On the Use of Higher Layer Information for Cognitive Networking

Cognitive radio networking research today mainly focuses on finding efficient ways to let secondary users access radio spectrum that is licensed to primary users, with minimal interference to the license owners. Physical layer cognition, though very important, is complex, expensive, and can provide only limited information about the higher layer traffic. We argue that, even in the absence of physical layer cognitive capability, higher layer traffic information can still be used to generate sufficient cognitive networking information to improve system performance. In this paper, we present an architecture for cognitive networking and an early prototype and experimental setup of a cognitive network access point (CogNet AP), and we describe our observations and lessons learned from this experimental activity. The CogNet AP gathers, processes, analyzes, and stores information available through its monitoring interface in order to build a cognitive local repository which holds the spatio-temporally tagged network traffic information. The inexpensiveness of the components used for building the CogNet AP shows the flexibility of building Cognitive Network elements compared to cognitive radio devices. The proposed cognitive networking architecture and prototype point to the many possible application scenarios and research potential for such systems which use temporal patterns of higher layer traffic information. From our experiments we found that the use of cognitive information derived from higher networking layers resulted in achieving better system throughput.