Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4, 2877-2892 2025 Publisher: Learning Gate DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate © 2025 by the authors; licensee Learning Gate History: Received: 17 February 2025; Revised: 12 April 2025; Accepted: 17 April 2025; Published: 29 April 2025 * Correspondence: AkinbolatiA@tut.ac.za In-depth study of path loss and coverage areas of selected GSM signals over a typical Sudan Savannah environment Bolanle T. Abe1, Akinsanmi Akinbolati2* 1,2Department of Electrical Engineering, Tshwane University of Technology, eMalahleni Campus, 1034, South Africa; abebt@tut.ac.za (B.T.A); AkinbolatiA@tut.ac.za (A.A.) Abstract: Studies on path loss and coverage areas are essential in GSM technology to ensure Quality of Service (QoS). This research was conducted on MTN and 9-Mobile 4G signals over Zaria City, Nigeria, using the COST-231 model. The measurement campaign was conducted using a drive test; Line of Sight (LOS) from the Transmitting Base Station(s) (TBS), Received Signal Strengths (RSS), elevation, longitude, and latitudes of data points were recorded. Data were collected at 100 m intervals along two routes from the TBS up to 1 km and then at 1 km intervals up to 5 km. The Spectrum Analyzer was used to measure the RSS while the GPS was used to capture the geographic parameters (latitude, longitude, elevation) for the data points. Path losses for the two routes and channels were evaluated with digital maps of coverage areas generated using GIS. Other key findings show strong RSS from TBS until about 4.0 km for both signals; beyond this range, RSS reduces drastically, approaching the sensitivity value. For GSM 4G, an Inter-TBS link of around 4 km is suggested for the study area and any similar city to improve QoS. Furthermore, a strong negative correlation coefficient (R) of -0.83 (mean value) exists between path loss and RSS, while a low R of -0.09 exists between it and elevation. A low level of signal enhancement by elevation is observed with R of +0.06 between it and RSS, and a low negative R of -0.09 between elevation and path loss. The COST-231 was optimized by generating a modified model (ZCOST-231) that incorporates some of the location-based transmission parameters. The optimized model presents valuable applications in the accurate prediction of path losses for GSM 3&4 G channels and Radio Link Design (RLD) over the study areas or any similar Sudan Savannah environment of Africa. The overall findings will significantly benefit radio link appraisal of existing GSM channels, power budgeting, and coverage area assessment in a typical Sudan Savannah City of Africa. Keywords: Coverage areas, GSM signals, Pathloss, Quality of service (QoS), Radio-link-design, Sudan savannah. 1. Introduction Radio communication simply means ‘wireless’ communication. It is the type of communication (sending signals or information from one point to the other) that does not require connecting wires directly from the transmitter to the receiver [1]. Instead, it is the transmission of the radio signal from the transmitter through space (atmosphere) to the receiver. A transmitter is an electronic device that processes, modulates, and amplifies a radio frequency energy and transmits it via the transmitting antenna [1]. A receiver is a radio signal (RF) processing device that periodically scans a given frequency range by performing a frequency ramp on the local oscillator [2] and delivers the inherent message. The signal of interest (data, voice, visual/picture, audio, etc) is generated at the source, processed (filtered, amplified, modulated) and transmitted through the transmitter via the transmitting antenna in the case of terrestrial transmission [1]. In the case of satellite communication, the signal is processed at the source and transmitted using the uplink satellite earth station, which beams the signal to the spacecraft/space station. The space station beams the signal back to the Earth with its coverage area equal to the transmitting mailto:abebt@tut.ac.za mailto:AkinbolatiA@tut.ac.za 2878 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate satellite's footprint. Radio communications include radio and television broadcasting, telephone, walkie- talkie, internet, data transmission, microwave link, etc. Path loss and coverage areas assessment are critical in wireless communications. This helps communication operators and regulatory government agencies appraise existing communication networks and make good link designs and power budgets for future networks to ensure efficient service delivery. This is why studies on coverage areas, path loss assessment, prediction, modelling and investigation of error margins of existing models for suitability in other environments find relevance in wireless communication research. Thus, the need for radio scientists and engineers to conduct studies in this regard becomes imperative [3]. This is the problem, and there is a shortage of such related studies in the selected City. This study will provide the quantitative attenuation profile for the studied GSM networks (MTN&9 Mobile) so that the service providers will be guided towards ensuring quality of service for their subscribers. Friis [4] introduced the Friis or free space equation, a propagation model to predict