Design of an NB-IoT Smart Metering solution Coverage and capacity planning.pdf

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/359501581
Design of an NB-IoT Smart Metering solution: Coverage and capacity
planning: Case of Yaoundé and Douala
Article in International Journal of Computer Applications · March 2022
DOI: 10.5120/ijca2022921972
CITATIONS
5
READS
813
4 authors, including:
Eric DEUSSOM
National Advanced School of Engineering Cameroon
34 PUBLICATIONS 104 CITATIONS
SEE PROFILE
David Magoudaya
SUP'PTIC
1 PUBLICATION 5 CITATIONS
SEE PROFILE
Michael Sone
University of Buea
26 PUBLICATIONS 91 CITATIONS
SEE PROFILE
All content following this page was uploaded by Eric DEUSSOM on 27 March 2022.
The user has requested enhancement of the downloaded file.

International Journal of Computer Applications (0975 – 8887)
Volume 184 – No.2, March 2022
20
Design of an NB-IoT Smart Metering solution: Coverage
and capacity planning: Case of Yaoundé and Douala
Deussom Djomadji Eric Michel
College of Technology
Department of Electrical and Electronic
Engineering
Magoudaya David
National Advanced School of Posts,
Telecommunications and Information and
Communication Technologies
Feudjio Cyrille
Engineering
Michael Ekonde Sone
Engineering
ABSTRACT
Advances in technology over the years have enabled the
Internet of Things to seize the untapped opportunities of
information and communication technologies. NB-IoT, one of
the LPWA technologies, is attracting a lot of attention of
universities and telecommunications companies. This
technology, standardized by 3GPP in June 2016, offers
advantages such as faster and simpler deployment using the
existing cellular network, a large coverage area, low cost and
low power consumption. It has great potential to meet the
enormous demand for machine type communication in the age
of the Internet of Things. The smart meter is an application
that potentially uses NB-IoT technology for water and energy
management. In the water and energy sector, the reading of
water and electricity consumption is done manually especially
in country like Cameroon. The meter reader staff enters
individuals' homes or businesses and reads the meters. Under
these conditions, a rigorous and regular reading of meter
indices proves to be restrictive, costly and even impossible;
and there is a big possibility for spreading COVID 19 virus.
This process poses a problem, of bills that do not correspond
to actual consumption. This research document proposes an
innovative solution to this societal problem, a smart meter
based on NB-IoT. This study aims to analyze the planning of
NB-IoT network in terms of coverage and capacity by relying
on an LTE network in 1800MHz band. The results of the
calculations show that the number of sites required is different
both in terms of coverage and capacity, and the highest
number of sites has been taken to meet the requirements.
Thus, the number of sites required is 19 for the city of Douala
and 17 for the city of Yaoundé. The initial results obtained in
this work are acceptable and encouraging.
General Terms
NB-IoT smart metering.
Keywords
Narrowband, IoT, NB-IoT, MTC, LPWA, Smart Meter.
1. INTRODUCTION
The Internet of Things is a global infrastructure for the
information society enabling advanced services by
interconnecting objects (physical and virtual) based on
existing and evolving interoperable information and
communication technologies [1]. It is expected that IoT will
manage the gigantic network of billions of devices to provide
many intelligent services to users. From the transmission rate
point of view, IoT communication services can be classified
into two categories: high-speed services (such as video
service) and low-speed services (such as counters) [2].
According to statistics from ATECH in 2017, narrowband
services account for more than 67% of total IoT services,
indicating that LPWAN technologies are really desirable
especially NB-IoT.
Narrowband Internet of Things is a Low Power Wide Area
technology proposed by 3GPP for the perception and
acquisition of data for intelligent applications at low
bandwidth. Smart metering is an application that potentially
uses NB-IoT technology for water and energy management,
which is expected to be implemented massively in the near
future.
The current system for recording the consumption indexes of
water and electricity meters in Cameroon presents a serious
problem, namely that of bills that do not correspond to actual
consumption. The planning analysis was performed by
examining NB-IoT technology as a connectivity technology
used to support smart meter service. The cities of Yaoundé
and Douala are dense urban areas. It is necessary to analyze
the minimum number of eNodeBs required from coverage and
capacity sizing to implement the NB-IoT network of smart
meter services for the two regions.
This document consists of several sections explaining the NB-
IoT network planning analysis for smart meter services.
Section 1 is an introduction and explanation of the planning
context. Section 2 describes the NB-IoT technology. Section 3
describes the methodology used for coverage and capacity,
and Section 4 presents the results. Finally, the last section,
Section 5, presents the conclusions of the study.
2. NB-IoT TECHNOLOGY
2.1 Definition and requirements
Since 2005, 3GPP has started extensive research on cellular
networks (such as GSM, UMTS and LTE) for machine type
communication services. NB-IoT is designed to connect a
large number of devices in a wide range of application areas
forming what is called the Internet of Things. Connected
objects must communicate through an existing cellular

