Professional Documents
Culture Documents
NB-IoT Optimization For Smart Meters Networks of S
NB-IoT Optimization For Smart Meters Networks of S
H O S T E D BY
Alexandria University
a
Electronics & Communications Dept., Ain Shams University, Cairo, Egypt
b
Nuclear Research Centre, Egyptian Atomic Energy Authority, Qalubia, Egypt
c
Faculty of Navigation Science and Space Technology, Beni-Suef University, Beni-Suef, Egypt
d
Radiation Technology Center, Egyptian Atomic Energy Authority, Cairo, Egypt
KEYWORDS Abstract The future demands for Smart Meters (SM) applications impose maximizing resources
5G; utilization. In this paper, we achieve two main objectives to comply with fifth-generation networks
NB-IoT; goals and serving the SM applications. First one is to enhance the Narrowband-Internet-of-Things
Smart meters; spectral efficiency. The second is to alleviate the signalling burden during each transmission request
Utilization; for SMs. First, we model the uplink scheduler for the NB-IoT access network using the state-
Spectral efficiency machine modelling methodology on the Simulink environment. The simulation results are verified
and validated with trusted modelling technique. Second, we optimize the NB-IoT uplink scheduler,
which exploits the periodicity nature of SMs applications. We achieve that by proposing an inte-
grated scheduling protocol which rearranges the transmission times of different smart meters util-
ities, and it draws a map for the transmission schedule of them to better utilize the sparse time
resources. The proposed approach comprises architecture, signalling, and algorithms. We evaluate
the optimization in term of numbers of SMs versus four performance indicators; resources utiliza-
tion, spectral efficiency, setup success percentage, and session drop rate according to a case study.
The simulation results prove the ability of our proposal to increase the spectrum utilization to
17.47%. This enhancement reflects on spectral efficiency that improved doubled.
Ó 2020 Faculty of Engineering, Alexandria University. Production and hosting by Elsevier B.V. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
* Corresponding author. Internet of Things (IoT) applications are one of the emerged
E-mail addresses: ebsosoft@yahoo.com, ahmed.abbas@eaea.org.eg technology applications that are considered a key digital trans-
(A.M. Abbas). formation enabler for several industries and utilities, including
Peer review under responsibility of Faculty of Engineering, Alexandria smart cities, factories, electricity grids, many other services.
University.
https://doi.org/10.1016/j.aej.2020.07.030
1110-0168 Ó 2020 Faculty of Engineering, Alexandria University. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
2 A.M. Abbas et al.
Recently, the IoT would be used as a component in the oper- to LTE. It is used in three modes: standalone, in-band, and
ational technologies of nuclear power plants and radiation guard-band [20]. The subcarrier bandwidth is 180 kHz in the
monitoring networks. Therefore, they need a reliable and effi- case of co-existence with current LTE networks to provide
cient network infrastructure for optimum operation. long periods of connections over mobile operator networks
Contrary to human traffic, the IoT communication profile because various objects have small amounts of data [20]. The
is completely different from what is experienced by smart- aim of NB-IoT is globally covering low-cost or complex ubiq-
phones over fourth-generation (4G) cellular networks. Where, uitous IoT devices worldwide [21].
a sufficient number of intelligent objects will become ubiqui- To cater for the massive IoT requirements, the 3GPP has
tous, enabling them to sense and manipulate information in standardized that the mMTC services will be served by further
real-time. Meanwhile, smartphones sessions start according evolving NB-IoT and LTE-M as integral parts of the 5G spec-
to the human profile. For periodic applications, a device trans- ifications, such that there will be no need for new standards in
mits the sensed data periodically to a network to deliver inno- the near future [22]. Although they are known as 4G technolo-
vative and real-time services [1]. For example, at smart cities, gies, the 3GPP has started a working group in the 5G core net-
the smart meter does not only register periodic reading but work to support LTE-M and NB-IoT radio access networks in
also it draws the profile for the object resources consumption. order to guarantee the compatibility of these technologies with
Where, this profile can be used for future predictions of sea- 5G [22].
sonal energy needs. So, smart meters play an important role
in decision making at different levels. Consequently, the SMs 1.2. Problem statement
became one of the most used and prevalence applications in
IoT domain [2–4].
