Cognitive radio network: architecture, algorithms, and testbed

Cognitive radio has been put forward in recent years to make an efficient use of the scarce radio frequency spectrum. However, the system architecture of cognitive radio was not defined. Moreover, to our best knowledge, no true real-time cognitive radio system nor cognitive radio network had been demonstrated. It is necessary to build a large-scale cognitive radio network test-bed. A cognitive radio network testbed will not only verify concepts, algorithms, and protocols for cognitive radio and cognitive radio network, but also identify practical problems for future research. In this dissertation, an architecture for cognitive radio systems, an architecture for cognitive radio network testbeds, and a design for the nodes of cognitive radio network testbeds, are proposed. The response delay is defined and the minimum response delays in hardware platforms are measured. Channel state prediction is proposed to help reduce the negative impact of the response delay. Moreover, algorithms for spectrum sensing, cooperative spectrum sensing, channel state prediction, and cooperative channel state prediction, are proposed and tested using measured real-world data. An algorithm for spectrum auction is also proposed and simulated. Experimental results show that the proposed algorithms are effective. In addition, a proposed algorithm for spectrum sensing is implemented and demonstrated on a hardware platform in real time. To our best knowledge, this is the first time that real-time spectrum sensing with controllable primary user devices is demonstrated. With the proposed architectures, algorithms, and design, a large-scale cognitive radio network testbed can be built further.