Hindawi

Wireless Communications and Mobile Computing

Volume 2022, Article ID 7387737, 9 pages
https://doi.org/10.1155/2022/7387737

Research Article
The Handover and Performance Analysis of LTE Network with
Traditional and SDN Approaches

Muhammad Emran,1,2 Vijey Thayananthan ,2 Muhammad Umair,2 Ivan Kotuliak,1

Muhammad Shuaib Qureshi ,3 and Muhammad Bilal Qureshi 4
1
Faculty of Informatics and Information Technology, Slovak Technology University, Bratislava, Slovakia
2
Faculty of Computing & IT, King Abdulaziz University, Jeddah, Saudi Arabia
3
Department of Computer Science, School of Arts & Sciences, University of Central Asia, Kyrgyzstan
4
Department of Computer Science & IT, University of Lakki Marwat, Lakki Marwat 28420, KPK, Pakistan

Correspondence should be addressed to Muhammad Shuaib Qureshi; [email protected]

Received 8 February 2022; Revised 4 March 2022; Accepted 14 March 2022; Published 14 April 2022

Academic Editor: Muhammad Zubair Asghar

Copyright © 2022 Muhammad Emran et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.

The mobility of user equipment (UE) in cellular network is a challenging issue in terms of its management. Current traditional
handover in Long-Term Evolution (LTE) network is managed by evolved Node B or eNodeB (eNB) which is a decentralized
solution. In contrast to existing technology, software defined network (SDN) has the capability of serving the packets of the
switching equipment without involving the SDN controller except for the first one. We have proposed an SDN-based
centralized solution for handover management in LTE network. Based on our solution, the handover is being managed by
SDN controller which keeps track of the overall network management and dictates the flow entries to the OpenFlow switches
in the network. In our testbed, two UEs are connected to two eNBs, and one of the UE performs a handover from one eNB to
other eNB. It enhances the performance of the network in terms of reducing the delay while performing the handover and
increasing the data rate of the running application. The initial delay will be bit higher due to the initial flow entry absence in
the flow tables of the switches; later, the delay will be reduced. It is evident from the results that our approach keeps track of
the overall network at centralized controller with improved performance.

1. Introduction fixed computing devices to wireless and mobile communica-

tion devices. The ease in use of ICT applications is supported
With the rapid growth in telecommunication industry, the by high-speed telecommunication networks like 4G/LTE,
generations of mobile network have evolved into multiple LTE Advanced, LTE Advanced Pro, and 5G. According to
generations 1G, 2G, 3G, 4G, and 5G where G stands for Huawei, in 2040, most of the presentation by humans will
generation and numbers 1, 2, 3, 4, and 5 represent the gen- be conducted by hologram technology using 6G networks.
eration number. The performance analysis of SDN-based This evolution in the field of telecommunication is the result
handovers in LTE network enhances the users’ potential of rapid advances in research and development in the field of
problems such as decision-making of autonomous mobile wireless and mobile communication.
vehicles (drones, driverless vehicles, etc.). The LTE network 3rd Generation Partnership Project (3GPP) is Interna-
merging with SDN technology is the evolution of the Tradi- tional Standards Organizational body which deals with the
tional LTE network. standardization of the protocols and standards in field of
The fast growth in Information and Communication telecommunications and was founded in 1998. This organi-
Technology (ICT) leads the use of ICT applications from zation is responsible for the development and maintenance
2 Wireless Communications and Mobile Computing