wireless signals' received power in unobstructed space. However, this model is limited in applicability to other environments beyond free space. Okumura, et al. [5] conducted extensive field strength measurements in Tokyo, Japan, resulting in an empirical formula for predicting power loss in urban areas. This formula considered factors like distance, transmission frequency, and antenna heights. Hata [6] modified Okumura's model to create the Okumura-Hata model, which is widely accepted over the UHF channel [7]. This model considers antenna height, frequency, and environmental conditions and is suitable for predicting path loss in urban, suburban, and rural environments. Other models were also proposed, such as those by COST- 231 [8] and Walfisch and Bertoni [9]. The authors in Sarkar, et al. [10] introduced numerical methods as a third alternative for predicting path losses alongside empirical and site-specific approaches. While empirical models’ applications are simple, site-specific models offer greater accuracy at the cost of requiring specific information about the area. The authors in Ajewole, et al. [11] examined UHF signal strength and its relationship with precipitation in Nigeria. A high negative correlation coefficient was established between UHF TV signal and precipitation. The authors in Prasad and Ahmad [12] conducted field strength measurements over the Indian subcontinent but didn't propose specific models based on the results. Armoogum, et al. [13] studied propagation models in digital television broadcast network design using Mauritius as a case study. They found that Okumura-Hata and COST-231 models gave better agreement. Nisirat, et al. [14] presented a novel terrain correction factor for the Okumura-Hata model, specifically at 900 MHz in Jordan. RMSE shows the predicted improvement by up to 3 dB compared to the Hata suburban model in most areas under study. Authors in Akinbolati and Ajewole [15] worked on investigating path loss and modelling over digital terrestrial television channels in Nigeria. The research areas covered Ikorodu-Lagos, Kaduna, and Katsina. Calculations and comparisons were based on Hata models. Modified- Hata models that incorporate some tropospheric parameters were proposed. The work in [1] analyzed UHF signal strength variations in Ondo State, Nigeria. The results show that elevation above ground level is crucial for UHF transmission and reception (location of transmitters, transmitting antenna’s height, directivity and gain). Akinbolati, et al. [16] worked on the assessment of the error margin of some path loss models over the Digital Terrestrial Television (DTTV) channel using the Star Times DTTV channel within the Katsina metropolis. The predicted COST-231 and ECC models are preferred for path loss prediction over the study areas and channels. Abiodun, et al. [17] evaluated path loss prediction models in microcellular environments of Ekiti State, Nigeria. They proposed a two-slope model that exhibited good agreement with measured values. The study concluded that the COST 231- Hata model could predict path loss in mobile microcell coverage. The focus of this study included the determination of the propagation curves that are useful in the prediction of coverage areas and assessments of the GSM signals. The evaluation of path loss (micro and macro cells) using the widely acceptable COST-231 model and the degree of relationship between path loss and the transmission variables was also carried out. The study also generated digital contour maps of the coverage areas of the two GSM service providers, providing functional parameters and solutions that will enhance QoS over the study area. Furthermore, the Modified COST-231 Model (ZCOST-231 2879 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate Model) that incorporates selected location-based transmission parameters of the environment was developed. 2. Theoretical Analysis 2.1. Path Loss Over the Radio Channel Path loss in radio communication signifies the decline in the intensity of a signal, brought about by an array of elements such as terrain characteristics and atmospheric circumstances encountered along the transmission route or channel. This diminishment can be gauged by the disparity in decibels between the power at the point of transmission and the power upon reception, as expressed in (1) 𝑃𝐿 (dB) = 10𝐿𝑜𝑔 𝑃𝑡 𝑃𝑟 (1) where 𝑃𝑡 𝑎𝑛𝑑 𝑃𝑟 are the transmitted and received power, respectively. This phenomenon illustrates the reduction in signal magnitude due to several factors, including the dissipation of energy as radio waves travel through open space, the reflection of surfaces, the bending of waves around obstacles, and the scattering of waves due to objects in the environment, amongst others [18]. Path loss refers to the typical reduction in signal strength encountered by a transmitted Radio Frequency (RF) signal during its propagation. This reduction becomes evident as the signal covers a distance equivalent to several wavelengths and ultimately arrives at the receiver. Path loss is helpful in the calculation of the link budget of communication systems. It serves as a tool for forecasting and enhancing coverage areas in radio communication [19]. Furthermore, it finds utility in anticipating spatial areas for TV and accommodating secondary