LTE Radio Access Network Dimensioning by Particle Swarm Optimization
No Thumbnail Available
Date
2020-02
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
In this research, Long Term Evolution Radio Access Network Dimensioning by Particle
Swarm Optimization (PSO) in the case of Addis Ababa, Ethiopia has been introduced.
Currently, mobile network service provision demand is increasing, especially customers’
data demand increases time to time. To accommodate this demand, LTE is in hand
solution. Preparing nominal plan after dimensioning is tiresome and error prone. The motive
of this research is to use PSO for addition of new sites to the existing e-NodeBs in order to
simplify LTE planning by considering traffic distribution. Using PSO has good result in order
to find optimal position of new site.
There are many researches done on LTE radio network dimensioning. However, they did
not consider actual customer distribution or traffic density, which will cause inefficient use
of the network resource. With this research, PSO has been checked for LTE dimensioning
in the case of Addis Ababa for LTE RAN planning and congestion optimization that it will
assist to put new sites on high traffic density area.
LTE coverage dimensioning have been done by two ways: by empirical way using Cost231
propagation model and by outdoor planning tool, Mentum planet. LTE capacity
dimensioning was also done by two scenarios: from traffic history prediction and
expectation of the coming big data demand of Internet of Things, Machine to Machine and
e-commerce. Live network traffic history and Time Advance measurement of Global
Solution for Mobile communication have been taken and average traffic probability
distribution function (PDF) and optimal traffic probability density function(pdf) have been
developed. From which, inter-site traffic share models have been formulated.
Finally, model for PSO has been developed. For each new site; searching radius, minimum
inter-site distance and required point by searching optimal points within searching area as
well as total number of sites were determined.
This research shows that number of sites found based on numerical calculation for capacity
is not enough to support expected traffic due to uneven traffic distribution. It confirmed that
using single type of LTE site is both inefficient and insufficient.
Description
Keywords
RAN dimensioning, particle swarm optimization, inter-site traffic share