of the following standards in the telecommunication voice calling on 4G and Wi-Fi depending upon the availabil-
industry: ity of either 4G or Wi-Fi network.
For the intertechnology handover, 2G/3G technology is
(i) Global System for Mobile Communication (GSM) extended by adding various entities including Mobility Man-
which covers 2G and 2.5G standards, including Gen- agement Entity (MME), Serving Gateway (S-GW), Packet
eral Packet Radio Service (GPRS) and Enhanced Data Network Gateway (PDN-GW), Home Subscriber
Data rates for GSM Evolution (EDGE) Server (HSS), and Policy and Charging Rules Function
(PCRF) in the EPC of 3G/4G.
(ii) Universal Mobile Telecommunication Systems
(UMTS) and related 3G standards, including High
Speed Packet Access (HSPA) and High Speed Packet 1.2. 5G Mobile Core Network. The 5G deployment can be
Access Plus (HSPA+) carried out in two possible ways. First is standalone deploy-
ment in which it uses independent 5G core network with
(iii) Long-Term Evolution (LTE) and its enhanced ver-
new 5G new radio (NR). This approach is not very common
sions including 4G standards, LTE Advanced, and
right now because it is not a cost-effective solution, but in
LTE Advanced Pro
future, it can replace the LTE network completely. Second
(iv) 5G New Radio (NR) is a new radio access technol- is nonstandalone deployment in which operator uses the
ogy (RAT) and its related 5G standards improved LTE core network along with new 5G NR. This
gives a leverage to mobile operators to use their current
The mobile communication has multiple parts in the LTE architecture. This approach is in use now a days
mobile network: the mobile user equipment (UE), radio commonly.
access network (RAN), and the core part of the mobile Along with the conventional use of mobile applications,
network. 5G will also provide the effective solution for the cloud-
In mobile network, different user equipments are used. based technologies and IoT applications in an optimized
This user equipment has dual wireless interfaces to be manner. With the technological advancements, more and
connected to cellular as well as with Wi-Fi networks based more applications are cloud based, and these applications
upon the availability of these networks. Mobile phones, require low latency which can be efficiently supported by
tables, laptops, and IoT devices come into the category of 5G [2]. The main contributions of this paper can be summa-
wireless and mobile user equipment. rized as follows:
There are different types of base station nodes available
in the RAN part of the mobile networks. As mobile networks (i) We propose the handover strategy in the LTE net-
have evolved from 2G (GSM) to 3G (UMTS), 4G (LTE), and work with different approaches which provide the
5G (NR) over time, so in the mobile RAN, different types of programmable facilities using SDN technology
base stations are used in these generations of cellular net-
work to provide the wireless connectivity to wireless devices (ii) We give a simple calculation method to show and
of the mobile users. prove the handover performance obtained through
In 2G mobile network, it was called base transceiver the proposed model and method
station (BTS). In 3G mobile network or Universal Mobile (iii) To reflect the user’s preferences regarding delay and
Telecommunication System (UMTS) or Evolved UMTS data rate to meet a higher utility, we discuss in detail
Terrestrial Radio Access Network (eUTRAN), it is called the relationship between handover and perfor-
NodeB, which connects with Radio Network Controller mance analysis of LTE network using traditional
(RNS). In 4G (LTE) mobile network, it is called eNB. So, and SDN approaches. Finally, we present the con-
in 4G mobile network, the eNB performs two tasks which cept of an acceptable handover performance and
are performed by NodeB and RNC together in 3G (UMTS) its optimization
mobile network [1].
(iv) To defend the best handover performance, we set up
the experimental procedure in Network Simulator-3
1.1. 4G (LTE) Mobile Core Network. LTE is a telecommuni- (ns-3)
cation standard for wireless broadband communication. The
development and evolution of LTE followed the path based The remainder of this paper is organized as follows.
on the GSM/EDGE/UMTS/HSPA/HSPA+. LTE is a regis- We introduce literature review in Section 2 and propose
tered trademark of European Telecommunication Standards the SDN-based handover model in Section 3. In this
Institute (ETSI). Its downlink data rate is 300 Mbit/s, and research, we have focused on SDN techniques in LTE net-
uplink rate is 75 Mbit/s. LTE provides the seamless handover work. The SDN controller allows us to improve the per-
support for voice and data for GSM, UMTS, and formance of the handovers in different situations and
CDMA2000 technologies. conditions. In Section 4, we explore the technical improve-
Evolved Packet Core (EPC) is an enhanced feature of ment through the necessary experiments and results. This
LTE which provides the capability of voice calling in LTE section also provides the analysis of the results obtained
called VoLTE or 4G calling, and it works with IP Multimedia from the traditional and SDN-based proposed approaches.
Subsystem (IMS). VoLTE is an added advantage to provide In Section 5, we discuss the challenges of the proposed
Wireless Communications and Mobile Computing 3