users, as highlighted by Faruk, et al. [20]. Path loss prediction encompasses the capability to accurately predict the attenuation of a radio signal as it traverses a communication channel. This entails the computation of the power reduction that occurs from the point of transmission, usually at a base station, to the reception point at the receiver. Path loss prediction quantifies the power dissipation the transmitted signal encounters during its journey between the origin and destination. Diverse prediction models have been formulated for wireless channels. Some are established through statistical analyses of field measurements, while others are developed analytically, considering factors like diffraction effects. These models are equipped with specific parameters tailored to achieve reasonable accuracy in their predictions. 2.2. Path Loss Prediction Models Path loss prediction involves accurately forecasting the impact of channel features on a radio signal, precisely the attenuation effect. It refers to accurately estimating the attenuation impact caused by the propagation channel on a radio signal. It involves predicting, modelling, or calculating the power losses between the transmission at the base station and the reception at the receiver. In wireless communication, various models have been developed to predict path loss. Some are derived statistically based on field measurements, while others are formulated analytically, considering the consequences of diffraction. Friis Transmission Equation or Free Space Model (FSPL): The Friis transmission equation [4] is a streamlined model for predicting path loss in radio wave propagation. Equation 2 is used to estimate path loss (L) between two isotropic antennas in an unobstructed space. (2) Where is the distance (Line of Sight, LOS) from a transmitter in meters, and λ is the wavelength in m. Plane Earth Model: The free space model simplifies by not accounting for the impact of ground reflection on signal propagation. When a radio wave travels over the ground, a portion of its energy is reflected due to the presence of the ground surface, eventually reaching the receiver. In this scenario, significant factors include the heights of the transmission and reception apparatus, the separation distance 2880 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate between them, and the specific attributes of the ground's reflective properties. The path loss equation in the context of the plane earth model is expressed as follows [18]: 𝐿𝑃𝐸 = 40𝑙𝑜𝑔10(𝑑) − 20𝑙𝑜𝑔10(ℎ1) − 20𝑙𝑜𝑔10(ℎ2) (3) Where the distance between the transmitting and receiving antenna in (km), h1 and h2 (m) are the height of the transmitting base station and receiving antenna, respectively. Okumura Model: Okumura's widely employed macroscopic propagation model was formulated during the mid-1960s following comprehensive investigations conducted within Tokyo and its environs. This model was tailored to function within the frequency spectrum from 200 to 1920 MHz, predominantly catering to urban propagation scenarios [5]. Okumura's model posits that the loss of signal strength between the transmitting and receiving ends within a terrestrial propagation environment can be effectively represented by: 𝐿50% = 𝐿𝐹𝑆𝐿 + 𝐴𝑚𝑢(𝑓, 𝑑) − 𝐺(ℎ𝑟𝑒) − 𝐺𝑎𝑟𝑒𝑎 (4) Where: 𝐿50% is the median (i.e., 50th percentile) value of propagation path loss expressed in dB. LFSL is the free space propagation loss in dB. Amu (f,d) is the essential median attenuation relative to free space in dB. G(hte) and G(hre) are the base station and receiver antenna height gain correction factor in dB, respectively, and GAREA is the gain dB due to the type of environment. Hata [6]: The Hata Model for Urban Areas, also called the Hata [6] model, represents an evolved iteration of the Okumura Model and is the most widely adopted radio frequency propagation model. This model is a robust tool for projecting the behaviors of cellular transmission in densely constructed regions [13]. The Hata model comprises distinct variants tailored for specific scenarios: one for sub-urban contexts and another for open areas. With a primary focus on terrestrial microwave communication, the model furnishes insights into the cumulative path loss experienced along a transmission link. The Hata model accommodates micro and macro cell outdoor prediction, rendering it a valuable tool for forecasting radio wave propagation across various scales. Its suitability extends to point-to-point and broadcast transmissions, making it versatile for multiple communication scenarios. Okumura-Hata models for urban and suburban areas are formulated as follows: (5a) For a large city with the wave frequency of transmission f ≥ 400 MHz, (5b) Where L(urban) is the path loss in Urban Areas in dB, hb is the height of the base station antenna in meters, hm is the height of the mobile station antenna in meters, f is the frequency of transmission in MHz, α(hm) the antenna height correction factor and d, is the line-of-sight distance between the base and mobile stations in (km). By specifications, the Okumura-Hata model has the following range for optimal results: Carrier frequency: 150MHz ≤ f ≤ 1500MHz; Base station height: 30m ≤ hb ≤ 200m; Mobile station height: 1m ≤ hm ≤ 10m; Distance between mobile and base station: 1km ≤ d ≤ 20km [14, 16]. For suburban areas, it is given by: (5c) The European Committee for Scientific and Technical Research Team (COST-231) Model: The COST-231 Hata model, also recognized as the Personal Communication System (PCS) Extension, is a prevalent tool for forecasting path loss within mobile wireless systems. Crafted by the European Co- operative for Scientific and Technical Research Team, this model has garnered substantial recognition in wireless communications. It represents an expanded version of the Okumura-Hata model, encompassing an extensive frequency spectrum from 0.5 to 2 GHz and suitable for medium to small cities [21]. It is a widely acceptable model for path loss and coverage assessment over wireless communication channels. The model's formulation is as presented in (6): 2881 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate 𝐿(𝑑𝐵) = 46.3 + 33.9 𝑙𝑜𝑔10(𝑓) − 13.82 𝑙𝑜𝑔10(ℎ𝑏) + [44.9 − 6.55 𝑙𝑜𝑔10(ℎ𝑏)] 𝑙𝑜𝑔10(𝑑) − 𝛼(ℎ𝑚) + 𝐶𝑚 (6) 𝑎(ℎ𝑚) is the receiver antenna correction factor as presented in (5b), while 𝐶𝑚 = 0 𝑑𝐵 for medium-sized city and suburban areas and 3 𝑑𝐵 for urban areas. The parameters ℎ𝑚 ℎ𝑏 𝑓𝑐 are presented in Table 1. This model was used for computation in this study. Its choice was because most results of similar studies in Nigeria predicted the COST- 231 model as the most preferred to be used on 3G/4G frequencies [22-28]. Internationally, COST-231 is gaining acceptability for path loss prediction over microcell outdoor prediction [29, 30]. 3. Materials and Methods The study was conducted in the urban city of Zaria, Nigeria (Latitude 11.085° N and Longitude 7.720° E). In terms of total land area, Zaria covers about 563 km2 with a population of 736,000 people as of 2006 [31]. Zaria has a tropical savannah climate with warm weather year-round, a wet season lasting from April to September, and a drier season from October to March. The experimental study was conducted on two GSM operators: MTN and 9-Mobile signals. Table 1 presents the transmission characteristics of experimental stations. Table 1. Transmission characteristics of the experimental Transmitting Base stations. S/N Parameters MTN Transmitting Base Station (TBS) 9-Mobile Transmitting Base Station (TBS) 1 Base station geographical coordinates (Lat.11.108°N, Long. 7.718° E) (Lat.11.158°N, Long. 7.652° E) 2 Base station rated power/transmitted output 36 dBm 36 dBm 3 Transmitting antenna’s gain 17 dBi 17 dBi 4 TBS frequency (𝑓𝑐) 2115 MHz 2175 MHz 5 Height of receiving/mobile antenna, ℎ𝑚 1.5 m 1.5 m 6 Height of transmitting/ base station antenna, ℎ𝑏 55 m 55 m 7 GSM band/ Technology (2100-2300 MHz)/3-4G (2100-2300 MHz)/3-4G 3.1. Method of Data Collection The Measurement of RSS from the Transmitting Base Stations (TBS) under investigation was conducted along various routes, with the transmitting station logged and served as the reference point throughout the measurement period using drive test protocol. Measurements were taken along two routes at 100 m intervals up to 1 km (microcell) and 1 km intervals up to 5 km (macro cell). The mean of the routes was taken and used for the study. Figures 1 and 2 present the Spectrum Analyzer (BK Precision2658A; 5 kHz-8.5 GHz) and GPS used to measure the RSS and geographic coordinates of data points, respectively. Figure 3 is the digital map indicating data location points. 2882 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate Figure 1. Spectrum Analyzer Used for RSS. Figure 2. GPS Garmin map 78s. 2883 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate Figure 3. The Digital Map indicating sampling points for the macro cell over Zaria City 3.2. Empirical and Statistical Tools The European Union Committee for Scientific and Technical Research Team model [32] a widely acceptable model for path loss prediction over the 3/4G-GSM channel, as presented in (6), was used for computation in this study. Karl Pearson’s Product Moment Correlation Coefficient for continuous data, presented in (7), was used to determine the degree of relationship between the variables and path loss, where x and y are the variables of interest. ( ) ( ) 22 2 cov var .var 2 n xy x y XY r X Y n x x n y y − = =    −            (7) 4. Results and Discussions 4.1. Spatial Distribution of RSS And Evaluated Pathloss of the Two Signals Tables 2 and 3 present the overall average of measured parameters (during the fieldwork) and the computed path loss based on the COST-231 model over the study areas for MTN and 9-mobile signals, respectively. From Tables 2 and 3, the mean values of the elevation, RSS, and path loss are 656.53 m, - 75.47 dBm, and 133.841 dB, respectively, for MTN and 683 m, -71.68 dBm, and 132.175 dB, for 9-mobile networks. 