Efficient and profitable IoT communication systems should be
The increased population density worldwide and the evolu-
low cost, size, weight, and power enabled. With the increased
tion of human and community lifestyles have led to the rapid
dependence on NB-IoT technology in people’s lives, a large
development of mobile communication technology. Thus, the
number of NB-IoT terminals need to be connected to the net-
3rd Generation Partnership Project (3GPP) has standardized
work in order to cater to the various needs of users. According
the vision for 5th generation mobile communication technol-
to predictions by Ericsson, the number of connected IoT
ogy (5G), which comprises the traditional LTE network along-
devices will reach 1.5 billion by 2022 [23].
side the future 5G network. The 5G and the 5G+ mobile
The rapid increase in the number of massive IoT devices
networks target to overcome the previous cellular standards
has added a growing challenge to 5G design owing to its con-
limitations and to become an enabling technology for future
trasting requirements [8] from two main perspectives. First, the
IoT applications. The 3GPP releases of 5G, starting from
narrow bandwidth may restrict the transmission performance
release 15 and beyond, introduced enhanced cellular communi-
of the IoT in addition to the communication overhead that
cations services such as enhanced mobile broadband (eMBB)
is required before data transmission [8]. Therefore, it is impor-
and massive machine-type communications (mMTC), and they
tant to achieve continuous improvements concerning the effi-
also introduced new services such as ultra-reliable low latency
cient use of these resources to support the continued growth
communication (URLLC) and time-sensitive communications
of IoT applications [24]. Second, access reservation protocols
(TSC) [5–7]. The URLLC and mMTC in 5G are closely related
designed for 5G networks face massive connectivity for IoT
to IoT [7]. Compared to the current 4G, it is expected that 5G
applications [3], because the device density is expected to be
mobile networks will efficiently support the basic IoT require-
larger than the capability of the techniques to realize access
ments, such as good coverage, high data throughput, low
reservation and RA procedures, and may significantly affect
latency, high scalability, high energy efficiency, and the ubiqui-
end-to-end delay. This issue would be critical, especially for
tous connectivity provision for IoT end-users [5]. 5G aims to
the scenarios of burst arrival [25]. The access requests and
maximize the capacity of the network to be 1000 times larger
the signalling associated with each periodic IoT transmission
than 4G [8].
process are a burden to both IoT device and network
The IoT 5G umbrella includes current technologies such as
resources. These problems are further transferred with uplink
unlicensed bands (short-range communication technologies
traffic and remain opened issues, as discussed in [20,26].
including Bluetooth, Zigbee, Wi-Fi, and the unused spectrum
Until now, 5G systems have been considered an evolution
band such as cognitive radio [9] and low-power wide-area net-
of the LTE-advanced (LTE-A) standard, where the 5G new
works (LPWANs). The latter are divided into non-3GPP net-
radio (NR) access reservation is also based on LTE-A random
works such as SigFox, LoRa, [3,4], and 3GPP networks such
access (RA) procedure [25,27]. Access class barring (ACB) and
as Long Term Evolution (LTE)-M in Release 12 and
its modifications is a standardized technique to process burst
narrowband-IoT (NB-IoT) in Release 13 [4,10–17]. Recently,
access in LTE and NR RA [25]. However, it may not be appro-
Fog radio access network (Fog-RAN) and Cloud radio access
priate for IoT applications [1]. Most adopted solutions focused
network (C-RAN) have been proposed for fifth-generation
on random-access phase optimization and other optimization
(5G) wireless communications that serve IoT applications
techniques, as we will present in Section 2. Currently, studies
[18]. For smart city applications, LPWANs are the most suit-
and developments for the NB-IoT have been widely adopted
able alternatives [3]. Therefore, we only focus on 5G cellular
by global mobile operators and manufacturing companies [24].
IoT in this paper.
NB-IoT achieves unrivalled LPWA performance in terms
1.3. Related works
of coverage and the variety of data rates and deployment sce-
narios compared to other LPWA technologies [19]. This tech-
nology is a modified design of LTE that aims to serve IoT The optimization works in MAC layer improve spectral and
traffic. Besides, it able to provide enhanced coverage compared energy efficiency to serve IoT traffic. This paper will adopt
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
NB-IoT optimization for smart meters networks of smart cities: Case study 3
spectral efficiency approach. The researches of this approach These features achieve by adopting different differential
adopted many different directions as follow. and integral equations types and techniques such as [38–
The authors of references [28] adopted the random access- 40], which were added to recent Simulink releases.
ing stage in MAC to improve the system throughput by using In the second step, we present an approach that employs the
Markov chain modelling technique. In [29], Begishev et al. persistent scheduling strategy based on the periodicity nat-
studied three BW allocation strategies; Static Allocation, ure of periodic IoT applications to develop integrated
Dynamic Allocation, and Dynamic Allocation with reserva- scheduling protocol. The proposal handles three modifica-
tion between LTE and NB-IoT. The study recommended the tion axes signalling, architecture, and processing algorithm.