21
infrastructure. 3GPP has also introduced different data rates
suitable for narrowband IoT, ranging from a few tens of Kbps
in 180 KHz bandwidth (LTE Cat-NB1) to a few hundred
Kbps [3, 4]. The NB-IoT is a low-power wide area network
solution that operates in licensed frequency bands. 3GPP
includes this technology in the LTE standards in order to
benefit from the large ecosystem offered by LTE technology
and mobile operators.
NB-IoT devices are designed with the following
requirements and objectives:
 Massive low-bandwidth device count: Support for at
least 52,547 connected devices in a cell site area.
This target is based on the use of 40 devices per
household, with household density based on the
City of London assumption provided in [5, 6]
(density of 1517 households per km2 and distance
between cell sites of 1732 m);
 Low power consumption: allow IoT devices to draw
a low current (on the order of a nanoampere) to
make it possible to charge a single battery for
several years (on the order of 10 years);
 Longer battery life: the goal is to offer a 10-year
battery life with a capacity of 5 Wh;
 Improvement of indoor and outdoor coverage: the
objective is to obtain an extended coverage of 20 dB
compared to cellular networks and other
technologies. A data rate of at least 160 bps should
be supported for both uplink and downlink ;
 Low complexity: the goal is to provide ultra-low
complexity devices to support IoT applications,
resulting in lower cost;
 Low latency: latency of 10s or less is the target for
99% of devices;
 Low cost: a cost of 5 USD per device. NB-IoT
devices are connected to the infrastructure and the
cellular network.
Figure 1: Advantages of NB-IoT.[15]
2.2 NB-IoT network architecture
Figure 2 shows the architecture of NB-IoT network,
which consists of five parts, namely NB-IoT device (things),
NB-IoT radio access, Evolve Packet Core (EPC) network,
NB-platform IoT and application server.
Figure 2: NB-IoT network architecture [7].
2.3 NB-IoT frequency bands
3GPP has defined a set of frequency bands for which NB-IoT
can be used. Here is an overview of the supported frequency
bands in the different regions [8]:
 North America : B4 (1700), B12 (700), B66
(1700), B71 (600), B26 (850) ;
 Asie-Pacifique : B1 (2100), B3 (1800), B5 (850),
B8 (900), B18 (850), B20 (800), B26 (850) et B28
(700) ;
 Europe : B3 (1800), B8 (900) et B20 (800) ;
 Latine America : B2 (1900), B3 (1800), B5 (850)
et B28 (700) ;
 Commonwealth of Independent States : B3
(1800), B8 (900) et B20 (800) ;
 Sub-Saharan Africa: B3 (1800) et B8 (900) ;
 Middle East and North Africa : B8 (900) et B20
(800) ;
These bands are only a subset of the bands supported by
3GPP. Version 13 is likely to be used: a total of thirteen
frequency bands (1, 2, 3, 4, 5, 8, 12, 18, 20, 26, 28, 66 and
71). After looking at the minimum band overlap for different
countries, it is estimated that a minimum of ten bands are
needed: 1, 2, 3, 4, 5, 8, 12, 20, 26 and 28 are needed for
coverage in every country.
2.4 Deployment mode
According to 3GPP specifications, there are three different
deployment scenarios: Standalone, the 200 KHz bandwidth is
used independently for NB-IoT outside of LTE; Guard Band,
a bandwidth of 200 KHz is allocated in the guard part of LTE;
In-Band, the 180 KHz bandwidth is occupied by LTE.
Figure 3: NB-IoT network deployment mode

22
3. METHODOLOGY
3.1 Planning
3.1.1 Definition
Planning is a process that aims to predict the
evolution of demand in terms of number of subscribers, the
volume of traffic induced by category of subscribers, the
equipment necessary for its routing under defined security
conditions and quality of service. This is an essential
operation for the deployment of a network.
3.1.2 Planning process
IoT networks and services are very different from
"classic cellular networks" in many aspects and especially
from a planning point of view. The following diagram
presents the planning procedure for a NB-IoT network.
Figure 4: planning process of a NB-IoT network.
3.1.3 IoT Specificities and impacts on planning
The specificities of IoT technologies have impacts on network
planning and design. IoT networks have different
characteristics from those of traditional cellular networks.
These characteristics have an impact on planning and design:
 Low power and large range
Gateways and terminals have high sensitivity. The
use of low frequencies implies strong signal penetration. The
use of narrowband allows a much greater reception range.
 Low deployment and running costs
The cost of gateways is low. 3GPP standards use existing
mobile networks. Therefore, the cost is low. A wide range
provides a large coverage area plus strong penetration (inside
buildings). We have a small number of sites or gateways. The
battery has a long life (100mA Rx current 100nA standby
current). Because most of the time, the device is in standby
mode and in connected mode only for transmission.
 Shared spectrum, interference management
This feature makes it possible to clearly evaluate the channels.
Frequency hopping and OFDMA/CDMA access and NOMA
technology help eliminate interference.
 Diversity of services
We have the diversity of traffic patterns and the diversity of
transmission modes.
 Low speeds, simple topology
Sharing the infrastructure of existing cellular
networks simplifies coverage of large areas. Having a low
throughput (hundreds to thousands of bits/sec) implies a
reduced number of gateways.
In the rest of our work, we will focus our attention on
the third and fourth phase of the planning process, namely
dimensioning as presented in figure 4.
3.2 Network dimensioning
3.2.1 Principe
In general, the dimensioning of mobile radio networks is
carried out according to the steps below:
 Traffic/data analysis;
 Estimation of coverage;
 Capacity assessment.
Cell dimensioning requires fundamental data and parameters.
These parameters include subscribers population, traffic
distribution, geographic area to be covered, frequency band,
allocated bandwidth, and coverage and capacity requirements.
The system-specific parameters such as the transmission
power of the antennas, their gains, the estimation of the
system losses, the type of antenna system used, must also be
known, before the start of the dimensioning, because each
network without thread has its own set of parameters.
Propagation models should be selected and modified if
necessary, depending on the area and frequency band. All
These are needed to estimate coverage.
3.3 Coverage dimensioning
Coverage of an area is ensured by a base station (BS).
Therefore, it is essential to determine the number of BS
needed to fully coverage requirements of an area, in order to
avoid access failure, communication failure and reduce the
handover rate. To do this, a link budget should be done.
Coverage planning procedure is presented by the following
figure.
Figure 5: Coverage planning procedure [9].
3.3.1 Link budget
The link budget allows to calculate the maximum
allowable path loss of a radio waves when it moves from one
point to another. Through the maximum allowable path loss
combine with an appropriate propagation model, the coverage
radius can be deducted. The input parameters required for the
link budget are: transmission power for the transmitter and
receiver, transmitter and receiver gains, cable losses,
interference margins and receiver sensitivity of both ends. In a
mobile radio network, the radio link is two-way: Uplink and
Downlink. This justifies the need to establish a link budget for
both directions. However, whatever the direction of the link,
the expression of the link budget remains the same, the main

