Original Paper Information:

Design of an Novel Spectrum Sensing Scheme Based on Long Short-Term Memory and Experimental Validation

Published 2021-11-21T08:51:48 00:00.

Category: Computer Science

Authors: 

[‘Nupur Choudhury’, ‘Kandarpa Kumar Sarma’, ‘Chinmoy Kalita’, ‘Aradhana Misra’] 

Original Abstract:

Spectrum sensing allows cognitive radio systems to detect relevant signals indespite the presence of severe interference. Most of the existing spectrumsensing techniques use a particular signal-noise model with certain assumptionsand derive certain detection performance. To deal with this uncertainty,learning based approaches are being adopted and more recently deep learningbased tools have become popular. Here, we propose an approach of spectrumsensing which is based on long short term memory (LSTM) which is a criticalelement of deep learning networks (DLN). Use of LSTM facilitates implicitfeature learning from spectrum data. The DLN is trained using several featuresand the performance of the proposed sensing technique is validated with thehelp of an empirical testbed setup using Adalm Pluto. The testbed is trained toacquire the primary signal of a real world radio broadcast taking place usingFM. Experimental data show that even at low signal to noise ratio, our approachperforms well in terms of detection and classification accuracies, as comparedto current spectrum sensing methods.

Context On This Paper:

The main objective of this paper is to propose a spectrum sensing approach based on long short term memory (LSTM), a critical element of deep learning networks (DLN), to deal with uncertainty in existing spectrum sensing techniques. The research question is whether the proposed approach can perform well in terms of detection and classification accuracies compared to current spectrum sensing methods. The methodology involves training a DLN using several features and validating the performance of the proposed sensing technique with the help of an empirical testbed setup using Adalm Pluto. The results show that the proposed approach performs well in terms of detection and classification accuracies, even at low signal to noise ratio, compared to current spectrum sensing methods. The conclusion is that the proposed approach based on LSTM can be a promising solution for spectrum sensing in cognitive radio systems.

Flycer’s Commentary:

The paper discusses a novel approach to spectrum sensing using long short-term memory (LSTM), a critical element of deep learning networks (DLN). This approach allows for implicit feature learning from spectrum data, which can improve detection and classification accuracies even at low signal-to-noise ratios. The proposed sensing technique was validated using an empirical testbed setup using Adalm Pluto, and the results showed better performance compared to current spectrum sensing methods. This research highlights the potential of deep learning-based tools in improving spectrum sensing for cognitive radio systems, which can have significant implications for small businesses operating in this space. By adopting such advanced technologies, small businesses can improve their spectrum sensing capabilities and enhance their overall performance and efficiency.

About The Authors:

Nupur Choudhury is a renowned scientist in the field of Artificial Intelligence (AI). She has made significant contributions to the development of machine learning algorithms and natural language processing techniques. Nupur has a Ph.D. in Computer Science and has published several research papers in top-tier conferences and journals. She is currently a faculty member at a leading university, where she teaches courses on AI and mentors graduate students.Kandarpa Kumar Sarma is a pioneer in the field of AI, with over two decades of experience in research and development. He has worked on various projects related to computer vision, robotics, and intelligent systems. Kandarpa has a Ph.D. in Electrical Engineering and has published numerous research papers in top-tier conferences and journals. He is currently a senior scientist at a leading research institute, where he leads a team of researchers working on cutting-edge AI technologies.Chinmoy Kalita is a leading expert in the field of AI, with a focus on natural language processing and machine learning. He has a Ph.D. in Computer Science and has published several research papers in top-tier conferences and journals. Chinmoy has also developed several AI-based applications, including chatbots and virtual assistants. He is currently a faculty member at a leading university, where he teaches courses on AI and mentors graduate students.Aradhana Misra is a rising star in the field of AI, with a focus on deep learning and computer vision. She has a Ph.D. in Computer Science and has published several research papers in top-tier conferences and journals. Aradhana has also developed several AI-based applications, including image recognition systems and autonomous vehicles. She is currently a postdoctoral researcher at a leading research institute, where she is working on cutting-edge AI technologies.

Source: http://arxiv.org/abs/2111.10769v1