‘‘dynamic resource sharing with reservation” strategy for The proposal modified the signalling by adding IoT type
heterogeneous IoT traffic. The evaluation process was exe- identification during Radio Resources Control (RRC) con-
cuted through formulating a model by using a two dimensional nection establishment messaging phase. At architecture
Markov chain. The first improvement trial for resources uti- axis, the medium access control (MAC) layer was chosen
lization was introduced in [30]. Where, Bing-Zhi et al. designed to execute the approached algorithm. The third axis repre-
an uplink (UL) scheduling algorithm that sort NB-IoT UEs sents processing algorithm.
based on minimum scheduling delay. Their algorithm was built
by using a Software Defined Network (SDN). The persistent Our algorithm is quite simple to alleviate the processing
scheduling strategy for periodic IoT applications was adopted burden for both the SM terminal and the access node. For get-
for the first time in [31]. Although it was addressed from an ting fair results, we divide the cell spectrum of NB-IoT equally
energy efficiency perspective, the authors focused only on on four data rates. We apply our proposal on one of the most
proposing prediction algorithm foresees the periodicity of famous periodic IoT applications that are smart meters. The
IoT traffic without reviewing the messaging procedures and optimization is applied to a case study for a designated geo-
their impact. Chen et al. [32] optimized the NB-IoT uplink graphical location has smart meters for two different utilities.
scheduling using the improved K-mean algorithm, and they The optimization is evaluated in term of the number of NB-
considered the IoT distribution model as an additional impact IoT UEs (in our case study, we call SM terminals or SM
on random accessing and its consequences on system through- UEs) versus four KPIs; resources utilization, spectral effi-
put. On the contrary, in [33], the periodic IoT traffic was ciency, RRC setup success percentage, and session drop rate.
addressed from a different perspective that focuses only on The simulation results prove the ability of our proposal to
alleviate the signalling burden. This perspective achieved by increase the spectrum utilization to 17.47%. This enhancement
eliminating some RRC procedures and only in limited cases. has a positive impact on spectral efficiency, which improved
The NB-IoT group communication technique was adopted doubled.
clearly in [34]. Their proposal was based on adapting Multime-
dia Broadcast Multicast Service (MBMS) protocol for serving 1.5. Paper organization
downlink traffic of IoTs. The cashing based group access was
adopted in [8] for Fog-RAN and in [35] for Industrial IoT cov- The remainder of the paper is organized as follow. Section 2
ered by small base stations. Jin in [36] introduced group based will describe our model design, our proposed optimization
accessing of same location and attributes under the group lea- and smart meters case study with its traffic profiles. The anal-
der. B. M. ElHalawany et al. in [37] maximized the cell ysis and the discussion for simulation results will be presented
throughput via adopting the Device to Device (D2D) commu- in Section 3. Finally, our conclusion will be presented in
nication methodology. However, their clustering idea was Section 4.
based on uniform deployment condition.
All the works, as mentioned above, did not consider the
periodicity nature of periodic IoT applications. Therefore, 2. NB-IoT uplink scheduler model
the issue of improving the system efficiency is still opened
due to the traffic management researches is still weak and This section describes our modelling methodology, our opti-
insufficient the study for the nature of IoT traffic types. This mization approach and applying the optimization on smart
paper presents an integrated vision to optimize this issue. meters case study at three parts as follow. The first one
describes the model design of the NB-IoT uplink scheduler.
1.4. Paper contribution The second subsection describes our optimization approach
for NB-IoT uplink scheduler. The case study is described in
This paper has two main objectives. The first one is to enhance the third subsection.
the NB-IoT spectral efficiency through increasing its spectrum
utilization. The second objective is to reduce the signalling bur- 2.1. Model description
den during each transmission request for periodic IoT applica-
tions. We achieved these objectives through the following two In general, the MAC layer charges of four main functions: ran-
steps: dom access control, uplink and downlink UEs data scheduling
control, Hybrid Automatic Repeat reQuest (HARQ) process,
First, we model NB-IoT uplink scheduler. The model is exe- and finally mapping between logical and transport channels
cuted using the state-machine modelling methodology over [41]. Fig. 1 explains the message sequence chart (MSC), start-
Simulink. We employ Simulink’s built-in features, which ing from random access request up to uplink data scheduling
provide continuous run at different simulating durations grant according to 3GPP release 14 [42,43]. In the figure, the
as desired for mathematical and state-machine models. messages sequencing are from one to seven.
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
4 A.M. Abbas et al.
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
NB-IoT optimization for smart meters networks of smart cities: Case study 5
Fig. 5 Geographical area has SMs, and each SM of utility has GID.
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
6 A.M. Abbas et al.
2.2.3. Algorithm
In our proposed procedures, the first task is initialization.