23
difference residing at the level of the transmitter and the
receiver, and consequently of the characteristics of each. A
radio link can indeed be illustrated by the figure below.
Figure 6: Link budget illustration
The following equation gives the mathematical formula of a
link budget. [10] (from standard TS 38.901 of 3GPP).
 For the Downlink :
𝑷 𝑹𝑿 𝑫𝑳,𝒅𝑩𝒎
= 𝑷 𝑺𝑪 𝒅𝑩𝒎 + 𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝒆𝑵𝑩 – 𝑪𝒂𝒃𝒍𝒆 𝑳𝒐𝒔𝒔𝒆𝑵𝑩
+ 𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝒈𝒂𝒊𝒏𝑼𝑬– 𝑪𝒂𝒃𝒍𝒆 𝑳𝒐𝒔𝒔𝑼𝑬
− 𝑷𝒂𝒕𝒉 𝑳𝒐𝒔𝒔 𝒑𝒓𝒐𝒑𝒂𝒈𝒂𝒕𝒊𝒐𝒏 𝒎𝒐𝒅𝒆𝒍 − 𝑷𝒆𝒏𝒆𝒕𝒓𝒂𝒕𝒊𝒐𝒏 𝑳𝒐𝒔𝒔
− 𝑰𝒏𝒕𝒆𝒓𝒇𝒆𝒓𝒆𝒏𝒄𝒆 𝑴𝒂𝒓𝒈𝒊𝒏 − 𝑺𝒍𝒐𝒘 𝑭𝒂𝒅𝒊𝒏𝒈
Where:
𝑷 𝑺𝑪 𝒅𝑩𝒎 = 𝑷 𝑻𝑿 𝒅𝑩𝒎 − 𝟏𝟎. 𝒍𝒐𝒈𝟏𝟎 𝑵𝑺𝒖𝒃 𝒄𝒂𝒓𝒓𝒊𝒆𝒓𝒔 𝐸𝑞. 1
 For the Uplink :
𝑷 𝑹𝑿 𝑼𝑳,𝒅𝑩𝒎 =
𝑷 𝑼𝑬 𝒅𝑩𝒎 + 𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝒆𝑵𝑩 – 𝑪𝒂𝒃𝒍𝒆 𝑳𝒐𝒔𝒔𝒆𝑵𝑩 +
𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝒈𝒂𝒊𝒏𝑼𝑬 – 𝑪𝒂𝒃𝒍𝒆 𝑳𝒐𝒔𝒔𝑼𝑬 −
𝑷𝒂𝒕𝒉 𝑳𝒐𝒔𝒔 𝒑𝒓𝒐𝒑𝒂𝒈𝒂𝒕𝒊𝒐𝒏 𝒎𝒐𝒅𝒆𝒍 − 𝑷𝒆𝒏𝒆𝒕𝒓𝒂𝒕𝒊𝒐𝒏 𝑳𝒐𝒔𝒔 −
𝑰𝒏𝒕𝒆𝒓𝒇𝒆𝒓𝒆𝒏𝒄𝒆 𝑴𝒂𝒓𝒈𝒊𝒏 − 𝑺𝒍𝒐𝒘 𝑭𝒂𝒅𝒊𝒏𝒈 Eq.2
Let assume that: 𝑻𝑷𝑳 = 𝑷𝒂𝒕𝒉 𝑳𝒐𝒔𝒔 𝑷𝒓𝒐𝒑𝒂𝒈𝒂𝒕𝒊𝒐𝒏 𝑴𝒐𝒅𝒆𝒍 +
𝑷𝒆𝒏𝒆𝒕𝒓𝒂𝒕𝒊𝒐𝒏 𝑳𝒐𝒔𝒔 + 𝑰𝒏𝒕𝒆𝒓𝒇𝒆𝒓𝒆𝒏𝒄𝒆 𝒎𝒂𝒓𝒈𝒊𝒏 +
𝑺𝒍𝒐𝒘𝑭𝒂𝒅𝒊𝒏𝒈 + 𝑪𝒂𝒃𝒍𝒆 𝑳𝒐𝒔𝒔𝒆𝑵𝑩 + 𝑪𝒂𝒃𝒍𝒆 𝑳𝒐𝒔𝒔𝑼𝑬 Eq.3
The link budget equation becomes:
𝑷 𝑹𝑿 𝑫𝑳,𝒅𝑩𝒎 = 𝑷 𝑺𝑪 ,𝒅𝑩𝒎 + 𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝒆𝑵𝑩 +
𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝑼𝑬 − 𝑻𝑷𝑳 Eq.4
𝑷 𝑹𝑿 𝑼𝑳,𝒅𝑩𝒎 = 𝑷 𝑼𝑬 ,𝒅𝑩𝒎 + 𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝑼𝑬 +
𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝒆𝑵𝑩 − 𝑻𝑷𝑳 Eq.5
Where:
PRX: Received signal;
PTX: BS transmitting power;
PSC: sub carrier power;
Subcarrier Quantity: This parameter designates the number
of sub-carriers used. It depends on the number of RBs used
and the configuration of the subcarrier. It can be calculated by
the formula:
𝑵𝒖𝒎𝒃𝒆𝒓𝒔𝒖𝒃−𝒄𝒂𝒓𝒓𝒊𝒆𝒓𝒔 =
𝑵𝒖𝒎𝒃𝒆𝒓𝑹𝑩 𝒖𝒔𝒆𝒅 × 𝑵𝒖𝒎𝒃𝒆𝒓𝒔𝒖𝒃−𝒄𝒂𝒓𝒓𝒊𝒆𝒓𝒔 𝒑𝒆𝒓 𝑹𝑩 Eq.6
TPL: represents the total loss on the path (BTS to UE);
Antenna Gain: represents the antenna gain;
Path Loss: is the maximal allowable path loss;
Penetration Loss: represents the penetration losses;
In fact, the penetration losses are model by the following
formula according to authors in [18]:
𝑷𝑳𝑷𝒆𝒏𝒆𝒕𝒓𝒂𝒕𝒊𝒐𝒏,𝑶𝟐𝑰 = 𝑷𝑳𝒃 + 𝑷𝑳𝒕𝒘 + 𝑷𝑳𝒊𝒏 + 𝑵 𝟎, 𝝈𝑷
𝟐
Eq. 7
Where :
PLb : is the path loss ;
PLtw : loss due to the penetration inside a building from an
external wall;
PLin : is the loss inside the building depending of the building
size ;
σ: is the standard deviation of the penetration loss.
 Receiver sensitivity
Receiver sensitivity is a measure of the minimum
signal strength that a receiver can detect. It tells us the
weakest signal that a receiver will be able to identify and
process. In other words, it is the minimum power below which
the quality of the link is degraded. It corresponds to a
propagation with maximum losses. We should always have
𝑷 𝑹𝑿 𝑫𝑳,𝒅𝑩𝒎 ≥ 𝑺 𝑹𝑿 𝑫𝑳,𝒅𝑩𝒎. At the reception threshold, the
received power is equal to the sensitivity and the Path loss
become the Maximum allowable path loss (MAPL)
Therefore, the expression for the radio link budget becomes:
𝑺 𝑹𝑿 𝑫𝑳,𝒅𝑩𝒎 = 𝑷 𝑺𝑪 ,𝒅𝑩𝒎 + 𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝒆𝑵𝑩 +
𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝑼𝑬 − 𝑻𝑷𝑳𝑫𝑳 Eq.8
𝑺 𝑹𝑿 𝑼𝑳,𝒅𝑩𝒎 = 𝑷 𝑼𝑬 ,𝒅𝑩𝒎 + 𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝑼𝑬 +
𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝒆𝑵𝑩 − 𝑻𝑷𝑳𝑼𝑳 Eq.9
And the MAPL is: Where MAPL is the Maximum Allowable
Path loss of the link.
Then we have:
𝑴𝑨𝑷𝑳 =
𝑷 𝑺𝑪 𝒅𝑩𝒎 + 𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝑮𝒂𝒊𝒏𝒆𝑵𝑩 – 𝑪𝒂𝒃𝒍𝒆 𝑳𝒐𝒔𝒔𝒆𝑵𝑩 +
𝑨𝒏𝒕𝒆𝒏𝒏𝒂 𝒈𝒂𝒊𝒏𝑼𝑬– 𝑪𝒂𝒃𝒍𝒆 𝑳𝒐𝒔𝒔𝑼𝑬 −
𝑷𝒆𝒏𝒆𝒕𝒓𝒂𝒕𝒊𝒐𝒏 𝑳𝒐𝒔𝒔 − 𝑰𝒏𝒕𝒆𝒓𝒇𝒆𝒓𝒆𝒏𝒄𝒆 𝑴𝒂𝒓𝒈𝒊𝒏 −
𝑺𝒍𝒐𝒘 𝑭𝒂𝒅𝒊𝒏𝒈 − 𝑺 𝑹𝑿 𝑫𝑳,𝒅𝑩𝒎 Eq.10
Indeed, the sensitivity depends on a set of parameters
including the thermal noise power at the receiver, the noise
factor and the SNIR. Depending on these parameters, the
expression of the receiver sensitivity is given below
𝑺𝑹𝑿 = 𝑭𝒊𝒈𝒖𝒓𝒆 𝑵𝒐𝒊𝒔𝒆 + 𝑻𝒆𝒓𝒎𝒂𝒍 𝒏𝒐𝒊𝒔𝒆 + 𝑺𝑰𝑵𝑹 Eq. 11
The noise figure depends on the frequency band, the duplex
gap and the allocated bandwidth. It is a parameter specific to
each manufacturer. The SINR, just like the noise factor, is a
parameter specific to each manufacturer.
3.3.2 Cell coverage radius calculation
From the link budget, we can have the MAPL, this
MAPL at the edge of the reception is equal to the propagation
model evaluate through a propagation model equation.
𝑴𝑨𝑷𝑳 = 𝑷𝒂𝒕𝒉 𝑳𝒐𝒔𝒔 𝒅 𝑷𝒓𝒐𝒑𝒂𝒈𝒂𝒕𝒊𝒐𝒏 𝑴𝒐𝒅𝒆𝒍 𝐸𝑞. 12
The path loss itself is a function of the distance between the
BTS and the user equipment, the path los scan also be written
on the form
𝑷𝒂𝒕𝒉 𝑳𝒐𝒔𝒔 𝒅 𝑷𝒓𝒐𝒑𝒂𝒈𝒂𝒕𝒊𝒐𝒏 𝑴𝒐𝒅𝒆𝒍 = 𝑨 + 𝑩𝒍𝒐𝒈 𝒅 𝐸𝑞. 13
3.3.3 Propagation model
The propagation model is used to estimate the loss
during propagation of the radio wave caused by terrain and
man-made environments. The propagation model is the
foundation of coverage planning. A good model means more
accurate planning. The propagation pattern depends on the
allocated frequency of the system. Different propagation
patterns have different allocated frequency ranges. It is
important to note that the propagation patterns depend on the
type of area. There are four types of areas commonly used in
network planning: dense urban, urban, suburban and rural.
Moreover, an indoor propagation model differs from an

