Traffic profile measurement and traffic driven network reconfiguration

Abstract

In multi hop ad hoc network, nodes are constrained by the power or the energy sources that are available. Efficient algorithms for saving the energy are required to manage the network in such a manner that the resources available, get properly utilized and nodes in the network do not drain out. Switching and working over idle and busy periods of a node, one can achieve this dual goal and it comes under the category of sleep scheduling. In this approach, the basic emphasis is on the traffic profiles of the network nodes which is estimated based on the network traffic flows. The traffic estimated becomes the weight of the path between the different source destination pairs available in the network and this value averaged either at the local level or the global level is needed to be reduced, hence saving the overall energy of the network. The traffic weighted paths which consume energy above the average or the threshold value become the prime candidates for the consideration. The aim is to go for shorter and lighter paths in the network which require the sleeping nodes in the proximity of heavy and longer paths to be awaken to be a part of network topology and saving the extra energy wasted on longer paths. This makes the nodes in longer paths go through a sleep state so that they can be utilized when they are required, hence making topology changes by reconfiguring the paths.