A predictive network resource allocation technique for cognitive wireless networks

Future wireless networks are evolving towards a heterogeneous cooperative and cognitive architecture to support broadband communication needs of different types of traffic. Future wireless networks will allocate network resources based on cooperative techniques. Resource controllers will apply the cognitive principle to find out status of various networks. We propose a predictive resource allocation strategy where we employ adaptive algorithms to predict the network loading. These algorithms can detect changes in the traffic characteristics, and adapt automatically. To validate our claim, we present the results of applying these adaptive algorithms on real-world network traffic traces.