM) [3] requirement of the users.…”

ACDMCP:An Adaptive and Completely Distributed Multi-Hop Clustering Protocol for Wireless Sensor Networks

“…Some other goals that are attributed to clustering include bandwidth reuse in the network, efficient data gathering and aggregation, target tracking, and supporting hierarchical routing techniques. Moreover, in some middleware approaches like [1,20] clustering not only helps in solving the scalability issue but it also helps in increasing the event reporting reliability. Additionally, clustering has also been proposed for solving security issues in mobile adhoc networks as suggested in [21].…”