Where, we create a Clusters Time Table, which formulated
as a matrix as depicted in Eq. (1). This matrix represents a
database for all SM clusters that will be registered on the
access network. Consequently, these SM terminals will avoid
the random access and the RRC connection establishment pro-
cedures during the next periodic data transmissions. Then, the
first condition to run the rest of our proposed procedures is to
Fig. 7 The overall proposed algorithms flow chart. determine the IoT traffic type. As previously mentioned, our
proposed approach focuses on periodic IoT traffic. As depicted
in Fig. 7, the overall proposed procedures consist of four
used and widespread applications at IoT industry like smart phases, each of which depends on the previous one and has
metering, and which is economically feasible as well as spec- its algorithm. We model our proposed procedures in the
trally efficient. Accordingly, it is necessary to add more func- RRC messaging phase by adding two sub-states have a red col-
tions to the signalling dictionary, which could enable our between the MSG3 and MSG4 sub-states, as depicted in
negotiation between nodes to achieve efficient operations in Fig. 8 during the transporting of the RRC messages via the
addition to the architectural changes proposed. MAC layer as well as after decoding and encoding the required
The proposed changes summarize in identifying the follow- information fields. These algorithms are implemented using
ing four new information fields that will be carried by the RRC MATLAB scripts that are linked as files with our Simulink
CONNECTION REQUEST: project. Fig. 9 presents the realization of the proposed model
Fig. 8 The comparison of the State level of RRC Messaging Phase before and after our proposal.
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
NB-IoT optimization for smart meters networks of smart cities: Case study 7
Fig. 9 Modelling of ‘‘RRC Messaging Phase – after proposal” using State flow toolbox of Simulink.
using State flow toolbox of Simulink. Table 1 describes the Phase 1: First Setting – during this phase, the start and end
notations in each algorithm. In the following section, we pre- of the first periodic transmission is determined and to
sent a description of the algorithm of each phase. accomplish the synchronization with the network. The
transmission duration and transmission end time are deter-
Initializations: the base station sets and creates the following mined according to the cluster payload.
variables:
Clusters Time Table Index = 1
Create Transmission Map array
Create Clusters Time Table matrix [4 n]
Algorithm 1: First Setting
Input: p,
2 3 Output: Cnstart , Cnstop
CID Cn
1: Compute Urf , Urf = Buffer-Size-Indexing (p)
6C Cn 7
6 p 7 2: if Clusters Time Table Index = 1 then
Clusters Time Table ¼ 6 7 ð1Þ
4 Urf Cn 5 3: tC1 = to, where n = 1
4: else tCn = tCn-1 + Urf-1
tCn Cn
5: end if
6: Cnstart = tCn
7: Cnstop = tCn + Urf
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
8 A.M. Abbas et al.
Fig. 10 The comparison of the Smart Meter sub-block Simulink model – State level before and after our proposal.
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
NB-IoT optimization for smart meters networks of smart cities: Case study 9
Fig. 11 The geographical location of the case study, (a) Satellite picture and (b) Google Map Layout depicts the exact total location area
and its perimeter distance.
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
10 A.M. Abbas et al.
Fig. 12 The screenshot depicts the comparison between the three scenarios for band resources reservation of data channels during one
second, (A) Traffic of SEM group by current scheduling scheme, (B) Traffic of SWM group by current scheduling scheme, and (C) Traffic
of SEM group & SWM group in proposed scheduling scheme.
(a)–(c) shows the three simulated scenarios. As previously the second duration. The third scenario presents the allocated
mentioned, the first and second scenarios follow the standard, channels for both groups and the allocated radio frames for
and the third follows our proposed scheme. According to our each group after applying our proposal. Fig. 12(c) shows that
model, the first scenario shows that all SWM Group UEs only our proposed approach utilized the resources wasted for sig-
succeeded to dominate all single-tone bands, and all SEM nalling and higher layer control messaging at each transmis-
Group UEs failed. The small blue solid circle represents the sion process. These resources were used instead in the
allocated resources for uplink RRC messaging. Then, the large transfer of more periodic IoT traffic. Also, the proposed
blue solid circle represents the allocated resources for the approach maximized the utilization of radio frame resources.
uplink payload. The second scenario shows that all SEM Fig. 13 presents the enhancement in the NB-IoT band
Group UEs only succeeded to dominate all single-tone bands, single-tone utilization with respect to the number of periodic
and all SWM Group UEs failed. The small green dotted circle NB-IoT UEs (in our case study are called SM terminals or
represents the allocated resources for uplink RRC messaging, SM UEs) during a 1-h (one hour) duration, where for the
and the large green dotted circle represents the allocated SWM Group traffic that uses the standard method the spec-
resources for the uplink payload. In Fig. 12(a) and (b), both trum saturated at 9.88%. Moreover, even when the payloads
the small and large red squares represent the tones and its in the SEM Group traffic increased with using the standard,
radio frames that are used to carry the control messages and the spectrum saturated at 22.05%. After applying our pro-
data payload. From Fig. 12(a) and (b), we observe free radio posed approach, the spectrum saturated at 39.52%. Conse-
frames, which represent unutilized tones during the rest of quently, the increase in channel utilization reached 17.47%.