24
outdoor propagation model. The different existing
propagation models are: the COST231-Hata model, the
Okumura-Hata model, the K-factor or SPM model, etc.
A propagation model can be optimized to be more precise for
specific town or environment, authors have proposed many
methods for propagation model optimization in [16] [17] [18]
[19].
In this work, we will use the basic COST231-Hata
propagation model, because the NB-IoT desired network to be
plan should operate at 1800MHz spectrum where in
Cameroon all existing operators have deploy 4G LTE
network.
This model is applicable for macro cell predictions when the
frequency range is between 1500 to 2000 MHz. The formula
for this model is given by:
𝑳 = 𝟒𝟔. 𝟑 + 𝟑𝟑. 𝟗 𝒍𝒐𝒈 𝒇 − 𝟏𝟑. 𝟖𝟐 𝒍𝒐𝒈 𝒉𝒃 + 𝟒𝟒. 𝟗 −
𝟔. 𝟓𝟓𝒍𝒐𝒈𝒉
𝒃
𝒍𝒐𝒈 𝒅 − 𝑨 𝒉𝒎 + 𝑪 Eq. 14
C=3dB for big cities and C=3dB for small or medium cities
or suburban areas and
𝑨 𝒉𝒎 = 𝟏. 𝟏𝒍𝒐𝒈𝒇 − 𝟎. 𝟕 𝒉𝒎 − 𝟏. 𝟓𝟔𝒍𝒐𝒈𝒇 − 𝟎. 𝟖
Eq.15
3.3.4 Cell surface calculation
Using the above equations, we can have:
𝑴𝑨𝑷𝑳 = 𝑨 + 𝑩𝒍𝒐𝒈 𝑹 𝐸𝑞. 16
That is:
𝑹 = 𝟏𝟎
𝑴𝑨𝑷𝑳−𝑨
𝑩 , 𝑹 𝒊𝒏 𝑲𝒎 𝐸𝑞. 17
Following the coverage planning process, we must calculate
the radii in UL and DL, and retain only the smallest value. It
is this which corresponds to the optimum radius of the cell.
Already knowing the radius of coverage, we can determine
the surface of the target area for trisector cell as follows:
𝑨𝒓𝒆𝒂𝒆𝑵𝑩 = 𝟗
𝟑
𝟖
∗ 𝑹𝟐
≈ 𝟏. 𝟗𝟓 ∗ 𝑹𝟐
Eq. 18
Thus, knowing the total surface of the zone to be covered, it is
possible to determine the number of BS, noted 𝑵𝒆𝑵𝑩,
necessary to ensure this coverage.
𝑵𝒆𝑵𝑩 =
𝐓𝐨𝐭𝐚𝐥 𝐒𝐮𝐫𝐟𝐚𝐜𝐞
𝑨𝒓𝒆𝒂𝒆𝑵𝑩
Eq. 19
3.3.5 Practical case of coverage dimensioning
The dimensioning will be done for 02 main towns
of Cameroon, namely Yaoundé and Douala. Administratively,
Cameroon now has ten regions, themselves divided into 58
departments. Yaoundé, is a city with seven hills, it is the
political capital of the country is populated by 4,100,000
inhabitants in 2019, the city of Yaoundé is located between 3°
52' north latitude and 11°31' east longitude. This city covers
an area of 183 km2. Douala is a port city in Central Africa on
the Wouri River estuary open to the Gulf of Guinea. The city
of Douala is located between 4° 03' north latitude and 9° 42'
east longitude. Its agglomeration has 3.7 million inhabitants
(2019 estimate), it is the economic capital of Cameroon, the
main business center of the country. It covers an area of 210
km2. Yaoundé and Douala are the two largest cities in the
country. According to the National Institute of Statistics, the
estimated number of households is 3,255,651 in the city of
Yaoundé and 3,322,170 in the city of Douala, with a density
of households per km2 of 17,790 and 15,820 respectively. .
Table 2 shows that planning the coverage of an NB-IoT
network in the cities of Yaoundé and Douala leads to
obtaining a number of eNodeBs in each city.
Table 1 : Coverage dimensioning result.
Parameters DL
(NPDSCH)
UL
(NPUSCH)
Bandwidth LTE (MHz) 20
Bandwidth NB-IoT (kHz) 180
Frequency of carrier (MHz) 1800
Number of resources Block 1RB
Number of subcarriers 12 SC
Bandwidth subcarrier (Hz) 15000
Area of the Yaoundé city
(Km2
)
183
Area of the Douala city
(Km2
)
210
Transmitter Tx = eNB Tx = UE
(a) Total power (dBm) 46 23
(b) Power for NB-IoT(dBm) 35 23
(c) Antenna gain Tx(dBi) 18 0
(d) Cable loss Tx(dB) -0.5 0
(e) PIRE 52.5 40.5
Receiver Rx-UE Rx-eNB
(f) Receiver noise factor (dB) 5 3
(g) Thermal noise (dBm/Hz)
= -174+10log (15000)
-132.24
(h) SINR (dB) -12 -12
(i) Receiver sensibility (dB)
= (f)+ (g)+ (h)
-139.24 -141.24
(j) Antenna gain Rx(dBi) 0 18
(k) Cable loss Rx(dB) 0 -0.5
(l) Interference margin(dB) 0
(m) Penetration loss (dB) -20
(n) Shadow fading(dB) -9.48
MAPL (dB) = (e) - (i)+ (m)+
(n)
162.26 /152.26
Propagation model COST231-Hata
Antenna heigh eNB(m) 30
Antenna heigh UE (m) 1.5
Ahu 0.04
MAPL = A + B log R Avec A = 139.20 et B = 35.22
Cell radius (Km) = 10
MAPL −A
B
4.52 2.35
Area of a tri-sector cell
(Km2
) = 1.95*R2
39.84 10.77
Number of sites in Yaoundé 17
Number of sites in Douala 19
From this planning we can see that respectively for Yaoundé
and Douala, 17 and 19 NB-IoT sites could be enough to
provide a smart metering solution which can include
electricity, water and gas. In the next section, we will discuss
about capacity planning in NB-IoT applied to the case of
Yaoundé.
3.4 Capacity planning
For NB-IoT network capacity planning, we have two
procedures.
3.4.1 First Procedure
During the initial feasibility study of NB-IoT in
3GPP version 13, the goal was to design a system capable of
supporting a load of 60,680 devices per km2 [12]. Later, in
version 14, it was decided that NB-IoT should also meet the
5G requirement for connection density of 1,000,000 devices
per km2 [11]. NB-IoT devices are expected in large numbers
in homes, cars, cities and municipalities [13], [14]. NB-IoT
Network Planning Capacity Analysis for Smart Metering

