International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 12 | Dec 2024
p-ISSN: 2395-0072
www.irjet.net
Development of an Adaptive Pathloss Prediction Model Using NeuroFuzzy Systems for Wireless Optimization in Urban Areas Sani Aminu Shuaibu1, Samuel Ndueso John 2, Joshua Sokowonci Mommoh3, Etinosa NomaOsaghe4, Ashraf Adam Ahmad5, Hayatu Idris Bulama6 . 1Department of Electrical Electronic Engineering, Nigerian Defence Academy, Kaduna, 800281, Nigeria 2Department of Electrical Electronic Engineering, Nigerian Defence Academy, Kaduna, 800281, Nigeria 3Department of Software Engineering, Mudiame University Irrua, Edo, 310112, Nigeria.
4Department of Electrical and Electronics Engineering, Olabisi Onabanjo University, Ogun, 120107, Nigeria. 5Department of Electrical Electronic Engineering, Nigerian Defence Academy, Kaduna, 800281, Nigeria
6Department of Electrical and Electronics Engineering, Air Force Institution of Technology, Kaduna, 800282,
Nigeria ---------------------------------------------------------------------***--------------------------------------------------------------------ubiquitous data access, has driven significant advancements Abstract - As the need for mobile phones and fast data
in wireless mobile communication technologies. Cellular networks have undergone remarkable evolution, progressing from the first generation (1G) to the imminent fifth generation (5G) [1]. Each successive generation has introduced substantial improvements in technological capacity, data transmission speeds, and service quality for users. This evolution underscores the transformative impact of wireless communication in modern society. The initial phase of wireless networks utilized analog technology, primarily supporting voice communication with limited data services. Operating at a modest 2.4 kbps, 1G offered basic functionality without features like caller ID or roaming capabilities. Transitioning to digital technology, 2G significantly enhanced data transmission rates (up to 64 kbps) and introduced caller ID, roaming, and support for text messaging [2],[3]. While 2G improved network performance, its data transmission speeds are slow by modern standards. Often hailed as a "data revolution," 3G networks delivered data rates of 144 kbps for mobile users, 384 kbps for walking users, and up to 2 Mbps for stationary users. This enabled broadband applications like large email transfers, video streaming, and secure online communication, marking a significant leap in mobile technology [4]-[6]. 4G networks represent a substantial improvement over 3G, achieving speeds up to 1.2 Gbps, approximately 20 times faster than 3G. Operating on the 700 MHz, 1800 MHz, and 2600 MHz frequency bands, 4G enables seamless downloading of large files, streaming of high-quality videos, and new services like mobile payments, video conferencing, and cloud-based gaming. Despite rapid technological advancements, wireless mobile networks in regions like Nigeria face persistent challenges. These include poor coverage, frequent call drops, signal interference, network congestion, and suboptimal quality of service (QoS) [7]. These issues primarily arise from the unpredictable and complex nature of mobile radio channels, which are influenced by environmental factors such as shadowing, path loss, and interference. Effective mitigation of these challenges requires robust network design and accurate prediction models for signal behavior.
services grows, so does the number of wireless communication technologies. This shows how important accurate path loss prediction models are for improving network performance. Traditional empirical models, such as Okumura-Hata and COST-231 Hata, often struggle to deliver reliable predictions in complex urban environments due to their limited adaptability to varying terrain and environmental factors. These limitations result in challenges such as poor coverage, signal interference, and suboptimal Quality of Service (QoS), especially in regions where networks are subject to unpredictable conditions. Existing models are often constrained by static configurations and outdated technologies. This study proposes a novel path loss prediction model using an Adaptive Neuro-Fuzzy Inference System (ANFIS) tailored to a 4G LTE network operating at 2600 MHz within the Zaria Government Reserved Area (GRA) in Nigeria. The model leverages a handheld spectrum analyzer, a GPS distance meter, and walk-test methodology to capture accurate path loss data. By integrating ANFIS with advanced statistical optimization, the proposed model achieves superior predictive performance compared to conventional models such as Okumura-Hata, COST-231 Hata, Egli, and ECC-33. The results demonstrate a significant reduction in error metrics, with a Chi-square error of 1.252063 dB, an RMSE of 0.32145 dB, and the highest Coefficient of Determination (R² = 0.976). These outcomes highlight the model's exceptional accuracy, reliability, and adaptability, paving the way for enhanced wireless network performance optimization and supporting the deployment of next-generation 5G networks in similar urban environments.
Key Words: Path Loss Prediction, Adaptive Neuro-Fuzzy Inference System (ANFIS), 4G LTE, Wireless Communication, Network Performance Optimization.
1.INTRODUCTION The demand for personal communication devices such as smartphones and tablets, combined with the need for
© 2024, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 424