2884 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate Table 2. Measured parameters and evaluated Path Loss for MTN 4 G. S/N LOS from TBS (km) Lat. (°N) Long. (°E) Elevation (m) RSS1 (dBm) RSS2 (dBm) Mean RSS (dBm) Path loss (dB) 1 0.0 11.108 7.718 662 -55.89 -57.25 -56.57 100.305 2 0.1 11.065 7.431 639 -62.98 -64.03 -63.505 102.764 3 0.2 11.064 7.431 638 -75.52 -72.41 -73.965 113.212 4 0.3 11.156 7.719 637 -74.43 -78.3 -76.365 119.324 5 0.4 11.105 7.718 637 -73.82 -74.77 -74.295 123.659 6 0.5 11.104 7.719 634 -79.81 -80.75 -80.28 127.023 7 0.6 11.103 7.186 633 -77.7 -76.97 -77.335 129.771 8 0.7 11.103 7.718 932 -76.21 -75.63 -75.92 132.095 9 0.8 11.101 7.718 631 -78.79 -79.75 -79.27 134.107 10 0.9 11.100 7.718 630 -73.92 -76.56 -75.24 135.883 11 1.0 11.099 7.718 628 -75.84 -71.81 -73.825 137.471 12 2.0 11.093 7.718 636 -76.91 -73.95 -75.43 147.918 13 3.0 11.087 7.722 643 -79.01 -79.16 -79.085 154.029 14 4.0 11.082 7.732 640 -84.99 -86.2 -85.595 158.366 15 5.0 11.080 7.739 628 -85.94 -84.91 -85.425 161.729 Mean 656.53 -75.45 -75.45 -75.47 133.841 Table 3. Measured parameters and evaluated Path Loss for 9-Mobile 4G. S/N LOS from TBS (km) Lat. (°N) Long. (°E) Elevation (m) RSS1 (dBm) RSS2 (dBm) Mean RSS (dBm) Path loss (dB) 1 0.0 11.158 7.652 683 -51.34 -54.8 -53.07 99.520 2 0.1 11.159 7.651 684 -59.65 -53.69 -56.67 103.175 3 0.2 11.159 7.650 685 -62.87 -64.74 -63.805 113.622 4 0.3 11.159 7.650 684 -68.91 -72.87 -70.89 119.734 5 0.4 11.160 7.649 686 -73.68 -70.94 -72.31 124.070 6 0.5 11.160 7.648 685 -73.76 -75.86 -74.81 127.433 7 0.6 11.161 7.648 685 -64.56 -64.61 -64.585 130.181 8 0.7 11.161 7.648 684 -70.37 -67.81 -69.09 132.505 9 0.8 11.161 7.646 685 -78.82 -78.22 -78.52 134.518 10 0.9 11.161 7.645 685 -73.32 -74.75 -74.035 136.293 11 1.0 11.162 7.644 684 -74.23 -72.49 -73.36 137.881 12 2.0 11.150 7.666 679 -76.16 -76.42 -76.29 148.329 13 3.0 11.146 7.673 676 -79.61 -78.58 -79.095 154.440 14 4.0 11.141 7.681 682 -82.82 -83.23 -83.025 158.776 15 5.0 11.138 7.689 678 -85.00 -86.32 -85.66 162.140 Mean 683.01 -71.67 -71.69 -71.68 132.175 4.2. Propagation Curves of The Signals and Digital Map Indicating Coverages Figure 4 shows the propagation curves of the two signals over the distance investigated. It could be observed that both signals depreciated gradually with distance. However, within the micro-cell (0-1km), the signals depreciated and improved again between 1-3 km of the macro cell and later depreciated around 4 km. The reason could be that line-of-sight was not adequately established between the transmitting antenna and the receiver in the micro-cell for this location. Perhaps it was due to some high-rise terrestrial features close to the TBS. From 1-3 km, LOS could now be established appropriately, thus improvement in the signal received. Beyond 3 km from the base station, both signals depreciated. The effects of wave divergence and the line-of-sight distance are manifesting. It can equally be observed that both signals from the two operators are nearly the same in their propagation trajectory. This indicates that the two operators operate within the same transmission characteristics permitted by the Communication’s 2885 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate regulator. One of the critical applications of this result is that an approximate distance of 4 km could be used for inter-TBS-link design in the study area or any similar city. This assertion is based on the propagation curves and the standard quality of 4G presented in Table 4 [33]. The signal is considered fair when RSS ≤ -85 dBm and poor when RSS ≤ -86 dBm. Here, the distance corresponding to -86 dBm is 4 km. The sensitivity value for a 4G signal at which no signal will be received is RSS ≤ -120 dBm. Figure 4. Propagation curves for the two 4G-GSM signals. Table 4. Classification of signal qualities according to 3G and 4G GSM [33]. S/N RSSI 3G (dBm) RSSI 4G (dBm) Signal Strengths Quality Description 1 ≥ -70 ≥ -65 Excellent Strong signal with maximum data speed 2 -86 ≤ RSS ≤ -70 -75 ≤ RSS ≤ -65 Good Strong signal with good data speed 3 -100 ≤ RSS ≤ -86 -85 ≤ RSS ≤ -75 Fair Fair, fast and reliable data speed could be achieved, but drop-out is possible 4 -110 ≤ RSS ≤ -100 -119 ≤ RSS ≤ -85 Poor Performance drops drastically 5 RSS≤ -110 RSS ≤ -120 No signal Sensitivity value (Disconnection) The collected data were interpolated using GIS software to create contour maps. Kriging interpolation techniques generated smooth, continuous surfaces representing the different grades of coverage-areas across the city. The resulting maps show regions of poor, weak, and strong signal strengths, this will find valuable applications for planning signal enhancement and the choice of quality of reception by GSM operators and subscribers. Figures 5 and 6 present the digital map for MTN and 9- mobile coverage profiles, respectively. -90 -80 -70 -60 -50 -40 -30 -20 -10 0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 R SS ( d B m ) LOS (km) MTN RSS 9-Mobile RSS 2886 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate Figure 5. Digital Map for MTN 4 G indicating the coverage profiles. Figure 6. Digital Map for 9mobile 4 G indicating the coverage profile. 2887 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate 4.3. The Path Loss Assessment and Modeling over the Study Area For the path loss analysis, Figure 7 depicts the comparison plots of the path loss associated with the two signals with distance. Mean values of 133.841 and 132.175 dB were obtained for MTN and 9-Mobile, respectively, while an overall mean of 133.008 dB was determined. The path-loss follows the conventional pattern of increase as distance increases, and the two are nearly the same. The path loss profiles justify the need for another transmitting base station at around 4 km intervals to ensure reception quality over the study locations. Figs. 8 and 9 present the variation of RSS and Path loss with LOS for MTN and 9- Mobile signals, respectively. It could be clearly seen that as the transmitter-receiver distance (LOS) increases, RSS reduces, and path loss increases for both signals. Figure 7. Variation of Path Loss over distance for both MTN and 9-mobile GSM Networks. Figure 8. Variation of Path Loss and RSS over distance for MTN signal. 