Multi-agent based context aware information gathering for agriculture using Wireless Multimedia Sensor Networks

Abstract

Wireless multimedia sensor networks (WMSN) can be used in a wide range of applications such as monitoring agriculture, infrastructures, military operations, disaster management and so on. Energy conservation is a major concern in WMSN applications. This paper proposes a multi-agent based context-aware information gathering using WMSN for monitoring agriculture. Three kinds of contexts are considered in this paper such as detecting an emergency, temporal and computational contexts for detection of diseased plants, weeds, fire and interpret the soil fertility based on the soil parameters. This work considers contexts driven by a sensor node. Whenever the context is detected the information will be sent to the sink node. The proposed scheme works as follows: Every sensor senses the information and updates the node knowledge base. Based on the sensed information node interprets the context such as disease affected plants, soil fertility, fire, and growth of weeds. The sensor nodes begin to transmit the stored information to the cluster heads with the help of Path Finding Agent (PFA). Cluster heads aggregate the information received by the sensor nodes in the field before sending this information with the help of Querry Agent (QA) to the sink node. At the sink node all the information will be sent to the end-user, but in case of the fire detection, the immediate action will be taken by the sink node itself to turn on the sprinklers. Once the sensor finishes the assigned task (sensing, communicating) then automatically it goes into sleep mode. To detect plant disease and weeds, content-based image retrieval is used to compare with the healthy or useful plant images respectively. For performance analysis, the proposed scheme is simulated using NS2. Some of the performance parameters considered in this work are context detection time, delay, fusion time and energy consumption. Keywords: Wireless multimedia sensor networks, Context-aware computing, Agent technology, Content-based image retrieval