Abstract
In this paper, the extended Stanford University Interim (ESUI) path loss model was evaluated and optimized based on empirically measured path loss obtained along a suburban road lined with Gmelina Arborea trees. The study was for a 1800 MHz cellular network located in Orlu Imo state. The field measurement was conducted in August which is in the rainy season with the entire trees blossom with their green leaves. G-NetTrack Lite 8.0 Adroid app installed on a Samsung Galaxy S8 mobile phone was then used to capture and log the received signal strength (RSSI) in dB, the geo-coordinates of the measurement points as well as the particulars of the 3G network base station. Two datasets were captured and one f the dataset was used for the model optimization while the other dataset was used for validation of the model. For the training dataset, un-tuned ESUI model had a root mean square error (RMSE) of 46.82 dB and maximum absolute prediction error of 46.64 dB whereas the RMSE-tuned ESUI model had a RMSE of 4.094dB and maximum absolute prediction error of 3.41dB. Similarly, for the validation dataset, the un-tuned ESUI model had a RMSE of 48.37dB and maximum absolute prediction error of 48.19 dB whereas the RMSE-tuned ESUI model had a RMSE of 4.39 dB and maximum absolute prediction error of 3.0 dB. In all, the tuned ESUI model was derived and it gave good path loss prediction performance for both the training and the validation datasets.
Keywords: Propagation loss, Cellular network, SUI model, RMSE-Tuned model, Extended sui model
Received: 10 October 2018 / Revised: 16 November 2018 / Accepted: 24 December 2018/ Published: 14 January 2019
Contribution/ Originality
The paper's primary contribution is the derivation of a root mean square error-based optimized extended Stanford University Interim (ESUI) path loss model for a suburban road lined with Gmelina Arborea trees.
1. INTRODUCTION
In wireless network communication systems, accurate knowledge of the path loss in any given environment is essential for estimating the network coverage area and attainable quality of service in the area [1-6]. As such, researchers and network designers have developed several path loss prediction mathematical models that are suitable for different terrain and situations [7, 8]. In this paper, the Extended Stanford University Interim (ESUI) model developed by 802.16 IEEE group is studied [8, 9]. The ESUI model is a modified version of the Stanford University Interim (ESUI) model that was developed by 802.16 IEEE group in collaboration with Stanford University [10-15] The ESUI is particularly suitable for the suburban areas and for terrains with light to heavy vegetation.
Consequently, in this paper, the ESUI model is used to characterize the path loss in a suburban road that is lined with Gmelina Arborea trees. The ESUI model was evaluated and tuned using a field measured data collected from the case study road. The model prediction performance was expressed in terms of root mean square error and maximum absolute prediction error. The optimized ESUI model was cross-validated by a second dataset captured along the same road. In all, the relevant mathematical expressions, the field measurement campaign procedure , the ESUI model optimization process, as well as the model performance evaluation and cross-validation, are presented. The essence of the study is to derive ESUI-based path loss model that will give better path loss prediction performance for the given tree-lined suburban road.
2. THE EXTENDED STANFORD UNIVERSITY INTERIM MODEL
Table-1. The values of the ESUI constants for the different terrains [9, 16]
Source: Kalu, et al. [9]; Erceg [16]
3. THE FIELD MEASUREMENT CAMPAIGN
The study was carried out on a Gmelina Arborea [17, 18] tree-line road in the suburban site located in Orlu in Imo state. The field measurement was conducted in August which is in the rainy season with the entire trees blossom with their green leaves. The road receives signal from a 3G 1800 MHz cellular network base station that is about 400 meters from the road. G-NetTrack Lite 8.0 Android app was installed on a Samsung Galaxy S8 mobile phone and was then used to capture and log the received signal strength (RSSI) in dB, the geo-coordinates of the measurement points as well as the particulars of the 3G network base station. The field data measured RSSI was used in a link budget equation to determine the measured path loss. Haversine formula was used to determine the distance between the base station and each of the measurement points. The data capture was done two times and one of the datasets was used as the training data for optimizing the ESUI model while the second dataset was used for the cross-validation of the optimized ESUI mode. The field measured RSSI, the measured path loss and their corresponding measurement point distance from the base station for the training and validation datasets are shown in Figure 1 and Figure 2 respectively.
