Model-Based Cognitive Radio Strategies

Many frequency bands for wireless services are severely underutilized by the primary users (PU) to which these bands are assigned. This motivates a new class of wireless communication devices known as cognitive radios (CR), which identify vacant spectrum and transmit accordingly. In this thesis, the PU traffic model knowledge as well as all the observations available to the CR are included in the CR transmission decisions. A transmission strategy is introduced that is based on comparing an a-posterior probability (APP) log-likelihood ratio (LLR) with a threshold. The objective is to maximize the utilization ratio (UR) subject to that the interference ratio (IR) is below a certain level. In papers A and B, we study CR transmission strategies that are based on all noisy observations of the PU activities, even when the CR itself is transmitting. Paper A demonstrates a more than 300% increase in UR over standard energy detection, for the same IR value, at the PU signal to CR noise power ratio (SNR) of −5 dB. Then, in paper B, we use a continuousoutput hidden Markov model for the received signal and calculate an APP LLR based on this model. This paper shows that this strategy is the optimum in the sense of maximizing the UR, given a certain maximum allowed IR, among all CRs. Moreover, two practical schemes for calculating the transmission threshold are introduced. Numerical results show that the first method yields a threshold that is close to optimum when the PU use a large fraction of the available spectrum (i.e., when the PU activity level is high). The second method is analytically proven to always give a valid threshold. Simulation results show a 116% improvement in UR with PU state estimation over energy detection, at an SNR of −3 dB and IR level of 10%. In paper C, we extend paper B to consider that PU activities cannot be observed when CR is transmitting, in other words they are censored. This new strategy, entitled CLAPP, calculates a new LLR, which is compared with a threshold. This threshold is computed with a bisection search method. Simulation results show that CLAPP has a 52% gain in UR over the best censored energy detection scheme for a maximum IR level of 10% and an SNR of −2dB. In paper D, we introduce new time-varying thresholds for sequential spectrum sensing. These new thresholds, for an SNR of −10 dB, in comparison with standard sequential detection with parallel (fixed) thresholds with similar probabilities of misdetection and false alarm, performs 54% faster in terms of maximum detection time (90 percentile).