View Article |
A quasi-moment-method-based calibration of basic pathloss models
Mowete, Ike1, Adelabu, Michael Adedosu2, Ayorinde, Ayotunde Abimbola3, Muhammed, Hisham Abubakar4, Okewole, Francis Olutinji5.
Using a technique similar to Harrington’s method of moments, this paper develops a very simple but remarkably efficient approach to the calibration of established (basic) mobile radio propagation pathloss models. First, the theoretical foundations of the process, here referred to as the ‘Quasi-Moment-Method (QMM)’, is succinctly presented. Thereafter, for validation purposes, pathloss predictions due to its use are compared with corresponding data reported in the open literature, for a model that derived from the application of the Adaptive Neuro-Fuzzy Inference System, ANFIS. Results of the comparisons reveal that the root-mean square error (RMSE) values for the QMM-models compare favorably with those reported for the more computationally involved ANFIS model; and that all the six QMM-calibrated models considered in the paper, provided better spread-correlated root-mean-square (SC-RMS) and standard deviation (SD) prediction errors. QMM cross-application prediction performance is also evaluated through comparisons with measurement data obtained by the authors, for the Nigerian cities of Ibadan and Abuja. Outcomes of the comparisons clearly show that the QMM crossapplication performance, particularly for the calibrated ECC-33 models, may be described as excellent.
Affiliation:
- University of Lagos, Nigeria
- University of Lagos, Nigeria
- University of Lagos, Nigeria
- University of Lagos, Nigeria
- University of Lagos, Nigeria
Download this article (This article has been downloaded 69 time(s))
|
|
Indexation |
Indexed by |
MyJurnal (2021) |
H-Index
|
1 |
Immediacy Index
|
0.000 |
Rank |
0 |
Indexed by |
MyAIS, DOAJ |
Impact Factor
|
5 yrly IF MyAIS=0.02, h index= 1 |
Rank |
0 |
|
|
|