Hybrid CNN-LSTM framework for cooperative spectrum sensing in Rayleigh and AWGN environments

View at publisher Summarise with VIRAT
Authors: Pankaj Verma, Sakshi Saxena, Aditi Dhiman, Mohit Dua
Cited by: 0
References: 25

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Abstract This paper presents a cooperative spectrum sensing (CSS) framework based on deep learning to enhance detection performance under Rayleigh, Rician, and AWGN channel conditions. A dataset generated using MATLAB, comprising energy statistics and sample covariance matrices from multiple secondary users (SUs), serves as the model input. Three deep learning architectures—Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM) network, and a hybrid CNN-LSTM model—are investigated. The results demonstrate that while the LSTM achieves the highest classification accuracy, the hybrid CNN-LSTM model exhibits superior robustness, attaining the highest AUC and faster improvement in detection probability, particularly at low SNR levels. The proposed framework operates without requiring prior knowledge of the primary user’s signal or noise power. Comparative analysis confirms the hybrid model’s effectiveness in balancing detection probability and false alarm rate, making it highly suitable for deployment in practical cognitive radio networks.

Fields of Study & Subjects

Cognitive Radio Networks and Spectrum Sensing Wireless Signal Modulation Classification Radar Systems and Signal Processing

Publication details

Journal / source
Physica Scripta
Volume
101
Issue
15
Pages
156001–156001
Publication date
25 March 2026
Language
EN