Evolutionary Computing Approach for Ad-Hoc Networks

Wireless ad-hoc networks are infrastructureless networks in which heterogeneous capable nodes assemble together and start communicating without any backbone support. These networks can be made truly dynamic and the nodes in these networks can move about freely while connecting and disconnecting with other nodes in the network. This property of ad-hoc networks to self-organize and communicate without any extrinsic support gives them tremendous flexibility and makes them perfect for applications such as emergencies, crisis-management, military and healthcare. For example, in case of emergencies such as earthquakes, often most of the existing wired network infrastructure gets destroyed. In addition, since most of the wireless networks such as GSM and IEEE 802.11 wireless LAN use wired infrastructure as their backbone, often they are also rendered useless. In such scenarios, ad-hoc networks can be deployed swiftly and used for coordinating relief and rescue operations. Ad-hoc networks can be used for communication between various stations in the battle-field, where setting up a wired or an infrastructure-based network is often considered impractical. Though a lot of research has been done on ad-hoc networks, a lot of problems such as security, qualityof- service (QoS) and multicasting need to be addressed satisfactorily before ad-hoc networks can move out of the labs and provide a flexible and cheap networking solution. Evolutionary computing algorithms are a class of bio-inspired computing algorithms. Bio-inspired computing refers to the collection of algorithms that use techniques learnt from natural biological phenomena and implement them to solve a mathematical problem (Olario & Zomaya, 2006). Natural phenomena such as evolution, genetics, and collective behavior of social organisms and functioning of a mammalian brain teach us a variety of techniques that can be effectively employed to solve problems in computer science which are inherently tough. In this Chapter and the chapter entitled, “Swarm Intelligence Approach for Wireless Ad Hoc Networks” of this book, we present some of the currently available important implementations of bio-inspired computing in the field of ad-hoc networks. This chapter looks at the problem of optimal clustering in ad-hoc networks and its solution using Genetic Programming (GP) approach. The chapter entitled, “Swarm Intelligence Approaches for Wireless Ad Hoc Networks” of this book, continues the same spirit and explains the use of the principles underlying Ant Colony Optimization (ACO) for routing in ad-hoc networks.