Journal
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS
Volume 68, Issue 1, Pages 241-245Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TCSII.2020.3010587
Keywords
Sensors; Support vector machines; Machine learning algorithms; Training; Fading channels; Wideband; Cascading style sheets; Cognitive radio; cooperative spectrum sensing; machine learning; support vector machine; wideband spectrum sensing
Categories
Ask authors/readers for more resources
Cooperative Spectrum Sensing addresses the Hidden Primary User problem by leveraging the spatial diversity of spectrum sensors, while introducing a novel fast soft decision algorithm based on Machine Learning theory for wideband Cooperative Spectrum Sensing, which outperforms traditional algorithms with a faster solution.
Hidden Primary User problem caused by fading and shadowing severely affects the detection rate of the cognitive radio systems with a single spectrum sensor. Cooperative Spectrum Sensing has been introduced to tackle this problem by using the spatial diversity of spectrum sensors. It is shown that the use of soft decision algorithms in fusion center has a better performance than hard decision algorithms. The problem of soft decision based on sensor measurements perfectly matches the Machine Learning paradigm. In this brief, a novel fast soft decision algorithm is proposed based on Machine Learning theory for wideband Cooperative Spectrum Sensing, which finds a decision boundary to classify the Power Spectral Density (PSD) measurement vectors of sensors into the empty and occupied channel classes. The statistical characteristics of sensors PSD samples are employed to derive a fast solution, which outperforms the SVM-linear algorithm. By solving an optimization problem a Constant False Alarm Rate algorithm is introduced and then it is improved to the Adaptive False Alarm Rate algorithm. The proposed algorithm reduces the average training time to one tenth of the SVM-linear training time, while the detection probability is the same. The hardware implementation of the proposed algorithm is described in Verilog HDL and the corresponding simulation results are presented.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available