25
Services refers to 3GPP Rel.13 TR 45 820. To assess NB-IoT
system capacity, two cities, London and Tokyo, are used as
models to determine population, household density, and
number of NB-IoT devices in use. This target is based on the
use of 40 devices per household. Table 2 shows the number of
devices per cell site area resulting from multiplying the area
of the cell site area by the density of households in a square
area, and the number of devices per household so that the
capacity of these services can at least support up to 52,547
devices in the cell site area.
Table 2: CIoT capacity for London and Tokyo urban
areas
City Household
Density
per km2
ISD
(m)
Area
of
Cell
Site
Sector
(km2
)
Number
of devices
per
household
Number
of
devices
within a
Cell
Site
Sector
London 1517 1732 0.866 40 52548
Tokyo 2316 1732 0.866 40 80226
The number of devices per site sector is given by the
following formula:
𝑵𝒂𝒑/𝒔𝒆𝒄 = 𝑺𝟎 ∗ 𝛒 ∗ 𝛔 Eq. 20
Where 𝑺𝟎 is the cellular site sector, 𝛒 is the density of
household per km2 and 𝛔 is the number of devices per
household.
In urban area, all sites have 3 sectors, then the number of
devices in a sites is equal to the number of site sector times 3.
𝑵𝒂𝒑/𝒔𝒊𝒕𝒆 = 𝟑 ∗ 𝑵𝒂𝒑/𝒔𝒆𝒄 Eq. 21
Le nombre total d‟appareils dans la zone cible est donné par :
𝑵𝒂𝒑 = 𝑵𝒉𝒉 ∗ 𝑵𝒅𝒆𝒗 𝒑𝒆𝒓 𝒉𝒉 Eq.22
𝑵𝒉𝒉 : Number of households
𝑵𝒅𝒆𝒗 𝒑𝒆𝒓 𝒉𝒉 : number of devices per household
NeNB: number of eNodeB is calculated as follows:
𝑵𝒆𝑵𝑩 =
𝐍𝐚𝐩
𝑵𝒂𝒑/𝒔𝒊𝒕𝒆
Eq. 23
Based on this model, the total number of devices projected for
electricity, water and gas smart metering services in the city
of Douala is 9966510 and in the city of Yaoundé 9766953.
Table 3: Capacity planning for the city Douala.
City Household
Density
per km2
Area
of
Cell
Site
Sector
(km2
)
Number
of devices
per
household
Number
of
devices
within a
Cell Site
Sector
Douala 15820 3.59 3 170381
Total number of
devices
Number of eNodeB
9966510 19
Table 4: Capacity planning for the city of Yaoundé.
City Densité
des
ménages
par km2
Area
of Cell
Site
Sector
(km2
)
Number of
devices
per
household
Number of
devices
within a
Cell Site
Sector
Yaoundé 17790 3.59 3 191598
Total number of
devices
Number of eNodeB
9766953 17
It follows from the calculation that the city of Douala requires
19 sites and that of Yaoundé 17 sites to carry Nb-IoT traffic
(see Tables 3 and 4).
3.4.2 Second procedure
In this part, we are modeling services and connected devices.
Modeling end devices, sensors, and other connected objects
yields the IoT service model. This service model is presented
as follows:
 Fleet management: the end device can send a packet
in the network every 30 seconds to track a vehicle;
 Logistics: a terminal device can send a packet in the
network every 5 minutes to signal its busy state;
 Water meter: can send a packet once a day to inform
about water consumption;
 Electricity meter: can send a packet once a day to
inform about electricity consumption.
In this second procedure, parameters such as the frequency of
transmission of packets at peak time, the number of devices
connected for each type of service, the number of packets per
day for a device, the break margin, the safety margin are taken
into account to scale the capacity. It consists of determining
the number of packets for each service and the total number of
packets per day for all services. Traffic characteristics and
technical requirements are the basis used to determine the
capacity needed for planning IoT networks for smart meter
services. To anticipate the peak hour load, a margin is
required as a solution. The solution consists of two margins, a
burstiness margin and a safety margin. Burst margin provides
the highest percentage of network overload assumptions to
minimize traffic spikes. The function of the headroom is to
manage burst traffic on a small time scale. As for the traffic
calculation for NB-IoT services, the percentage value of the
Burstiness margin and the safety margin assumptions were
20% and 10%. The total number of data packets needed in one
day for the cities of Yaoundé and Douala can be seen in
Tables 5 and 6 below.
This model describes the different services provided by the
network by identifying the characteristic parameters of the
different types of services. The capacity of a site is 1800000
packets. The following formulas are used.
𝑵𝒑 𝒊 = 𝟐𝟒 ∗ 𝑭 𝒊 ∗ 𝑵 ∗ 𝒏 𝟏 + 𝑴𝒃 𝟏 + 𝑴𝒔 Eq. 24
𝑵𝒑(𝒊) = 𝟑𝟏. 𝟔𝟖 ∗ 𝑭(𝒊) ∗ 𝑵 ∗ 𝒏 Eq.25
Pour Ms = 10% et Mb = 20%
Np: number of packets per day for service i
F: packet transmission frequency at busy hour for service i
N: number of end devices
n: number of packets per day for a device
Mb: packet break margin
Ms: safety margin
Tpacd: Total number of packets per day
The Total number of data packets needed in a day is given by:
𝑻𝒑𝒂𝒄𝒅 = 𝑵𝒑
𝒌
𝒊=𝟏 𝒊 Eq. 26

