RNN-BIRNN-LSTM based spectrum sensing for proficient data transmission in cognitive radio
E. Vargil Vijay, K. Aparna · e-Prime - Advances in Electrical Engineering Electronics and Energy · 2023
With the advancements that are taking place in the wireless communications field, the number of users who are utilizing resources is also increasing; as a result, the wireless spectrum is scarce. In this article, RNN-BIRNN-LSTM with Gaussian noise (RBRLG)-based spectrum sensing (SS) for QAM16, CPFSK, QPSK, and BPSK modulation schemes has been proposed. Recurrent Neural Networks for sequential data use recurrent connections to capture temporal dependencies; BIRNN extends RNN by processing input in both forward and backward directions, capturing past and future context; and finally, LSTM, using specialized memory cells, efficiently manages long-term dependencies in sequential data. In order to create a spectrum sensing model, RNN units, BIRNN units, and LSTM units were cascaded in this paper. Open-source dataset RadioML2016.10B has been used for the investigation. The experimental results show that the proposed RBRLG-based SS has higher accuracy on the dataset, especially at -20 dB, a lower probability of miss detection percentage of 7.19%, and a lower sensing error (SE) percentage of 10.79%. The evaluation of performance indicators for our suggested model, such as the F1 Score, Jaccard Index, and Matthew's correlation coefficient, demonstrates that the proposed model provides improved SS performance.