A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Reduced Kernel PCA Model for Nonlinear Spectrum Sensing in Cognitive Radio Network
https://orcid.org/http://orcid.org/0000-0003-0224-7464
Cognitive radio is a device designed for the efficient usage of channels of radio spectrum. In the success of cognitive radio networks, spectrum sensing functions play a key role. Energy detection is one of the cognitive radio spectrum sensing techniques. Principal component analysis (PCA) is a basic technique for dimension reduction for primary user (PU) detection. A Kernel PCA (KPCA) is the non-linear version that makes best use of a complex high-dimensional spatial framework where a feature of the kernel indirectly transits. The traditional KPCA does not work correctly, although it succeeds sensibly because of complexity, high memory requirements, and huge training data, which is a major constraint. To resolve this issue, a reduced KPCA (RKPCA) method is new to detect the presence or absence of the PU in a cognitive radio network. RKPCA is a new learning approach that combines reduction in dimensionality, guided training, and the collection of the kernel. This alternative approach results in measurements that retain important observations. The proposed system performance is evaluated in conditions of probability of detection, false alarm rates, and number of data samples with varying SNR, which are linearly related to collected observations. At the end of the simulation studies, we made the spectrum sensing more accurate by choosing the best threshold as per the cumulative percentage of variance.
Reduced Kernel PCA Model for Nonlinear Spectrum Sensing in Cognitive Radio Network
https://orcid.org/http://orcid.org/0000-0003-0224-7464
Cognitive radio is a device designed for the efficient usage of channels of radio spectrum. In the success of cognitive radio networks, spectrum sensing functions play a key role. Energy detection is one of the cognitive radio spectrum sensing techniques. Principal component analysis (PCA) is a basic technique for dimension reduction for primary user (PU) detection. A Kernel PCA (KPCA) is the non-linear version that makes best use of a complex high-dimensional spatial framework where a feature of the kernel indirectly transits. The traditional KPCA does not work correctly, although it succeeds sensibly because of complexity, high memory requirements, and huge training data, which is a major constraint. To resolve this issue, a reduced KPCA (RKPCA) method is new to detect the presence or absence of the PU in a cognitive radio network. RKPCA is a new learning approach that combines reduction in dimensionality, guided training, and the collection of the kernel. This alternative approach results in measurements that retain important observations. The proposed system performance is evaluated in conditions of probability of detection, false alarm rates, and number of data samples with varying SNR, which are linearly related to collected observations. At the end of the simulation studies, we made the spectrum sensing more accurate by choosing the best threshold as per the cumulative percentage of variance.
Reduced Kernel PCA Model for Nonlinear Spectrum Sensing in Cognitive Radio Network
https://orcid.org/http://orcid.org/0000-0003-0224-7464
Cognitive radio is a device designed for the efficient usage of channels of radio spectrum. In the success of cognitive radio networks, spectrum sensing functions play a key role. Energy detection is one of the cognitive radio spectrum sensing techniques. Principal component analysis (PCA) is a basic technique for dimension reduction for primary user (PU) detection. A Kernel PCA (KPCA) is the non-linear version that makes best use of a complex high-dimensional spatial framework where a feature of the kernel indirectly transits. The traditional KPCA does not work correctly, although it succeeds sensibly because of complexity, high memory requirements, and huge training data, which is a major constraint. To resolve this issue, a reduced KPCA (RKPCA) method is new to detect the presence or absence of the PU in a cognitive radio network. RKPCA is a new learning approach that combines reduction in dimensionality, guided training, and the collection of the kernel. This alternative approach results in measurements that retain important observations. The proposed system performance is evaluated in conditions of probability of detection, false alarm rates, and number of data samples with varying SNR, which are linearly related to collected observations. At the end of the simulation studies, we made the spectrum sensing more accurate by choosing the best threshold as per the cumulative percentage of variance.
Reduced Kernel PCA Model for Nonlinear Spectrum Sensing in Cognitive Radio Network
J. Inst. Eng. India Ser. B
Pallam, Venkatapathi (author) / Khan, Habibulla (author) / Surampudi, Srinivasa Rao (author) / Immadi, Govardhani (author)
Journal of The Institution of Engineers (India): Series B ; 106 ; 181-187
2025-02-01
7 pages
Article (Journal)
Electronic Resource
English
Reduced Kernel PCA Model for Nonlinear Spectrum Sensing in Cognitive Radio Network
| Springer Verlag | 2025
Mitigation of Spectrum Sensing Data Falsification Attack (SSDF) in Cognitive Radio Network
| Springer Verlag | 2022