LSSNN: A Novel LSTM-SNN Hybrid Architecture For Energy-Efficient Primary User Detection In Low SNR Cognitive Radio Networks

Author(s)	Ms. MANSHI PRAVINKUMAR SHAH, Dr. PARESH M DHOLAKIA
Country	India
Abstract	Cognitive radio networks require dependable spectrum sensing to recognize primary user activity in severe noise. Low SNR detection is hard because energy detection, matched filtering, and classical deep models are noise sensitive or capture limited temporal context. We present LSSNN, a hybrid Long Short Term Memory and Spiking Neural Network for accurate, energy aware sensing. The LSTM module learns sequential features from received samples, and the SNN produces event driven decisions. A synthetic CRN dataset was generated using BPSK, QPSK, and 16 QAM signals over AWGN and Rayleigh fading, with SNR from -20 dB to +4dB. The model was trained in Python with Adam and binary cross entropy. LSSNN achieves a detection probability of 0.94 at -10 dB, F1 score of 0.93, overall accuracy of 94.5%, and 6.05 µJ per inference, surpassing conventional baselines.
Keywords	Cognitive Radio, Spectrum Sensing, SNR, Long Short-Term Memory, Spiking Neural Network, Energy Efficiency, Signal Classification
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 8, Issue 1, January-February 2026
Published On	2026-01-15
DOI	https://doi.org/10.36948/ijfmr.2026.v08i01.66567

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research