Cache Optimization in Wireless Sensor Networks Using Soft Computing Techniques

Abstract

Wireless sensor networks (WSN) are vastly distributed networks of small, lightweight nodes capable of sensing, processing and communicating environmental parameters They are capable of operating in harsh environmental conditions with very little or no human intervention. WSNs are competent enough to deal with heterogeneous nodes, high mobility, frequent path breaks, node failures and communication failures. These characteristics makes WSN use in certain applications indispensable. However, WSNs are constrained by limited power, memory, processing capacity and therefore are faced by many challenges. Energy consumption of a sensor node is a key issue to be addressed by researchers so as to keep it to minimum for an enhanced network lifetime. This trade-off between energy usage of a sensor node and that of satisfying application level requirements is very crucial to sensor network design and worthiness of WSNs. At the network layer, cause of high energy dissipation by sensor nodes is the routing technique used for data transmission. Routing in WSNs is data centric than address centric. Multiple sensors do the same task and message redundancy is generated. Sometimes, due to energy scavenging sensors may turn off their circuitry for a while. In such situations, alternative paths need to be present between the source and the sink. This would also help in load sharing. Routing in WSNs is multi-hop therefore, path breaks, node disconnections, node failures have to be accounted for. Literature gives numerous routing algorithms that solve data routing in WSNs. However, even the best routing protocols are not efficient enough to conserve network energy [Rahman 2007]. Another possible way to minimize power consumption is by use of caching popular data. Caching of popular data can save a sensor network from exploitation of many of its scarce resources [Xu 2008]. To mention a few, caching helps in reducing amount of communication between nodes in the network, reduces network wide transmission, hence, ....