Article Preview
TopIntroduction
A MANET is formed by a cluster of mobile hosts without any predefined infrastructure. Hosts in a MANET can roam and communicate with other at will. The main advantage of a MANET is that it can be rapidly deployed since no base station (BS) or fixed network infrastructure is required. MANETs can be applied where redeployment of network infrastructure is difficult or impossible (for example, in fleets on the oceans, armies on the march, natural disasters, battlefields, festival grounds, and historic sites, etc.).
A large number of routing protocols have been developed for MANETs (Broch et al., 1998; Corson et al., 1995; Johnson & Maltz, 1996; Perkin & Bhagwat, 1994; Dube et al., 1997; Buttyan & Hubaux, 2003; Siva Kumar et al., 2003; Chen et al., 1999; Liao et al., 2001; Chang & Chang, 2001; Doshi et al., 2001; Song et al., 2004; Chang & Tu, 2002) which are characterized by unpredictable network topology changes, high degree of mobility, energy-constrained mobile nodes, bandwidth-constrained intermittent connection, and memory constrained. The routing problem has been well researched in infrastructure based wireless networks with the following major goals: efficient detection and adaptation to the network topology, scalability, and convergence. Since MANETs are inherently more dynamic, so these solutions can not be directly applied to solve the routing problems in MANETs.
Use of Agents is an attractive choice for routing decisions in heterogeneous networks over the past few years (Patel & Garg, 2001; Patel & Garg, 2004; Neeraj et al., 2007; Manvi et al., 2008). Agent based routing scheme offers flexibility, scalability, efficiency, adaptability and maintainability because of the following reasons: Agents are flexible to accommodate varieties of services to facilitate routing in MANETs (Patel & Garg, 2001; Patel & Garg, 2004). Also MAs are scalable to any size of a network since link and path stability are computed in a distributed manner (Manvi et al., 2008). Moreover they increase the packet delivery ratio and decrease packet delivery latency with increase in the mobility. So the efficiency of the network will not decrease even in the case of high mobility (Patel & Garg, 2001; Patel & Garg, 2004; Manvi et al., 2008). Agents are capable of adapting themselves by change in network conditions (Patel & Garg, 2001; Patel & Garg, 2004). We can easily debug the agent components and also replace the old agent components with new ones without affecting the other components, because MAs can be coded to perform aggregate tasks such as identification of reliable and intermediate nodes, multicast routes set up, multicast route management, etc.