Figure-1. The field measured RSSI versus measurement point distance from the base station for the training and validation datasets
Figure-2. The field measured path loss versus measurement point distance from the base station for the training and validation datasets
4. RESULTS AND DISCUSSION
Table -1. The ESUI model predicted path loss and the optimized ESUI model predicted path loss versus distance for the training dataset
d (km) |
Measured Path Loss (dB) |
Un-Tuned ESUI For Terrain B (dB) |
RMSE-Tuned ESUI For Terrain B (dB) |
0.4542 |
136 |
93 |
140 |
0.4524 |
137 |
93 |
140 |
0.4621 |
136 |
93 |
140 |
0.4697 |
139 |
94 |
141 |
0.4768 |
138 |
94 |
141 |
0.4852 |
138 |
94 |
141 |
0.4964 |
140 |
95 |
142 |
0.504 |
140 |
95 |
142 |
0.5141 |
139 |
95 |
142 |
0.5298 |
141 |
96 |
143 |
0.5379 |
140 |
96 |
143 |
0.5395 |
139 |
96 |
143 |
0.5414 |
140 |
96 |
143 |
0.5433 |
140 |
96 |
143 |
0.5455 |
140 |
96 |
143 |
0.5469 |
142 |
96 |
143 |
0.5496 |
140 |
97 |
143 |
0.5528 |
144 |
97 |
143 |
0.5592 |
143 |
97 |
144 |
0.5643 |
145 |
97 |
144 |
0.5767 |
148 |
97 |
144 |
0.5893 |
146 |
98 |
145 |
0.6013 |
150 |
98 |
145 |
0.6144 |
152 |
99 |
145 |
0.6154 |
152 |
99 |
145 |
0.6173 |
154 |
99 |
145 |
0.6324 |
156 |
99 |
146 |
Figure-3. The ESUI model predicted path loss and the optimized ESUI model predicted path loss versus distance
The un-tuned ESUI model had a RMSE of 46.82 dB and maximum absolute prediction error of 46.64 dB whereas the RMSE-tuned ESUI model had a RMSE of 4.094dB and maximum absolute prediction error of 3.41dB . The validation dataset also gave good prediction performance the ESUI tuned with the RMSE value of 46.64 dB obtained from the training dataset. According to the validation dataset result, the un-tuned ESUI model had a RMSE of 48.37dB and maximum absolute prediction error of 48.19 dB whereas the RMSE-tuned ESUI model had a RMSE of 4.39 dB and maximum absolute prediction error of 3.0 dB .
5. CONCLUSION
The extended Stanford University Interim (ESUI) path loss model was presented. The study was for a 3G network in a suburban area along a road lined with Gmelina Arborea trees. Field measurements were conducted along the road and the measured path loss data was used to optimize the ESUI model for better path loss prediction performance. The data field capture was conducted two times and one of the datasets was used for the model optimization while the second dataset was used for the cross-validation. In all, the root mean square error tuned ESUI model gave good prediction performance for both the training and the validation dataset. Finally, the tuned ESUI model for the case study road was also derived based on the filed data results.
| Funding: This study received no specific financial support. |
| Competing Interests: The authors declare that they have no competing interests. |
| Contributors/Acknowledgement: All authors contributed equally to the conception and design of the study. |
REFERENCES
[1] C. Phillips, D. Sicker, and D. Grunwald, "Bounding the practical error of path loss models," International Journal of Antennas and Propagation, vol. 2012, pp. 1-21, 2012.Available at: https://doi.org/10.1155/2012/754158.
[2] D. Sharma and R. Singh, "Analysis of handover initiation using path loss to sustain QoS," International Journal of Scientific & Engineering Research, vol. 3, pp. 1-6, 2012.
[3] S. I. Popoola, A. A. Atayero, and N. Faruk, "Received signal strength and local terrain profile data for radio network planning and optimization at GSM frequency bands," Data in Brief, vol. 16, pp. 972-981, 2018.Available at: https://doi.org/10.1016/j.dib.2017.12.036.