Fig. 13 Comparison of NB-IoT Resources Utilization for the current standard and the proposed scheme.
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
NB-IoT optimization for smart meters networks of smart cities: Case study 11
Fig. 14 Comparison of NB-IoT system throughput for current standard and the proposed scheme.
3.2.2. Throughput terminals doubled after our proposal, which they increased
The utilization enhancement reflects on NB-IoT system from 720 terminals to 1680 after the proposal.
throughput. Fig. 14 depicts the system throughput versus
increasing the numbers of SM UEs. For SWM Group pay- 3.2.3. Accessibility
loads, the throughput is 10.7 Mbps. Even with an increased
Fig. 17 presents the RRC Setup Success percentage versus
payload in SEM Group, the maximum achievable throughput
increasing the numbers of SM UEs. The figure depicts the
is 23.82 Mbps when using the standard. After applying our
enhancement of the RRC Setup Success percentage from 660
proposed scheme, the maximum achievable throughput
SM terminals to 100% success at 1540 SM terminals.
becomes 42.49 Mbps. Fig. 15 presents the relation between
the spectral efficiency and the increased numbers of SM
UEs, where the maximum spectral efficiency is 5.95bb/s/Hz 3.2.4. Retainability
for the SEM Group traffic when using the 3GPP standard. Fig. 18 represents the relation between Session Drop Rate
The maximum spectral efficiency becomes 10.62b/s/Hz with (SDR) and increasing the numbers of SM UEs. Also, improving
our proposed approach. Additionally, the number of SM the utilization contributed to save and increase the resources not
UEs that can be served until the NB-IoT spectrum reach to only for more smart meters devices but also to complete their
saturation is depicted in Fig. 16. Where, the numbers of SM session. Where, at our proposed optimization, the SDR starts
Fig. 15 Comparison of NB-IoT Spectral efficiency for current standard and the proposed scheme.
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
12 A.M. Abbas et al.
Fig. 16 Comparison of maximum numbers of served NB-IoT UEs for current standard and the proposed scheme.
Fig. 17 Comparison between RRC Setup Success Rate for the current standard and the proposed scheme.
at 560 devices with the lowest SDR rate, as depicted in the figure, tion achieved was 22.05% then after the proposal
instead of 240 devices at standard procedures. became 39.52%. The enhancement in the utilization
strongly influenced on system throughput and spectral
3.3. Results discussion efficiency.
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
NB-IoT optimization for smart meters networks of smart cities: Case study 13
Fig. 18 Comparison of Session drop rate for current standard and the proposed scheme.
(2) The second main KPI is the number of served smart (1) On the IoT terminal side, it reduces the signalling and
meter terminals from different utilities which simultane- access request burden on SM terminals at each transmis-
ously request access. We examined our optimization on sion request,
one of the major applications, as the smart meters for (2) On the NB-IoT eNB side, it enhances the system spectral
two different utilities that serve the described geograph- efficiency,
ical area. The number of SMs doubled after our pro- (3) At both the smart meter terminal and NB-IoT eNB, it
posal. Where, the served SMs increased from 720 will result in a positive impact on energy efficiency,
terminals to 1680 terminals after the proposal. Only
20% of the eNB total bandwidth served this number Well scheduling for radio frames, avoid the accumulative
of SM terminals. Also, this optimization was validated delay due to multiple trials of random access requests until suc-
by RRC Setup Success percentage KPI. Where, we com- cess to acquire data channel even with using alternative tech-
pared our optimization values with the model in [28]. niques as D2D [37].
Where, the lowest probability of channel busy was
9.37% and could serve only 5 IoT terminals per one 4. Conclusion
channel with payload size 512Byte. However, our pro-
posal and with the same calculation methodology served
The future increasing demands’ for smart meters businesses
7 SM terminals per one channel per second.
require maximizing the resources utilization to serve maximum
(3) The third main KPI is SDR. We achieved maximum
possible numbers of NB-IoT UEs. Additionally, the traffic man-
served terminals with zero session drops reached to
agement subject over NB-IoT for efficient spectrum is a newly
560 SMs of 1560 SMs of RRC setup success. Our pro-
opened issue at IoT 5G. In this paper, we achieve two primary
posal enhanced the session drop by 35.7% of only
goals. The first one is to improve the NB-IoT spectral efficiency
20% of the eNB total bandwidth, referring to the session
to serve the increasing numbers of smart meters. The second one
drop probabilities in reference [29]. Where, in reference
is to reduce the signalling burden during each transmission
[29], the session drop probability was 23% for 60%
request for smart meters businesses. Therefore, this paper intro-
NB-IoT static reservation from total eNB bandwidth.
duces a new integrated approach to enhance the NB-IoT spec-
So, their session probability will increase by 69% if they
tral efficiency that serves the smart meters applications. First,
used only 20% of eNB total bandwidth for NB-IoT.
we model the protocol stack of the NB-IoT access network by
However, our proposal achieved only 33.33% session
using the state-machine modelling methodology on the Simu-
drop probability with only 20% of eNB total bandwidth
link environment. According to our goals, we focus on the
for NB-IoT.
