Nonlinear Quantization Scheme Considering Quantization Error for Soft Decision Cooperative Spectrum Sensing in Cognitive Radios

Yuuki TANI  Takahiko SABA 

Publication
B - Abstracts of IEICE TRANSACTIONS on Communications (Japanese Edition)  Vol.J95-B  No.7  pp.907-917
Publication Date: 2012/07/01
Online ISSN: 1881-0209
Print ISSN: 1344-4697
Type of Manuscript: Special Section PAPER (Special Section on Frontiers by Young Researchers)
Category: 
Keyword: 
cognitive radio,  energy detector,  soft decision cooperative spectrum sensing,  quantization, 

Full Text(in Japanese): PDF(1.3MB)

Summary: 
Recently, cognitive radio techniques which recognize and reuse the spectrum for licensed users are actively studied for the efficient use of spectrum. For the realization of cognitive radios, a highly accurate sensing technique to detect the presence of licensed users is essential. For such a purpose, soft decision cooperative spectrum sensing techniques in which the cognitive base station collects the received signal energy information from multiple users and see if the licensed users are present are proposed. In soft decision cooperative spectrum sensing, since the bandwidth of the common control channel used for collecting the energy information is limited, it is necessary to quantize the received signal energy information in each user. In this paper, we propose a quantization scheme of energy information in each user. In our proposed scheme, the quantization threshold is decided using the information of the number of sensing users as well as the information of noise power of each user, and the quantized energy values are decided using the expected values of the energy observed between each threshold. Furthermore, we derive the theoretical expressions of the false alarm and the detection probabilities in the cognitive base station when the proposed quantization scheme is employed. The probabilities obtained from the theoretical expressions agree well with the results obtained from computer simulation. We also show that the our proposed scheme can achieve the desired probability of false alarm under the environment where SNR of each sensing user is different from those of others because the decision threshold is determined appropriately with the consideration of the quantization error.