Cognitive Routing with Stretched Network Awareness through Hidden Markov Model Learning at Router Level

The routing of packets are generally performed based on the destination address and forward link channel available from the instantaneous Router without sufficient cognizance of either the performance of the forward Router or forward channel characteristics. The lack of awareness of forward channel property can lead to packet loss or delayed delivery leading to multiple retransmissions or routing to an underperforming pathway. This paper describes an application of Cognitive Network to improve the network performance by implementing a Hidden Markov Model (HMM) algorithm for learning and predicting the performance of surrounding routers continuously while a routing demand is initiated. The cognition segment/domain of every router can gain knowledge about the quality of forward network. The information of the current network conditions is shared between routers by the ‘Forward Channel Performance Index’ FCPI. This enables complete cognition of surroundings and efficient delivery of messages in various paradigms of performance.