EVALUATION OF RAIN EFFECT ON THE QUALITY OF SERVICE OF WIRELESS LOCAL AREA NETWORK USING FRACTAL DIMENSION IN MUBI METROPOLIS

Abstract

This study, titled "Evaluation of Rain Effect on the Quality of Service of Wireless Local Area Network Using Fractal Dimension in Mubi Metropolis," examines the influence of seasonal rainfall on the performance of Wireless Local Area Networks (WLANs). The analysis focuses on key Quality of Service (QoS) parameters, including throughput, latency, and traffic volume, while employing fractal dimension (F.D) and Hurst index measurements to assess the self-similarity, complexity, and predictability of network traffic patterns. Data was collected using Wireshark protocol analysis software over four months—two months each for the rainy and dry seasons. A drive test format was applied, capturing network traffic for six hours a day, two days per week. The Transport Control Protocol (TCP) data was filtered, exported to Excel, and then analyzed using the SELQOS tool to determine the Autocorrelation Function, Hurst index, and fractal dimensions, providing a comprehensive understanding of network behavior across different environmental conditions. Results indicate significant seasonal variations in network performance. During the rainy season, the total captured traffic was 6,978,936 packets, compared to 4,764,884 packets in the dry season. June recorded the highest throughput (123.88 bps) and a low latency of 0.008 seconds, while July saw a decline in throughput (37.67 bps) and an increase in latency (0.027 seconds). The Hurst index values during the dry season ranged from 0.65 to 0.72, indicating strong long-range dependence and predictable traffic patterns. Conversely, rainy season values fluctuated between 0.58 and 0.63, reflecting reduced network stability. Similarly, higher fractal dimension values (e.g., 1.329 for R/S and 1.346 for A.M in December) were observed in the dry season, highlighting consistent traffic behavior, whereas lower values (e.g., 1.257 for R/S in May) indicated sporadic traffic during rainfall. These findings underscore the negative impact of rainfall on WLAN performance, resulting in increased packet loss, latency, and reduced throughput. To mitigate these effects, the study recommends adaptive network strategies such as dynamic power control, error correction mechanisms, and signal amplification technologies to maintain consistent QoS across varying weather conditions.