NB-IoT uplink scheduler in MAC layer model. For fair results,
Whatever the payload increase or uses of other techniques, we divide the cell spectrum of NB-IoT equally on the four data
many radio frames are still not effectively utilized (even when rates. Also, we assume ideal propagation without any transmis-
they will be allocated later). A comparison of our records with sion impairments. The simulation results are verified according
the previously mentioned methods shows the ability of our to 3GPP standard and are validated with other trusted of differ-
approach to utilize the resources wasted for signalling and ent modelling technique. Second, we optimize the NB-IoT
higher layer control messaging at each transmission process uplink scheduler, which exploited the periodicity nature of
and to maximize the utilization of radio frame resources. smart meters. We executed that by proposing an integrated
Therefore, our proposed approach achieves the following: scheduling protocol which rearranges the transmission times
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
14 A.M. Abbas et al.
of different smart meters utilities and draws a map for the trans- 2; GPP Technical Specifications document (TS) 6.00/r14 (14.4.0)
mission schedule of them. Accordingly, the comprehensive uti- (September 2017).
lization of free time resources achieves. We applied the [13] 3rd Generation Partnership Project; Technical Specification
optimization on a case study for a designated geographical loca- Group Radio Access Network; Evolved Universal Terrestrial
Radio Access (E-UTRA); User Equipment (UE) procedures in
tion has smart meters for two different utilities. The proposed
idle mode; GPP Technical Specifications document (TS) 6.04/
approach encompasses three visions that comprise signalling, r14 (v 14.4.0) (September 2017).
architecture, and algorithm. Each vision represents an update [14] 3rd Generation Partnership Project; Technical Specification
on in current NB-IoT 3GPP standard. The optimization is eval- Group Radio Access Network; Evolved Universal Terrestrial
uated in term of the number of smart meters terminals versus Radio Access (E-UTRA); User Equipment (UE) radio access
four KPIs; resources utilization, spectral efficiency, RRC setup capabilities; GPP Technical Specifications document (TS) 6.06/
success percentage, and session drop rate. Finally, the simula- r14 (v 14.4.0) (September 2017).
tion results prove the ability of our proposal to increase the spec- [15] 3rd Generation Partnership Project; Technical Specification
trum utilization to 17.47% according to the used traffic profiles. Group Radio Access Network; Evolved Universal Terrestrial
This enhancement reflects on spectral efficiency that improved Radio Access (E-UTRA); User Equipment (UE) Radio
Transmission and Reception; GPP Technical Specifications
two times.
document (TS) 6.101/r15 (v 15.0.0) (September 2017).
[16] 3rd Generation Partnership Project; Technical Specification
Declaration of Competing Interest Group Radio Access Network; Evolved Universal Terrestrial
Radio Access (E-UTRA); Base Station (BS) Radio
The authors declare that they have no known competing Transmission and Reception; GPP Technical Specifications
financial interests or personal relationships that could have document (TS) 6.104/r15 (v 15.0.0) (September 2017).
appeared to influence the work reported in this paper. [17] 3rd Generation Partnership Project; General Packet Radio
Service (GPRS) Enhancements for Evolved Universal
Terrestrial Radio Access Network (EUTRAN) access; GPP
References Technical Specifications document (TS) 2.401/r15 (v 15.2.0)
(December 2017).
[1] S. Verma, Y. Kawamoto, N. Kato, Energy-efficient group [18] A. Nassar, Y. Yilmaz, Reinforcement learning for adaptive
paging mechanism for QoS constrained mobile IoT devices over resource allocation in Fog RAN for IoT with heterogeneous
LTE-A pro networks under 5G, IEEE Internet Things J. 6 (5) latency requirements, IEEE Access 7 (2019) 128014–128025.
(2019) 9187–9199. [19] J. Cao, P. Yu, X. Xiang, et al, Anti-quantum fast authentication
[2] https://www.britishgas.co.uk/thesource/yourhome/smarter- and data transmission scheme for massive devices in 5G NB-IoT
living/what-is-a-smart-meter,8-01-2020. system, IEEE Internet Things J. 6 (6) (2019) 9794–9805.
