Effective Combining Scheme for Channel Sensing in Multiuser Cognitive Radio Systems

In cognitive radio (CR) systems, channel sensing should be performed very accurately in order to protect primary user (PU) and to achieve high spectral efficiency. Recently, there are some efforts to improve the accuracy of the channel sensing in the CR systems. In particular, it is shown that the correctness of channel sensing can be improved by combining the multiple sensing results from CR nodes [1]. However, previous works do not consider the fact that CR nodes receive very dissimilar energy from PU with respect to each other, according to their distances from the PU. For example, if a node is close to a PU, it can receive large amount energy from the PU. On the other hand, a node located far away from the PU receives weaker energy from the PU. In this situation, without considering the dissimilarity of received energy among CR nodes, optimal combining for the sensing information cannot be achieved. In this paper, we propose an effective combining scheme for the multiuser CR systems. In contrast to the previous works, the proposed scheme combines multiple sensing results reported from CR nodes, considering the heterogeneity of their received energy caused by the diversity of their distances from the PU. To accomplish this, the proposed scheme investigates the probability distribution of statistics from the energy detector by using the virtual grid.