Cooperative wideband spectrum detection based on maximum likelihood ratio for CR enhanced VANET

Cognitive radio (CR) enhanced vehicular ad hoc network (CR-VANET) is a special VANET that each vehicle could explore local spectrum gap and reduce data congestion by mitigating interference by dynamically accessing spectrum holes. One of the most important challenges for CR-VANET is to identify the presence of primary users (PU) over wideband spectrum at a particular time and specific geographic location. In this paper, maximum likelihood ratio (MLR) detection is proposed as an effective scheme for wideband spectrum detection in orthogonal frequency division multiplexing (OFDM) based overlay CR-VANET. OR fusion rule with square-law selection (SLS) is implemented at roadside to make global detection decision by means of the combination of each vehicle's local sensing results. Hence, cooperative wideband spectrum detection is implemented in dynamic topology CR-VANET to enhance the reliability of spectrum detection. Specifically, robust cooperative sensing technique is investigated with the consideration of sensing channel SNR, reporting channel fading, as well as inter-user error rate, with the purpose of exploiting diversity in multi-vehicle cooperative sensing scenario. Limitations of cooperative wideband spectrum sensing are presented with MLR detection in space-time block code (STBC) and time division multiple access (TDMA) protocols for practical consideration. Simulation results are presented to confirm our theoretical derivations. It is revealed that, compared with conventional energy detection (ED), cooperative wideband spectrum detection with MLR has adaptive detection thresholds. Moreover, robust cooperative sensing by roadside SLS OR rule and its limitations are also performed with receiver operation characteristics (ROC). It is proven that, MLR detection is available in robust cooperative sensing with dynamic threshold variations for CR-VANET practical applications.