Optimal Truncated Ordered Sequential Cooperative Spectrum Sensing in Cognitive Radio

The cost for increasing the accuracy of cooperative spectrum sensing in cognitive radio requires a long time for collecting sensing data. Sequential fusion, in which the sensing data reports are sequentially collected, combined, and tested with two thresholds for deciding to wait for the next data report or concluding the presence or absence of the primary signal, is considered to reduce the collecting time. A further reduction in the average sequential report number (ASN) can be achieved by arranging the reports in descending order of data quality where the better sensing data is reported earlier. In this paper, the ASN and the error probability of the ordered sequential fusion are analytically calculated based on the direct method. The results are then adopted to design an optimal method for obtaining the optimal sequential thresholds and truncated point. The optimal truncated point enables one to minimize the ASN by discarding the sensing data with low quality, while the optimal sequential thresholds ensure the maintaining of the sensing performance of sequential fusion be the same level to that of the conventional parallel fusion.