TCRR: A Channel Quality Prediction Framework for Industrial Wireless Communication

Published

September 21, 2026

Author(s)

Oishy Saha, Mohamed Hany, Jing Geng, Shuvra Bhattacharyya, Richard Candell

Abstract

Reliable wireless communication in industrial Internet of Things (IoT) environments helps to manage varying operational conditions and ensure robust connectivity among devices, actuators, and machines. The anticipation of a channel's behavior can accelerate the processes of identifying channel degradation and choosing an optimal channel and lead to a more dependable and energy-efficient network. This paper proposes a novel clustering-based ensemble approach for channel impulse response (CIR) prediction using machine learning in industrial wireless environments. Initially, the training dataset undergoes clustering to capture the intrinsic characteristics and similarities within specific types of channel conditions. Then each cluster is used to train a random forest regressor model to make predictions that are best fit to types of CIR patterns and variations associated with the cluster. The results of these different random forest regressor models are then combined together by averaging to construct an integrated prediction framework. During inference, input data is first preprocessed to filter out noise and to compress the data into a form that can be processed efficiently. The output of preprocessing is then processed by the ensemble of random forest regressor models. Extensive experimentation conducted under realistic industrial conditions demonstrates the effectiveness of the proposed method. The results demonstrate the potential of the proposed method for significantly enhancing the accuracy of CIR prediction, thereby contributing to enhanced reliability, efficiency, and robustness in industrial wireless network applications, particularly in dynamic and challenging industrial environments.

Proceedings Title

Proceedings of the 4th International Conference on Intelligent Computing, Communication, Networking and Services (ICCNS2026)

Conference Dates

September 8-11, 2026

Conference Location

Bucharest, RO

Conference Title

International Conference on Intelligent Computing, Communication, Networking and Services (ICCNS2026)

Pub Type

Conferences

Download Paper

Local Download

Keywords

industrial wireless network, channel impulse response (CIR), cluster analysis, random forest regressor, ensembling, CIR prediction

Citation

Saha, O. , Hany, M. , Geng, J. , Bhattacharyya, S. and Candell, R. (2026), TCRR: A Channel Quality Prediction Framework for Industrial Wireless Communication, Proceedings of the 4th International Conference on Intelligent Computing, Communication, Networking and Services (ICCNS2026), Bucharest, RO, [online], https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=961496 (Accessed September 22, 2026)

Additional citation formats