A Cognitive Architecture for Personal Networks

This paper proposes a three-layer cognitive architecture for pervasive and intelligent computing of personal networks. The key element of the proposed architecture is the cognitive layer, which consists of five components, namely the context cognition, the personalization cognition, the resource cognition, the network cognition, and the cognition management. In order to demonstrate the purpose of each one of the cognition component, we present a motivation example on session mobility and show that the proposed architecture enables proactive configuration and thus hides the latency of configuration. Finally, this paper identifies the research issues that need to be addressed in order to implement the cognitive architecture for personal networks.