Schemes for self management in wireless networks for throughput and reliability improvement

Abstract

In the current age of Information Technology revolution, quick availability and use of newlineinformation to make speedy decisions is becoming a competitive advantage for many newlinebusinesses. In such an environment, to have all decision-makers connected in, ubiquitous newlinecommunications has become the need of the hour. In turn, wireless networks are gaining newlineimportance as they are a key enabler of the ubiquitous communication paradigm. With newlinebusinesses depending heavily on speedy decision-making, loss of communication for small newlineamounts of time can potentially lead to very large losses to the business. Communication newlinenetworks in general, and wireless networks in particular are thus needed to provide a newlinehigh degree of availability and reliability, in addition to being able to adopt quickly to newlinemobility and changing network conditions. newlineSeveral factors affect the quality of communication over a wireless network: bad network newlinecoverage, fading, interference, failure of different network elements or failure of just a newlinesingle RF (Radio Frequency) channel. This is in contrast to wired networks, where newlineloss of communication can be either due to congestion or faulty physical links. In spite newlineof congestion effects, loss of communication in wired networks is much rarer than in newlinewireless networks. Loss of communication can be stated to be more a norm than an newlineexception in wireless networks. To keep the user transparent to such failures and to newlineprovide a high degree of availability, wireless networks must be able to recover from newlinethese failures and switch to alternate network resources so that the user sees no breakage newlinein services. This recovery must be automated and should be carried out without newlinethe intervention of an operator. Also, effects like fading, radio interference and other newlinechanging network conditions lead to a lower-than-expected throughput being observed newlineon wireless networks. In such cases, wireless networks should be able to change their newlineoperating parameters like power of transmission, RF channel in use, predictive buffering newlineof packets.