Abstract
Spectrum sensing (SS) is an important tool in finding new opportunities for spectrum sharing. The users, called Secondary Users (SU), who do not have a license to transmit without hindrance, need to employ SS in order to detect and use the spectrum without interfering with the licensed users’ (primary users’ (PUs’)) transmission. Deep learning (DL) has proven to be a good choice as an intelligent SS algorithm that considers radio environmental factors in the decision-making process. It is impossible though for SU to collect the required data and train complex DL models. In this paper, we propose to employ a Federated Learning (FL) algorithm in order to distribute data collection and model training processes over many devices. The proposed method categorizes FL devices into groups by their mean Signal-to-Noise ratio (SNR) and creates a common DL model for each group in the iterative process. The results show that detection accuracy obtained via the FL algorithm is similar to detection accuracy obtained by employing several DL models, namely convolutional neural networks (CNNs), specialized in spectrum detection for a PU signal with a given mean SNR value. At the same time, the main goal of simplification of the SS process in the network is achieved.
Highlights
In today’s wireless communication systems, the radio spectrum has become a scarce resource
To correctly evaluate the final results, we calculated Pd and Pfa for the same Secondary Users (SU) data, but employing separate convolutional neural networks (CNNs) that specialize in every Signal-to-Noise ratio (SNR) value from a considered range, and we used these results for comparison with Federated Learning (FL) algorithm results
It has been proven that the FL method makes it possible for SU to perform intelligent sensing without the need for extensive data collection and CNN model training
Summary
In today’s wireless communication systems, the radio spectrum has become a scarce resource. The individual sensing performed by each single SU and employing ML is computationally complex and may be inaccurate This is because ML algorithms require many training data to be able to recognize the time, frequency, and location dependencies existing in the transmitted and received signal. SUs exchange their sensing results or collected data to decide cooperatively on the current spectrum state This approach solves the problem of generating an ML model, which can be created by the elected end user device or the so-called fusion center or a central server, but it still does not answer the problem of collecting labeled data by SUs. A promising solution to the problems presented above is Federated Learning (FL), which an iterative procedure that edge devices (called FL nodes) that create their ML models on their local data.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