26
The number of eNodeB can be calculated as follow:
𝒏𝒖𝒎𝒃𝒆𝒓 𝒐𝒇 𝒆𝑵𝒐𝒅𝒆𝑩 =
𝑻𝒑𝒂𝒄𝒅
𝒔𝒊𝒏𝒈𝒍𝒆 𝒔𝒊𝒕𝒆 𝒄𝒂𝒑𝒂𝒄𝒊𝒕𝒚
Eq. 27
Table 5: Capacity dimensioning of Yaoundé city, Source
[11].
NB-IoT
services
Packet
transmissio
n
frequency
at BH(F)
End
device
number(
N)
Number
of packets
per day
for one
device(n)
Burstine
ss
Margin(
Mb)
Security
Margin(
Ms)
Number of
packets(Np
)
Sensor 1 5000 24 20% 10% 3801600
Metering
water
0,04 600000 1 20% 10% 760320
Metering
electric
0,04 600000 1 20% 10% 760320
Public
lightning
2 10000 1 20% 10% 13200
Parking
manageme
nt
2 4000 48 20% 10% 12165120
Tracking
logistic
2 1000 48 20% 10% 3041280
Asset
Tracking
3 1000 72 20% 10% 6842880
Agriculture 1 1800 24 20% 10% 1368576
Wearables 0,5 6000 12 20% 10% 1140480
Home
automation
0,5 2500 12 20% 10% 475200
Total number of packets per day 30368976
𝟑𝟎𝟑𝟔𝟖𝟗𝟕𝟔
𝟏𝟖𝟎𝟎𝟎𝟎𝟎
= 𝟏𝟕 ,
17 eNodeB are needed to provide service in Yaoundé.
Table 6: Capacity dimensioning of Douala city, Source
[11].
NB-IoT
services
Packet
transmissio
n
frequency
at BH(F)
End
device
number(
N)
Number
of packets
per day
for one
device(n)
Burstine
ss
Margin(
Mb)
Security
Margin(
Ms)
Number of
packets(Np
)
Sensor 1 5000 24 20% 10% 3801600
Metering
water
0,04 500000 1 20% 10% 633600
Metering
electric
0,04 500000 1 20% 10% 633600
Public
lightning
2 10000 1 20% 10% 13200
Parking
manageme
nt
2 4000 48 20% 10% 12165120
Tracking
logistic
2 1000 48 20% 10% 3041280
Asset
Tracking
3 1100 72 20% 10% 7527168
Agriculture 1 1800 24 20% 10% 1368576
Wearables 0,5 6000 12 20% 10% 1140480
Home
automation
0,5 2500 12 20% 10% 475200
Total number of packets per day 30799824
𝟑𝟎𝟕𝟗𝟗𝟖𝟐𝟒
𝟏𝟖𝟎𝟎𝟎𝟎𝟎
= 𝟏𝟖 , 18 eNodeB are
needed to provide service in Douala.
3.5 Design of an android App for NB-IoT
planning
In this part we present the design of an android App
developed with the following functionality for smart metering
solution:
 Perform a link balance to determine the maximum
allowable losses on that link Calculate the coverage
radius of a cell;
 Calculate the number of base stations required to provide
full coverage of the area;
 Calculate the maximum number of packets per sector;
 Calculate the total number of packets in a cell;
 Determine the number of base stations required to
support the capacity of the target area;
 Present the curves and the amount of consumption per
utility (electricity, water or gas);
 Access to real-time data;
 Checking and control of monthly consumption.
Design is an important phase in the development cycle of a
project. The support of this phase is done by appropriate
techniques and tools to produce a high quality software and
applications. The design should take into account the needs,
experience and capabilities of the user. In our approach, we
will use the industrial standard of object modeling UML, to
have a good understanding of the functioning of the tool to be
produced. For this purpose, we will produce diagrams such as:
class diagrams, system sequence diagrams and use case
diagrams.
3.5.1 Use case diagram
The set of use cases describes the objectives of the
system. It constitutes the use case diagram whose graphical
representation is as follows:
Figure 7: Example of Use Case Diagram.
They are represented by the use cases, the system (with its
boundary), the actors (primary and secondary) and the
associations between actors and use cases. The actors and use
cases of a system are linked by an association relationship
which materializes the communication between an actor and a
use case and is represented by a continuous line (see
representation above).
To this end, we can, in view of the previously defined actors,
highlight the following use case diagrams for our application:
Figure 8: NB-IoT application use case diagram.