[4] S. Kumar and D. Lobiyal, "Sensing coverage prediction for wireless sensor networks in shadowed and multipath environment," The Scientific World Journal, vol. 13, pp. 1-11, 2013.Available at: https://doi.org/10.1155/2013/565419.
[5] T. S. Rappaport, S. Sun, R. Mayzus, H. Zhao, Y. Azar, K. Wang, G. N. Wong, J. K. Schulz, M. Samimi, and F. Gutierrez Jr, "Millimeter wave mobile communications for 5G cellular: It will work!," IEEE Access, vol. 1, pp. 335-349, 2013.
[6] H. van den Berg, F. Roijers, M. Castro, C. Gomez, J. Paradells, T. Staub, R. de Oliveira, J. Karlsson, S. Avallone, and P. Hurni, "Multi-hop wireless networks," 2009.
[7] A. Zreikat and M. Dordevic, "Performance analysis of path loss prediction models in wireless mobile networks in different propagation environments," in Proceedings of the 3Rd World Congress on Electrical Engineering and Computer Systems and Science(EECSS'17) Rome, Italy –June 4 –6, 2017, 2017.
[8] G. Senarath, "Multi-hop relay system evaluation methodology (channel model and performance metric)," A Technical Paper of IEEE 802.16 Broadband Wireless Access Working Group2007.
[9] C. Kalu, B. U.-A. Stephen, and M. C. Uko, "Empirical valuation of multi-parameters and RMSE-based tuning approaches for the basic and extended stanford university interim (SUI) propagation models," Mathematical and Software Engineering, vol. 3, pp. 1-12, 2017.
[10] A. HashibSiddique, A. Arifuzzman, and M. Tarique, "Performance study of IEEE 802.16 d under Stanford University Interim (SUI) channel," International Journal of Computer Networks & Communications, vol. 5, pp. 137-152, 2013.Available at: https://doi.org/10.5121/ijcnc.2013.5211.
[11] D. Alam and R. H. Khan, "Comparative study of path loss models of WiMAX at 2.5 GHz frequency band," International Journal of Future Generation Communication and Networking, vol. 6, pp. 11-24, 2013.
[12] V. Abhayawardhana, I. Wassell, D. Crosby, M. Sellars, and M. Brown, "Comparison of empirical propagation path loss models for fixed wireless access systems," in Vehicular Technology Conference, 2005. VTC 2005-Spring. 2005 IEEE 61st, 2005, pp. 73-77.
[13] P. K. Sharma and R. Singh, "Comparative analysis of propagation path loss models with field measured data," International Journal of Engineering Science and Technology, vol. 2, pp. 2008-2013, 2010.
[14] Y. A. Alqudah and A. Tahat, "Path loss and propagation models at 3.5 GHz using deployed WiMAX network," in Information Networking (ICOIN), 2011 International Conference on, 2011, pp. 301-305.
[15] A. Halifa, E. T. Tchao, and J. J. Kponyo, "Investigating the best radio propagation model for 4G-WiMAX networks deployment in 2530MHz Band in Sub-Saharan Africa," Communications on Applied Electronics, vol. 7, pp. 28-36, 2017.
[16] V. Erceg, "Channel models for fixed wireless applications," A Technical Paper of IEEE 802.16 Broadband Wireless Access Working Group2001.
[17] D. Pathala, A. Harini, and P. L. Hegde, "A review on gambhari (Gmelina arborea Roxb.)," Journal of Pharmacognosy and Phytochemistry, vol. 4, pp. 127-132, 2015.
[18] M. Akeem Azeez, J. E. Andrew, and B. B. Sithole, "A preliminary investigation of Nigerian Gmelina arborea and Bambusa vulgaris for pulp and paper production," Woods Science and Technology, vol. 18, pp. 65-78, 2016.
Views and opinions expressed in this article are the views and opinions of the author(s), Review of Computer Engineering Research shall not be responsible or answerable for any loss, damage or liability etc. caused in relation to/arising out of the use of the content. |