Multisensor Data Fusion: From Algorithm and Architecture Design to Applications Decision Fusion in Cognitive Wireless Sensor Networks

In this chapter, we focus on a cognitive wireless sensor network (WSN), where a primary WSN (PWSN) is co-located with a cognitive (or secondary) WSN (CWSN). The shared frequency spectrum can be “freely” assigned by the PWSN. In particular, this spectrum is divided into disjoint “subchannels” and each subchannel is assigned (in a unique way) to a node of the PWSN. We assume that the nodes of the CWSN cooperate to sense the frequency spectrum and estimate the free subchannels which can be used to transmit their data. The key contribution of this work is two-fold. First, the sensing correlation among the secondary nodes is exploited to improve the reliability of the decision, taken by a secondary fusion center (FC), on the occupation status (by a node of the PWSN) of each subchannel. In this context, we provide a simple model for characterizing the local sensing performance per subchannel, in terms of probabilities of missed detection (MD) and false alarm (FA). Then, we analyze the use of joint source channel coding (JSCC) schemes in the CWSN, in order to further exploit the source correlation to make the final decision of the secondary FC on the status of the frequency subchannels more reliable.