0 20 40 60 80 100 120 140 160 180 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 P at h lo ss in ( d B ) LOS (km)Path loss MTN Path loss 9-mobile -100 -50 0 50 100 150 200 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 2 3 4 5P at h L o ss a n d R SS LOS from TBS MTN Path loss(db) RSS(dBm) 2888 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate Figure 9. Variation of Path Loss and RSS over distance for 9-Mobile signal. 4.4. The Correlation Coefficient Between Path Loss and Some Location-Based Parameters The degree of relationship between path loss and some of the location-based transmission parameters (LOS, Elevation and RSS) for MTN and 9-Mobile were determined and presented in Tables 5 and 6, respectively. Correlation coefficient (R) of +0.88, -0.05 and -0.83 exists between path loss and LOS, elevation and RSS, respectively, for MTN. Similarly, correlation coefficient (R) of +0.95, -0.13 and -0.83 exists between path loss and LOS, elevation and RSS, respectively, for 9-mobile. This shows a strong positive R of 0.92 (mean value) between GSM Path loss and LOS, meaning that path loss increases by about 92% over the Sudan Savannah City as LOS increases. This result will serve as a good guide for wireless link design over the study area. Table 5. Degree of the relationship between path loss and some location-based transmission parameters for MTN 4 G. LOS (km) Elevation (m) RSS (dBm) Path Loss (dB) LOS (km) 1 Elevation (m) -0.13 1 RSS (dBm) -0.66 0.06 1 Path Loss (dB) 0.88 -0.05 -0.83 1 Table 6. Degree of the relationship between path loss and some location-based transmission parameters for 9-Mobile 4 G. LOS (km) Elevation (m) RSS (dBm) Path Loss (dB) LOS (km) 1 Elevation (m) -0.13 1 RSS (dBm) -0.87 0.05 1 Path Loss (dB) 0.95 -0.13 -0.83 1 In addition, a strong negative R of -0.83 (mean value) exists between path loss and RSS, while a low R of -0.09 exists between it and elevation. The implication is that the higher the RSS, the lower the path loss. A low level of signal enhancement by elevation is observed in this study, with a low +0.06 (mean value) between it and RSS and a low negative R of -0.09 between elevation and path loss. The study area -100 -50 0 50 100 150 200 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 2 3 4 5 P at h lo ss a n d R SS LOS from TBS 9-Mobile RSS(dBm) 9-mobile Path loss(dB) 2889 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate has a balanced land terrain, with an averagely high elevation with mean values of 669.5 m and a low deviation from one location to the other. 4.5. Optimization of COST-231 Model by Incorporating Local Transmission Parameters of the Environment COST-231 model was optimized by generating a local path loss model that incorporates some of the location-based transmission parameters to serve as enhanced alternative to the COST-231 model, a linear regression was generated, using the path loss as the dependent variable and the parameters of RSS, LOS, and Elevation as independent variables. Thus, a modified COST-231 model was obtained. Table 8 presents the parameters used for modelling. Table 8. Mean values of dependent and independent variables used for modelling. S/N Path Loss (dB) LOS (km) RSS (dB) Elevation (m) MTN 1. 102.76 0.1 - 63.50 639 2. 127.50 0.5 - 80.28 634 3. 137.471 1.0 -73.825 628 4. 147.918 2.0 -75.43 636 5. 154.029 3.0 -79.085 643 6. 158.366 4.0 -85.595 640 7. 161.729 5.0 -85.425 628 9 – Mobile 1. 103.17 0.1 -56.67 684 2. 127.43 0.5 -74.81 685 3. 137.881 1.0 -73.36 684 4. 148.329 2.0 -76.29 679 5. 154.440 3.0 -79.095 676 6. 158.776 4.0 -83.025 682 7. 162.140 5.0 -85.66 678 Equation 8 represents the general format of the expected path loss model developed in this study. It is the Modified COST -231 Path Loss (ZCOST-231Model) for Zaria City. Equations (9) and (10) represent the actual models for MTN and 9-mobile with coefficients of multiple determination (𝑅2) of 0.82 and 0.88, respectively. Equation (11) represents the mean model for the two, which is the real predicted model for the environment with 𝑅2= 0.85. This indicates a strong and reliable relationship between the dependent and independent variables. This model (s) can be used for the accurate prediction of path loss of 3G ang 4G signals in the study area or any environment with similar terrain and climate. (8) Where: • LOS (km) is Tx-Rx distance • RSS (dBm) is Received Signal Strength • Elev(m) is elevation • β1, β1 and β3 are the model coefficients • k is a constant ZCOST − 231(dB) = 10.915𝐿𝑂𝑆 − 0.098𝑅𝑆𝑆 + 0.619𝐸𝑙𝑒𝑣 − 283.721𝑘 (9) ZCOST − 231(dB) = 9.335𝐿𝑂𝑆 − 0.132𝑅𝑆𝑆 − 1.416𝐸𝑙𝑒𝑣 + 1074 (10) ZCOST − 231(dB) = 10.125𝐿𝑂𝑆 − 0.115𝑅𝑆𝑆 − 0.399𝐸𝑙𝑒𝑣 + 395.15 (11) 5. Conclusion Path loss and coverage areas’ assessment of 4G-9-mobile and MTN signals along a 5 km distance over Zaria, Nigeria, have been investigated successfully. The Received Signal Strength (RSS) and 2890 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate geographic parameters such as latitude, longitude, elevation, and