Lifting Wavelet-Assisted EM Joint Estimation and Detection in Cooperative Spectrum Sensing

Area of Science:

• Electrical Engineering
• Signal Processing
• Wireless Communications

Background:

• Cognitive radio (CR) enhances spectrum utilization by identifying spectral holes for dynamic resource allocation.
• Accurate wireless environment information is crucial but difficult to obtain in real-world scenarios.
• Key parameters like signal-to-noise ratio (SNR), noise variance, and channel occupancy rate are challenging to ascertain beforehand.

Purpose of the Study:

• To propose a full-blind spectrum sensing method for cognitive radio that overcomes limitations of prior parameter estimation.
• To enhance detection performance and efficiency in dynamic wireless environments.

Main Methods:

• A lifting wavelet-assisted Expectation-Maximization (EM) joint estimation and detection technique is employed.
• Lifting wavelet is utilized for noise variance estimation to boost detection probability and convergence speed.
• A stream learning strategy is incorporated for dynamic estimation of SNR and channel prior occupancy rate, accommodating SU mobility.

Main Results:

• The proposed method achieves full-blind detection, estimating critical parameters without prior knowledge.
• Lifting wavelet integration improves noise variance estimation, leading to faster convergence and higher detection probability.
• Stream learning effectively handles SU mobility for SNR and channel occupancy rate estimation.
• Simulation results show the method's detection performance is comparable to semi-blind EM techniques.

Conclusions:

• The developed lifting wavelet-assisted EM method offers a robust solution for full-blind spectrum sensing in cognitive radio.
• This approach enhances spectrum utilization by improving the accuracy and efficiency of spectral hole detection.
• The method's adaptability to mobile secondary users (SUs) makes it suitable for practical, dynamic wireless environments.