Article Preview
Top1. Introduction
Sensor nodes are densely deployed in the monitoring area to sense some physical phenomenon. There are various applications of WSN such as traffic monitoring, disaster management, fire detection, weather monitoring, underwater monitoring, underground monitoring for agricultural surveillance, security surveillance, etc. (Khoufi et al., 2017; Chelbi et al., 2018; Verma et al., 2016; Verma et al., 2017, Osamaa et al.2016). For such applications, various protocols have been developed to solve many research issues like energy efficiency, security, QoS etc. (Wang et al., 2018; Shahbazian et al., 2018; Mohamed et al., 2018; Dwivedi et al., 2012). This paper focuses on the energy efficient protocols. Some of these protocols are non-agent based (Kulik et al., 1999; Kulik et al., 2002; Intanagonwiwat et al., 2000; Heinzelman et al., 2000; Yu et al., 2001) while some are agent-based protocols (Gupta et al., 2012; Konstantopoulos et al., 2010; Chen et al., 2007; Chen et al., 2011). Agent based protocols are the protocols which use mobile agents. Mobile agent is a software program which migrates to sensor nodes for collecting data and it can move autonomously (Sasirekha et al., 2017; Jing et al., 2017; Vijayalakshmi et al., 2016). Some authors (Gupta et al., 2011; Chen et al., 2006) explained in their research that mobile agents help to increase power efficiency and reduce communication cost as well as network traffic. Existing researches also show that under certain conditions, agent based WSN protocols give better performance in terms of energy than non-agent-based or client server based WSN protocols (Gupta et al., 2011; Chen et al., 2009; Chen et al., 2011). In agent-based protocols, mobile agents are used to carry the processing code assigned by the sink for the different tasks. This processing code needs local processing at the sensors. Data aggregation can be implemented using these agents in which redundant data is removed. This helps in power saving of sensor batteries. Agent based protocols can use either single agent or multiple agents. There are some drawbacks of single agent-based protocols in large scale networks such as large delay, unbalanced load and insecurity of data loss with large accumulated size of agent. Simulation results show that the multi-agent-based protocols improve the energy efficiency and minimize the delay as compared to the single agent based schemes of similar computation complexity (Gupta et al., 2015; Pantazis et al., 2013). This paper focuses on the energy efficient multi agent-based data dissemination protocols of WSN. Some of the existing multi agent based protocols such as MMADIDD (Gupta et al., 2012) and TBID (Konstantopoulos et al., 2010) have shown that if multiple agents are dispatched for data gathering from center of the circular monitoring region to its various sub divisions individually then power efficiency can be improved. These protocols outperform in terms of energy efficiency at easy-to-reach areas but not suitable for hard-to-reach territories due to some limitations such as difficulties in placing sink at center of the monitoring area of hard-to-reach territories, improper selection of the type of sensor nodes and less efficient agent migration process etc. Therefore, research towards development of an energy aware data dissemination protocol of WSN for hard-to-reach territories is the demand of this generation. In this regard, a Multi Agent based energy and fault aware protocol for Hard-to-reach Territories (MAHT) is designed and presented in this paper. MAHT removes the limitations of the existing protocols. Selection of only source nodes for data dissemination and modified agent migration approach of MAHT improve the network lifetime and energy efficiency of the protocol. MAHT uses dynamic itinerary planning that makes this protocol fault tolerant. Simulation results describe that MAHT performs better than the existing protocols TBID and MMADIDD.