Cross-Channel Estimation Using Supervised Probing and Sensing in Cognitive Radio Networks

Prior work in implementing spectrum sharing scenarios for cognitive radio networks has relied on the unlikely assumption of full cross-channel knowledge made available to the cognitive transmitter (CT) and/or the target primary receiver (PR). However, estimation of this cross-channel knowledge is not only important from a practical stand-point, but also in limiting the real interference power felt at the PR due to concurrent CT transmissions. We propose a supervised probing and sensing model, which enables the CT to gain a decent estimate of the cross-channel, assess the spectrum opportunities in its region of interference, and according to its throughput needs safely share or access primary user spectrum. With the probing power increment playing an essential part in the probing model, its optimization with respect to cross-channel estimation success and its influence on mean square error of the cross-channel estimation are discussed.