Enhancing Wireless Communication Efficiency Through Advanced Frequency Band and Path Loss Prediction with Random Forest Algorithms
V. Revathi, Yagateela Pandu Rangaiah, Amit Dutt, Dinesh Kumar Yadav, Tamam Ali Abd Ulabbas Abedi · 2025
Mobile communication systems are popular in today's world, whereas estimating the frequency bands and the path loss is complicated due to the changes in environment and operations. This research introduces a new method involving the use of Random Forest algorithms in identification of the frequency bands and path loss in various wireless communication scenarios accurately. The methodology incorporates voluminous databases that consist of the environmental aspects such as type of terrain, Building density and foliage and operating parameters consisting of transmission power, antenna height & frequency band. This is made possible through the incorporation of Random Forests which has the ability to learn from the ensemble in cases of complex and non-linear input data. The investigation reveals that Random Forests are capable of improving the predictive performance far much better than the conventional path loss models such as Okumura-Hata and COST-231 models. The suggested approach not only contributes to the enhancement of the selection of frequency bands, thus increasing the spectrum resource efficiency but also offers accurate calculation of the path loss for networks, which is vital in the network design and implementation. Real-world applications and numerous simulations prove the applicability of the model demonstrating its capacity to meet the flexible needs of 5G and beyond context. This thesis benefits the progress of predictive analytics in wireless communication while providing a versatile application to improve the network and its dependability.