[3] G.A. Akpakwu, B.J. Silva, G.P. Hanck, A survey on 5G [20] J. Xu, J. Yao, L. Wang, et al, Narrowband internet of things:
networks for the internet of things: communication technologies evolutions, technologies, and open issues, IEEE Internet Things
and challenges, IEEE Access 6 (2017) 3619–3647. J. 5 (3) (2018) 1449–1462.
[4] L. Chettri, R. Bera, A comprehensive survey on internet of [21] Y. Zhang, F. Ren, A. Wu, et al, Certificateless multi-party
things (IoT) towards 5G wireless systems, IEEE Internet Things authenticated encryption for NB-IoT terminals in 5G networks,
J. 7 (1) (2020) 16–32. IEEE Access 7 (2019) 114721–114730.
[5] I.B.F. de Almeida, L.L. Mendes, J. Rodrigues, 5G waveforms [22] O. Kodheli, S. Andrenacci, N. Maturo, et al, An uplink UE
for IoT applications, IEEE Commun. Surv. Tutorials 21 (3) group-based scheduling technique for 5G mMTC systems over
(2019) 2554–2567. LEO satellite, IEEE Access 7 (2019) 67413–67427.
[6] A. Ghosh, A. Maeder, M. Baker, 5G evolution: a view on 5G [23] S.K. Goudos, M. Deruyck, D. Plets, et al, A novel design
cellular technology beyond 3GPP release 15, IEEE Access 7 approach for 5G massive MIMO and NB-IoT green networks
(2019) 127639–127651. using a hybrid Jaya-differential evolution algorithm, IEEE
[7] S. Li, Q. Ni, Y. Sun, et al, Energy-efficient resource allocation Access 7 (2019) 105687–105700.
for industrial cyber-physical IoT systems in 5G era, IEEE Trans. [24] F. Al-Turjman, E. Every, H. Zahmatkesh, Small cells in the
Ind. Inf. 14 (6) (2018) 2618–2628. forthcoming 5G/IoT: Traffic modelling and deployment
[8] Q. Wang, D. Chen, N. Zhang, et al, LACS: A lightweight label- overview, IEEE Commun. Surv. Tutorials 21 (1) (2019) 28–65.
based access control scheme in IoT-based 5G caching context, [25] M. Vilgelm, S.R. Liñares, W. Kellerer, Dynamic binary
IEEE Access 5 (2017) 4018–4027. countdown for massive IoT random access in dense 5G
[9] B. Alzahrani, W. Ejaz, Resource management for cognitive IoT networks, IEEE Internet Things J. 6 (4) (2019) 6896–6908.
systems with RF energy harvesting in smart cities, IEEE Access [26] N. Xia, H. Chen, C. Yang, Radio resource management in
6 (2018) 62717–62727. machine-to-machine communications—a survey, IEEE
[10] 3rd Generation Partnership Project; Technical Specification Commun. Surv. Tutorials 20 (1) (2018) 791–828.
Group GSM/EDGE Radio Access Network; Cellular System [27] M. Centenaro, L. Vangelista, S. Saur, et al, Comparison of
Support for Ultra Low Complexity and Low Throughput collision-free and contention-based radio access protocols for
Internet of Things; GPP Technical Report document (TR) the internet of things, IEEE Trans. Commun. 65 (9) (2017)
45.820/r1 (v 1.1.0) (August 2015). 3832–3846.
[11] 3rd Generation Partnership Project; Evolved Universal [28] Y. Sun, F. Tong, Z. Zhang, Throughput modeling and analysis
Terrestrial Radio Access (E-UTRA); NB-IoT; Technical of random access in narrow-band internet of things, IEEE
Report for BS and UE Radio Transmission and Reception; Internet Things J. 5 (3) (2018) 1485–1493.
GPP Technical Specifications document (TS) 6.802/r1 (v 1.0.0) [29] V. Begishev, V. Petrov, A. Samuylov, et al, Resource allocation
(June 2016). and sharing for heterogeneous data collection over conventional
[12] 3rd Generation Partnership Project; Technical Specification 3GPP LTE and emerging NB-IoT technologies, Comput.
Group Radio Access Network; Evolved Universal Terrestrial Commun. (2018).
Radio Access (E-UTRA) and Evolved Universal Terrestrial [30] B.-Z. Hsieh, Y.-H. Chao, R.-G. Cheng, Design of a UE-Specific
Radio Access Network (E-UTRAN); Overall description; Stage Uplink Scheduler for Narrowband Internet-of-Things (NB-IoT)
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030
NB-IoT optimization for smart meters networks of smart cities: Case study 15
systems, in: 2018 3rd International Conference on Intelligent [42] LTE; Evolved Universal Terrestrial Radio Access (E-UTRA);
Green Building and Smart Grid (IGBSG), 22-25 April 2018, Yi- Radio Resource Control (RRC); Protocol Specification; 3GPP
Lan, Taiwan, Technical Specifications document (TS) 3GPP TS 36.331/r14 (v
[31] J. Lee, J. Lee, Prediction-based energy saving mechanism in 14.2.2) (May 2017).
