Spatiotemporal Modeling of Short Message Service Traffic Distribution
No Thumbnail Available
Date
2020-02
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Short Message Service (SMS) is a widely used text messaging service on mobile devices.
Business owners and companies look forward to new ways of promoting their services
and products to reach and attract the intended users. However, SMS distribution is not
investigated in Ethio Telecom with time and location. The current way of selecting
customers to send bulk SMS is very costly and resource wastage.
The concern of the thesis is to understand the SMS distribution and develop a
spatiotemporal model of SMS traffic using city of Addis Ababa, Ethiopia as a case study.
The dataset is extracted from SMS traffic of Ethio Telecom network system. The model
selection is performed in time and space domain using Fourier Transform methods. By
applying Fourier Transform, five major frequency components are identified with
different magnitude of coefficients and phases. Then, the coefficient and phase values are
transformed from spatial to spectral-domain by applying Two Dimensional Discrete
Cosine Transform. The tools used are Matlab for traffic modeling, MapInfo to map
geographical location and SPSS for analysis.
Validation of the SMS traffic distribution model is done based on the R
2
coefficient of
determination statistically. The accuracy of the proposed temporal model is evaluated
with real SMS traffic data. The results show that these models can accurately describe the
variation pattern of real SMS traffic. In the spatial model, the effect of the level of
truncation on the frequency of transformed phase and coefficient on the model
performance is evaluated. Finally, a mathematical model that captures the SMS traffic
variation of any location in the city at any time of the day is developed. Furthermore,
correlation analysis is done between outgoing and incoming SMS traffic. The general SMS
usage trend of customers that respond to SMS advertisements was analyzed.The distribution model helps the Company to know the SMS traffic distribution, which
in a way helps to understand the behaviours of its customers thereby saving cost and
reducing resource wastage, and plan its marketing strategy accordingly.
Description
Keywords
SMS Traffic, bulk SMS, Spatiotemporal Modeling, SMS Traffic Distribution