Estimation of channel profile and signal-to-noise ratio in wireless receivers using deep learning

Ebrahim Toosi, Mohammad Ebrahim; Izadi Onji, Seyed Amin Allah

doi:10.22034/joeds.2024.473590.1062

Estimation of channel profile and signal-to-noise ratio in wireless receivers using deep learning

Document Type : Original Article

Authors

mohammad ebrahim ebrahim toosi ¹

seyed amin allah Izadi onji ²

¹ iran , semnan , damghaMaster of Science in Aerospace, University of Science and Technology, Tehran, Iran.

² Assistant Professor, Faculty of Electrical Engineering, Shahid Sattari University of Science and Technology

10.22034/joeds.2024.473590.1062

Abstract

Signal parameters are critical in determining the performance of wireless networks and the user experience they provide. These parameters, such as the channel profile and signal-to-noise ratio (SNR), can significantly affect the quality of wireless communications, either directly or indirectly. Accurate estimation of these parameters is essential in modern wireless communication systems. A promising approach to achieve this is through the application of deep learning techniques, which are widely used for estimating the channel profile and SNR due to their ability to capture complex patterns in signal data. By learning from training data, deep learning models can effectively extract key features related to the channel and SNR, resulting in more precise estimations. This paper proposes a novel method for estimating the channel profile and SNR for both LTE and 5G wireless standards using a deep learning-based approach, specifically a hybrid LSTM-CNN network architecture. The trained model demonstrates an accuracy of approximately 94% on the test dataset.

Keywords

Key words: channel profile

signal-to-noise ratio

deep learning

LTE

LSTM-CNN