line of Sight (LOS) for data points were measured using Spectrum Analyzer and Global Positioning System (GPS), respectively. Measurements were carried out along two routes within the range of 0 – 5 km (Line of Sight) from the transmitter’s base station. These parameters were used to analyze path loss and generate the digital map for the coverage areas. The digital maps show regions of poor, weak, and strong signal strengths that are useful for planning signal enhancement and quality of reception by the GSM operators and subscribers, respectively. Other key findings show that strong RSS were recorded from the base station till about 4.0 km for both signals. Beyond these distances, the RSS reduces drastically, approaching the sensitivity value with greater path loss values recorded. The implication is that the signal quality would not always be reliable beyond this distance from the TBS. For GSM 3G and 4G, an inter-transmitting base station (TBS) link of around 4 km is suggested in the study area and in any environment with similar terrain and climate to improve signal quality. In addition, the mean values of the elevation, overall RSS and path loss are 656.53 m, -75.47 dBm and 133.841 dB, respectively, for MTN and 683 m, -71.68 dBm and 132.175 dB, for 9- mobile networks. In addition, a strong negative R of -0.83 (mean value) exists between path loss and RSS, while a low R of -0.09 exists between it and elevation. The implication is that the higher the RSS, the lower the path loss. A low level of signal enhancement by elevation is observed in this study, with a low +0.06 (mean value) between it and RSS and a low negative R of -0.09 between elevation and path loss. Furthermore, the Modified COST-231 Model, which incorporates selected location-based transmission parameters of the environment ZCOST-231 Model, was developed. This model presents valuable application in the accurate prediction of path losses and GSM Radio-Link-Design over the study areas or any similar Sudan Savannah environment of Africa. The overall findings of this study will be of great benefit at ensuring accurate appraisal, power budgeting, and coverage areas’ assessment by all stakeholders, to ensure the quality of services of GSM transmission and reception in the study areas and in similar environments in general. Funding: This research received no external funding Transparency: The authors confirm that the manuscript is an honest, accurate, and transparent account of the study; that no vital features of the study have been omitted; and that any discrepancies from the study as planned have been explained. This study followed all ethical practices during writing. Acknowledgment: Authors wish to thank the department of Telecommunications Engineering, Ahmadu Bello University Zaria, for the spectrum analyzer used. The support of Tshwane University of Technology, South Africa is equally acknowledged. Copyright: © 2025 by the authors. This open-access article is distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). References [1] A. Akinbolati, O. Akinsanmi, and K. Ekundayo, "Signal strength variation and propagation profiles of UHF radio wave channel in Ondo state, Nigeria," International Journal of Microwave and Wireless Technologies, vol. 6, no. 4, pp. 12-27, 2016. https://doi.org/10.5815/ijigsp.2023.06.01 [2] A. Akinbolati, M. O. Ajewole, A. T. Adediji, and J. S. Ojo, "The influences of meteorological parameters on digital terrestrial television (DTT) signal in the tropics," International Journal of Digital Information and Wireless Communications, vol. 7, no. 3, pp. 161-173, 2017. https://creativecommons.org/licenses/by/4.0/ https://doi.org/10.5815/ijigsp.2023.06.01 2891 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate [3] R. Panteyev, F. Yanovsky, A. Mykolushko, and V. Shutko, "Radio Receiver with Internal Compression of Input Signals Using a Dispersive Delay Line with Bandpass Filters," International Journal of Image, Graphics and Signal Processing (IJIGSP), vol. 15, no. 6, pp. 1-12, 2023. [4] H. T. Friis, "A note on a simple transmission formula," Proceedings of the IRE, vol. 34, no. 5, pp. 254-256, 1946. [5] T. Okumura, E. Ohmori, and K. Fukuda, "Field strength and its variability in VHF and UHF land mobile service," Review Electrical Communication Laboratory, vol. 16, pp. 824–870, 1968. [6] M. Hata, "Empirical formula for propagation loss in land mobile radio services," IEEE Transactions on Vehicular Technology, vol. 29, no. 3, pp. 317-325, 2013. [7] A. Akinbolati, M.O. Ajewole, A. T. Adediji, and J. S. Ojo, "Propagation curves and coverage areas of digital terrestrial television base stations in the tropical zone," Heliyon, vol. 6, no. 3, p. e03599, 2020. [8] COST- 231, Urban transmission loss models for mobile radio in the 900 and 1800 MHz bands (revision 2)”. In: COST 231 TD (90) 119 Rev. 2. Netherlands: The Hague, 1991. [9] J. Walfisch and H. L. Bertoni, "A theoretical model of UHF propagation in urban environments," IEEE Transactions on Antennas and Propagation, vol. 36, no. 12, pp. 1788-1796, 1988. [10] T. K. Sarkar, Z. Ji, K. Kim, A. Medouri, and M. Salazar-Palma, "A survey of various propagation models for mobile communication," IEEE Antennas and Propagation Magazine, vol. 45, no. 3, pp. 51-82, 2003. [11] M. O. Ajewole, A. Akinbolati, A. T. Adediji, and