|
|
 |
| |
|
|
|
Performance Analysis of Cooperative Spectrum Sensing under Noise Uncertainty |
|
|
|
PP: 587S-593S |
|
|
|
Author(s) |
|
|
|
Yunxue Liu,
Fen Li,
Guoying Hu,
Ying Wu,
|
|
|
|
Abstract |
|
|
| Spectrum sensing is a critical technology for primary user detection in cognitive radio networks, in which energy detector is widely used for spectrum sensing due to its generality and low complexity. However, noise uncertainty degrades the sensing performance of energy detection severely because of its dependence on the noise power. In this paper, we take the average power as the decision statistic and derive the closed-formed expressions for the average detection and false alarm probabilities under noise uncertainty. We then demonstrate the performance of the three kinds of well known hard fusion rules (OR, AND and Majority) over AWGN and Rayleigh channels, and we identify the best rule in different conditions. Extensive simulations indicate that in AWGN channels, Majority rule is optimal at the higher SNR, while AND rule performs best for most situations at low SNR. Moreover, OR rule always exhibits the best performance in Rayleigh channels at low SNR. However, when SNR is higher, the optimal rule is changed with the variation of number of cognitive users. Our research is very helpful and meaningful for selecting the proper hard fusion rule in practical cognitive radio networks. |
|
|
|
|
|
|
|