27
3.5.2 Class diagram
It is presented by the following figure.
Figure 9: Class diagram for proposed NB-IoT App.
3.5.3 Activity Diagram
Some activity diagrams are presented below for our
proposed system:
Figure 10: Sequence diagram of the case « Authentication »
Figure 11:Sequence diagram of the case « Enter »
Figure 12: Sequence diagram for the case « inscription »
This part was devoted to the modeling of the App to be
developed. All this work led to results which will be presented
in the next chapter, the result of coverage and capacity
planning were already include inside the related part because
of the pedagogic approach which obliged us the develop the
idea till the end by presenting the related results for the
specific case of the selected towns.
4. RESULT AND COMMENT
The methodology used to solve the problem posed
above has allowed us to obtain the results that will be
presented in this chapter. To do this, we will proceed to a
presentation of the different interfaces constituting the
developed software, accompanied by some related comments.
4.1 The authentication page
When you access the application, you are directed
to this screen. It contains the application registration and login
form. The user must fill in the fields to have access to the
modules allowing him to carry out the task he wants to
accomplish. After filling in these fields, to authenticate, he
clicks on the connect button. Figure 13 presents the
authentication page. After a successful login, we have the
interface presents in figure 14.
Figure 13: Authentication page
4.2 Moduls interface
Figure 14: Moduls interface.
Once registered on the platform, users have access to the main
screen. This interface presents the available modules, namely
coverage sizing, capacity sizing and a captive data access
portal.
4.3 Coverage planning interface
This interface allows to plan the coverage of the NB-IoT
network for a given city. It has a multi steps form where the
user enters the information and performs the calculation. This
process is broken down into different tasks, including: Link
budget, MAPL, coverage radius of a cell or eNodeB site,
number of eNodeBs needed to cover the chosen area. This
interface is a prompt to enter the various parameters of the
link budget. These parameters, which are those of the UE and
of the base station, are of two types: the transmission
parameters and the reception parameters. Therefore, the user
will have to fill in this information. The different fields to be
filled in concern respectively:
 The total powers of the base station and of the UE;
 The gains of the base station and UE antennas;
 Cable losses at BS and UE respectively;
 The SINR of the base station and of the UE;
 UE and base station noise figures;
 Receiver sensitivity in DL and Ul;
 The frequency, the heights of eNodeB and UE, etc.
The following figure presents an interface for coverage
planning.

28
Figure 15:Interface for coverage planning case.
4.4 Interface capacity planning
4.4.1 Based on first procedure
Figure 16:Interface for capacity planning based on first
procedure.
This interface is used to do the capacity planning of
the NB-IoT network for a given city. It has a multi-step form
where the user enters the information and performs the
calculation. This gives the result for the calculation of: the
number of devices in a cell site sector; number of devices in a
cell; total number of devices in the chosen area; total number
of eNodeBs needed to handle all traffic. The "CALCULATE"
button provides the number of devices in a sector, the number
of devices in the cell and the number of base stations.
5.4.2 Based on second procedure
Figure 17:Interface for capacity planning.
Through this interface, a capacity planning can be
done based on the second method. Once clicked on the
“CACULATE” button, we obtain the number of packages, the
total number of packages per day.
4.5 Customer service access interface
Figure 18: Customer service access interface.
This interface allows households to access their panel in order
to consult their consumption. This will allow them to properly
control their monthly consumption. The user must first
register by entering his name, his telephone number and his
code, PIN code. Then, to access the data, the user must enter
their phone number and the PIN code. After these actions, the
user can log in and view the consumption curves.
4.6 Customer data interface
On this interface, households can consult their
consumption on the services they have registered on the
platform. The customer can see the indexes of his electricity
and water meter. He can also visualize the consumption curve
to better take control of his consumption as shown in figure
33. This is the consumption curve according to the months. It
presents the months on the abscissa, therefore from January to