J. S. Ojo, "Precipitation effect on the coverage areas of terrestrial UHF television stations in Ondo state, Nigeria," International Journal of Engineering and Technology, vol. 4, no. 9, pp. 524-535, 2014. [12] M. Prasad and I. Ahmad, "Comparison of some path loss prediction methods with VHF/UHF measurements," IEEE Transactions on Broadcasting, vol. 43, no. 4, pp. 459-486, 1997. [13] V. Armoogum, K. Soyjaudah, N. Mohamudally, and T. Fogarty, Propagation models and their applications in digital television broadcast network design and implementation. New York: Christos Bouras, 2010. [14] M. A. Nisirat, M. Ismail, L. A. Nissirat, and S. Alkhawaldeh, "A terrain roughness correction factor for hata path loss model at 900 MHz," Progress In Electromagnetics Research C, vol. 22, pp. 11-22, 2011. [15] A. Akinbolati and M. O. Ajewole, "Investigation of path loss and modeling for digital terrestrial television over Nigeria," Heliyon, vol. 6, no. 6, p. e04101, 2020. [16] A. Akinbolati, I. Shaibu, and C. Okpala, "Investigation of error margin of some path loss models over digital terrestrial television channel in Katsina Metropolis," Nigerian Journal of Theoretical and Environmental Physics, vol. 1, no. 1, pp. 1- 8, 2023. [17] C. Abiodun, S. Azi, J. Ojo, and P. Akinyemi, "Assessment of path loss prediction models for wireless propagation channels at l-band frequency over different micro-cellular environments of Ekiti State, Southwestern Nigeria," International Journal of Electronics and Communication Engineering vol. 11, pp. 1103-1109, 2017. [18] S. Ranvier, Path loss models”, S-72.333 Physical layer methods in wireless communication systems. Finland: Radio Laboratory/TKK, 2004. [19] A. Akinbolati and O. J. Agunbiade, "Assessment of error bounds for path loss prediction models for TV white space usage in Ekiti State, Nigeria," International Journal of Information Engineering and Electronic Business (IJIEEB), vol. 12, no. 3, pp. 28-39, 2020. [20] N. Faruk, A. Ayeni, and Y. A. Adediran, "On the study of empirical path loss models for accurate prediction of TV signal for secondary users," Progress In Electromagnetics Research B, vol. 49, pp. 155-176, 2013. [21] M. Mollel and K. Michael, "Comparison of empirical propagation path loss models for mobile communication," Computer Engineering and Intelligent Systems, vol. 5, pp. 1–10, 2014. [22] A. Ayeni, N. Faruk, L. Olawoyin, M. Y. Muhammad, and M. I. Gumel, "Comparative assessments of some selected existing radio propagation models: A study of Kano City, Nigeria," European Journal of Scientific Research, vol. 70, no. 1, pp. 120-127, 2012. [23] S. I. Popoola and O. F. Oseni, "Empirical path loss models for GSM network deployment in Makurdi, Nigeria," International Refereed Journal of Engineering and Science, vol. 3, no. 6, pp. 85-94, 2014. [24] L. Akinyemi, N. Makanjuola, O. Shoewu, and F. Edeko, "Evaluation and analysis of 3G network in Lagos metropolis, Nigeria," International Transaction of Electrical and Computer Engineers System, vol. 2, no. 3, pp. 81-87, 2014. [25] C. Umesi, S. Ilo, K. Akpado, and U. S. Osondu, "Analysis of artificial neural network model against empirical model (Okumara-Hata Model) for predicting pathloss in GSM network," Invention Journal of Research Technology in Engineering & Managemen, vol. 3, no. 3, pp. 33-40, 2019. [26] I. C. Abiodun and J. Idoko, "The study of empirical path loss models for accurate prediction of GSM signal for applications in microcellular environments," European Journal of Science, Innovation and Technology, vol. 2, no. 4, pp. 1- 3, 2022. [27] A. Akande, F. Semire, Z. Adeyemo, and C. Agubor, "A modified empirical path loss model for 4G LTE network in Lagos, Nigeria," Iranian Journal of Electrical and Electronic Engineering, vol. 19, no. 2, pp. 2081-2081, 2023. [28] O. A. Aremu, A. Olaosebikan, O. S. Makinde, and L. Oyinkanola, "Optimization of COST-231 Hata Radio Propagation Prediction Model for GSM 4G FDD-LTE operating at 1800 MHz in tropical region," Quest Journals Journal of Electronics and Communication Engineering Research, vol. 9, no. 1, pp. 09-16, 2023. 2892 Edelweiss Applied Science and Technology ISSN: 2576-8484 Vol. 9, No. 4: 2877-2892, 2025 DOI: 10.55214/25768484.v9i4.6689 © 2025 by the authors; licensee Learning Gate [29] R. Mardeni and T. S. Priya, "Optimised COST-231 Hata models for WiMAX path loss prediction in suburban and open urban environments," Modern Applied Science, vol. 4, no. 9, pp. 1-10, 2010. [30] S. A. Mawjoud, "Path loss propagation model prediction for GSM network planning," International Journal of Computer Applications, vol. 84, no. 7, pp. 1-10, 2013. [31] The National Population Commission of Nigeria, "The national population commission of Nigeria," Retrieved: https://nationalpopulation.gov.ng, 1988. [32] COST-231, European cooperation in the field of scientific and technical research: The COST-231 Final Report. Europ: European Union, 1999. [33] A. F. Akpaneno and A. Akinsanmi, "Spatial variation of the received signal strength of mobile telephone network (MTN) over Dutsin-Ma town, Katsina State, Nigeria," Journal of Basics and Applied Sciences Research, vol. 2, no. 1, pp. 28-34, 2024. https://doi.org/10.33003/jobasr-2024-v2i1-9 https://nationalpopulation.gov.ng/ https://doi.org/10.33003/jobasr-2024-v2i1-9