Volume-based method for spectrum sensing

It is recently shown that algorithms derived from random matrix theory (RMT) can provide superior performance for spectrum sensing, which corresponds to the task of detecting the presence of primary users in cognitive radio. The essence of the RMT-based methods is to utilize the distribution of extremal eigenvalues of the received signal sample covariance matrix (SCM), namely, the Tracy-Widom (TW) distribution. Although the TW distribution is quite useful in spectrum sensing, computationally demanding numerical evaluation is required because it does not have an explicit closed-form expression. In this paper, we devise two novel volume-based detectors by exploiting the determinant of the SCM or volume to distinguish between the signal-presence and signal-absence cases. With the use of RMT, we accurately produce the theoretical decision threshold for one of the detectors under the Gaussian noise assumption. Simulation results are included to illustrate the effectiveness of the volume-based detectors.