29
December, and on the ordinate, the indexes sent by the smart
meter in m3 or KWh.
Figure 19: Users data interface.
4.7 Services interface (Water or electricity)
Figure 20:Interface for adding services
On this interface, households can add new services.
This involves, for example, registering a new electricity meter
to be able to monitor consumption. To this end, the customer
enters the number of his NB-IoT smart meter and his personal
information such as name, telephone number, etc. Once
informed, he can consult his consumption curve.
The results of the calculations show that the number
of sites required is different both in terms of coverage and
capacity, and the highest number of sites has been taken to
meet the requirements. Thus, the number of sites required is
19 for the city of Douala and 17 for the city of Yaoundé. The
initial results obtained in this work are acceptable and
encouraging.
5. CONCLUSION
At the end of our work, which focused on NB-IoT:
Smart Metering in the cities of Yaoundé and Douala, it is
important to take stock of the different articulations
addressed. Firstly, our objective was to develop a planning
solution that could enable an operator using 1800MHz band to
deploy the future NB-IoT network, whose operation will be
subordinate to the existing 4G network infrastructure, in order
to create a new source of revenue in the digital era.
Subsequently, we presented the coverage and capacity
planning of the NB-IoT network. Referring to previous works,
including articles, theses and publications, on similar topics,
we adopted a methodology that allowed us to clearly identify
the expected solution. The number of sites required is 19 for
the city of Douala and 17 for the city of Yaoundé. We have
also propose and App which can embedded all this solution.
6. REFERENCES
[1] ITU, (2016), ITU-T Y.4000/Y.2060 (06/2012), ITU-T
Recommandations.
https://doi.org/http://handle.itu.int/11.1002/1000/11559.
[2] Y. Li and M. Chen, „„Software-defined network function
virtualization: A survey,‟‟ IEEE Access, vol. 3, pp.
2542–2553, 2015.
[3] 3GPP Std. 36.300, (2017), “Overall description”, Stage
2, v15.0.0.
[4] 3GPP Std. 36.401, (2017), “Architecture description”,
v15.0.0.
[5] 3GPP, Tech. Rep. 36.888, (June 2013), “Study on
provision of low-cost machine-type communications
(MTC) user equipment (UEs) based on LTE,” v12.0.0.
[6] 3GPP, Tech. Rep. 45.820, (2015), “Cellular system
support for ultra-low complexity and low throughput
Internet of Things (CIoT),” v13.1.0.
[7] Sakshi P., Rakesh K., & Sanjeev J., (2018), A Survey on
Energy Efficient Narrowband Internet of things (NB-
IoT): Architecture, Application and Challenges”, (7),
16739 – 16776, IEEE
https://ieeexplore.ieee.org/document/8536384
[8] Afzal J., (2017), „‟ NB-IoT Frequency Bands (As per
3GPP Rel. 13, 14 and 15)‟‟,
http://www.techplayon.com/nb-iot-frequency-bands-as-
per-3gpp-rel-13-14-and-15/
[9] LTE network planning Huawei Technologies, 48 pages.
[10] 3GPP, Tech. Rep. 45.820, (2016), “Cellular System
Support for Ultra-low Complexity and Low Throughput
Internet of Things”, v13.0.0
[11] Sami TABBANE, (2016), “IoT Network Planning”,
Bangkok, Thailand 1-208
[12] Lauridsen, M.; Kovacs, I.Z.; Mogensen, P., Sorensen,
M., Holst S., “Coverage and Capacity Analysis of LTE-
M and NB-IoT in a Rural Area”, In Proceedings of the
2016 IEEE 84th Vehicular Technology Conference
(VTC-Fall), Montreal, QC, Canada, 18–21 September
2016.
[13] 3GPP, Tech. Rep. 45.820, (2016), “Cellular System
Support for Ultra-low Complexity and Low Throughput
Internet of Things”, v13.0.0.
[14] 3GPP, Tech. Rep. 22.861, “Feasibility study on new
services and markets technology enablers for massive
internet of things; stage 1,” Sep. 2016, v14.1.0.
[15] https://www.u-blox.com/en/blogs/innovation/iot-and-
four-reasons-why-licensed-spectrum-technologies-have-
been-worth-wait
[16] Deussom E. and Tonye E. «New Approach for
Determination of Propagation Model Adapted To an
Environment Based On Genetic Algorithms: Application
to the City Of Yaoundé, Cameroon», IOSR Journal of
Electrical and Electronics Engineering, Volume 10,
pages 48-49, 2015.
[17] Deussom E. and Tonye E. Optimization of Okumura
Hata Model in 800MHz based on Newton Second Order
algorithm. Case of Yaoundé, Cameroon, IOSR Journal of

30
Electrical and Electronics Engineering (IOSR-JEEE) 10
(2), 16-24
[18] Deussom Djomadji Eric Michel, Kabiena Ivan Basile,
Tonye Emmanuel, Propagation model optimization based
on Artificial Bee Colony algorithm: Application to
Yaoundé town, Cameroon. IOSR Journal of Electrical
and Electronics Engineering (IOSR-JEEE) (Mar – Apr
2020), PP 14-26.
[19] E DEUSSOM, E TONYE, New Propagation Model
Optimization Approach based on Particles Swarm
Optimization Algorithm- International Journal of
Computer Applications, 2015.
IJCATM : www.ijcaonline.org
View publication stats

Design of an NB-IoT Smart Metering solution Coverage and capacity planning.pdf

More Related Content

Similar to Design of an NB-IoT Smart Metering solution Coverage and capacity planning.pdf (20)

Recently uploaded (20)

Design of an NB-IoT Smart Metering solution Coverage and capacity planning.pdf