3GPP NB-IoT networks, Sensors 2008 (17) (2017). [43] LTE; Evolved Universal Terrestrial Radio Access (E-UTRA);
[32] X. Chen, Z. Li, Y. Chen, Performance analysis and uplink Medium Access Control (MAC) Protocol Specification; 3GPP
scheduling for QoS-aware NB-IoT networks in mobile Technical Specifications document (TS) 36.321/r14 (v 14.2.1)
computing, IEEE Access 7 (2019) 44404–44415. (May 2017).
[33] P.A. Maldonado, P. Ameigeiras, J.P.-Garzon, et al., Optimized [44] https://www.google.com/maps/@30.0648042,31.485601,17.25z,
LTE data transmission procedures for IoT: device side energy 4/1/2020.
consumption analysis, in: 2017 IEEE International Conference on [45] M. Lauridsen, H. Nguyen, B. Vejlgaard, et al., Coverage
Communications Workshops, May 21–25, 2017, Paris, France. comparison of GPRS, NB-IoT, LoRa, and SigFox in a 7800
[34] G. Tsoukaneri, M. Condoluci, T. Mahmoodi, et al, Group km2 area, 2017 IEEE 85th Vehicular Technology Conference
communications in narrowband-IoT: architecture, procedures, (VTC Spring), 4–7 June 2017, Sydney, NSW, Australia.
and evaluation, IEEE Internet Things J. 1 (1) (2018) 1–10. [46] K. Mekki, E. Bajic, F. Chaxel, A comparative study of LPWAN
[35] P. Duan, Y. Jia, L. Liang, et al, Space-reserved cooperative technologies for large-scale IoT deployment, Elsevier-The
caching in 5G heterogeneous networks for industrial IoT, IEEE Korean Institute of Communications and Information
Trans. Ind. Inf. 14 (6) (2018) 2715–2724. Sciences (KICS), ICT Express 5 (2019) 1–7.
[36] J. Cao, P. Yu, M. Ma, Fast authentication and data transfer [47] R.S. de Carvalho, P.K. Sen, Y.N. Velaga, et al, Communication
scheme for massive NB-IoT devices in 3GPP 5G network, IEEE system design for an advanced metering infrastructure, MDPI
Internet Things J. 6 (2) (2019) 1561–1575. Sens. 3734 (18) (2018).
[37] B.M. ElHalawany, O. Hashad, K. Wu, Uplink Resource [48] W. Luan, D. Sharp, S. LaRoy, Data traffic analysis of utility
Allocation for Multi-Cluster Internet of Things Deployment smart metering network, in: 2013 IEEE Power & Energy Society
Underlaying Cellular Networks, Springer Science + Business General Meeting, 21–25 July 2013, Vancouver, Canada.
Media, LLC, Part of Springer Nature, Mobile Netw. Appl. (25) [49] eNodeB V100R005C00 KPI Reference, Huawei Technologies
(2020) 300–313. Co., Ltd, Issue 01, 2012-03-30.
[38] F. Soleymani, A. Akgül, E.K. Akgül, On an improved [50] Universal Mobile Telecommunications System (UMTS); LTE;
computational solution for the 3D HCIR PDE in finance, Telecommunication Management; Key Performance Indicators
Analele Universitatii‘‘ Ovidius” Constanta-Seria Mathematica (KPI) for Evolved Universal Terrestrial Radio Access Network
27(3), 207–230. (E-UTRAN): Definitions, 3GPP Technical Specifications
[39] A. Akgül, A. Cordero, J. R. Torregrosa, A fractional Newton document (TS) 32.450/r14 (v 14.0.0) (April 2017).
method with 2ath-order of convergence and its stability, [51] Universal Mobile Telecommunications System (UMTS); LTE;
Elsevier-Appl. Math. Lett. 98, 344–351. Telecommunication Management; Key Performance Indicators
[40] D. Baleanu, A. Fernandez, A. Akgül, On a fractional operator (KPI) for Evolved Universal Terrestrial Radio Access Network
combining proportional and classical differintegrals, MDPI- (E-UTRAN); Requirements, 3GPP Technical Specifications
Mathematics 8(3), 360. document (TS) 32.451/r14 (v. 14.0.0) (April 2017).
[41] Long Term Evolution IN BULLETS, Chris Johnson,
Northampton, England, 2 (1) (2012).
Please cite this article in press as: A.M. Abbas et al., NB-IoT optimization for smart meters networks of smart cities: Case study, Alexandria Eng. J. (2020), https://
doi.org/10.1016/j.aej.2020.07.030