PALM-CSS: a high accuracy and intelligent machine learning based cooperative spectrum sensing methodology in cognitive health care networks

Spectrum sensing is the most crucial importance in cognitive radios. We propose a novel machine-learning algorithm for spectrum sensing in cognitive radio networks, which plays an essential role in medical data transmission. In this regard, high-speed pre-emptive decision-based multi-layer extreme learning machines are implemented for co-operative spectrums sensing in CR health care networks. For a radio channel, different vectors such as energy levels, distance, Channel ID, sensor values are determined at CR devices and are considered as a feature vector and thus used to feed into the proposed classifier for the determination of the availability of the channel. The classifier further categorizes the parameters such as user identification i.e., primary and secondary users, availability of channels, and the most crucial predictive decision of the available channels. The proposed PALM-CSS consists of two major phases, such as classification and prediction. Before the online classification and prediction, datasets are generated, and these datasets are used for the training of the proposed classifier. The proposed classifier uses the principle of high-speed priority-based multi-layer extreme learning machines for the classification and prediction. The experimental testbed has designed based Multicore CoxtexM-3 boards for implementing the real-time cognitive scenario and various performance parameters such as prediction accuracy, training and testing time, Receiver operating characteristics, and accuracy of detection. Furthermore, the proposed algorithms has also compared with the other existing machine learning algorithm such as artificial neural networks, support vector machines, K-nearest neighbor, Naïve Bayes and ensemble machine learning algorithms in which the proposed algorithm outperforms the other existing algorithms and finds its more suitable for cognitive health care networks.