model with the emerging SDN-based technologies. Some deployment of the SDN control plane supports the collection
conclusions are given in Section 6. of real-time network attributes and mobility information to
calculate the target network, thus reducing unnecessary
2. Literature Review handovers and decreasing the handover delay. Here, the
fuzzy analytic hierarchy process (FAHP) and multipath
According to the 3GPP, latency issues during offloading can transmission control protocol (MPTCP) is applied to select
be improved when 3GPP applies the LTE-WLAN aggrega- the most appropriate target network and execute seamless
tion (LWA). Paper [3] proposed the SDN based approach handover, respectively. A FAHP- and MPTCP-based seamless
for LWA, named as LWA under SDN Assistance (LWA- handover method in LTE and WLAN is proposed for multiple
SA), which supports to improve the latency issues during service scenarios. The FAHP-based network selection algorithm
the offloading and handover issues. In this approach, and MPTCP-based handover mechanism support a compre-
LWA-SA maximizes the throughput of user equipment, hensive decision and service continuity during handover,
and SDN initiates aggregation appropriately between LTE respectively.
and an optimal wireless local area network (WLAN) access The original SDN controller enhances the facilities of
point (AP), which avoids frequent reconnections and handover processing and simplifies the traffic between wire-
deprived services. less interfaces on one node. The extended SDN controller
Reference [4] shows that distributed coordination proves that it provides a promising solution for handover
between colocated wireless domains using either single or in the heterogeneous wireless network [11]. When handover
multiple radio technologies (such as Wi-Fi and LTE) between two Wi-Fi interfaces is considered, the packet loss
supports the service provider for customer infrastructure was better, but when a handover between Bluetooth and
sharing. These days, much of the handover traffic is off- Wi-Fi is applied, the packet loss rate was increased up to
loaded from the cellular LTE spectrum to Wi-Fi provided 37.2%.
by APs. Cross-layer OpenFlow switch handover algorithm for
The main objective of Session Initiation Protocol (SIP) is SDWNs helps the controller to function effectively and to
to support the mobility management protocols to optimize drop the packet loss during the handover [12]. It also mini-
the handover process between heterogeneous networks ver- mizes the hardware complexity and controller failure scenar-
tical handover (VHO) such as Wi-Fi. Despite the optimiza- ios occurring due to high control traffic load and/or poor
tion techniques, handover takes more time with the SIP, link quality. Authors in [13] introduced the SDN-based
but paper [5] focuses on minimizing the handover time Hierarchical MIPv6 (HMIPv6) as a proposed approach for
using vertical handover architecture without SIP. improving the handover in HMIPv6 networks. The main
According to [6], the performance of Internet usage by target here is to reduce the involvement of the mobile node
using VHO between the cellular network and WLAN is (MN) in the handover process while still benefiting from
evaluated. When the data rate of WLAN is more than the layered architecture of HMIPv6.
one of cellular network, VHO enhances the throughput. Although paper [14] focuses on handover authentication
In order to improve network performance and handover with the SDN, the technique used for developing handover
delay in a mobile environment, integrating handover man- authentication is useful to 5G-based heterogeneous network
agement with SDN would be better because the SDN concept (HetNet) and other networks. SDN-based 5G HetNet, which
simplifies the architecture and design of the integrated supports the large number of low-power small cells includ-
system [7]. Also, SDN introduces mechanisms to improve ing picocell, femtocell, microcell, and other access points
the connections, client’s mobility, load balancing, and Wi- such as Wi-Fi, is deployed in the coverage of macrocell.
Fi network management. Regarding the performance evalua- The virtual representation of the MNs (vMN) not only
tion of handover association mechanisms, SDN played an supports the handover process of the physical MNs but also
important role during the integration, which includes the act as proxies for delivering quick information to other clos-
Mininet-Wi-Fi. In this handover mechanism, the authors est MNs. Here, the SDN controller simplifies and manages
measure three performance metrics, i.e., transfer (KB), jitter the handover processing and procedures efficiently. Within
(ms), and packet loss (%). With an extension to Mininet as the vMN environments, SDN controller uses this informa-
the most popular SDN emulation environment, paper [8] tion for improving the optimized handover decisions, which
claims that the performance of Least-Loaded-First (LLF) is influence the technology-agnostic flow mobility manage-
a better result than Strongest-Signal-First (SSF). But ment mechanism for heterogeneous networks [15]. Accord-
Mininet-Wi-Fi does not provide the support for LTE ing to [16], mobility SDN (M-SDN) reduces the traffic pause
network. time caused by a host-initiated layer-2 handover considered
According to [9], the authors introduced the virtualization- in the SDN-based enterprise network. Here, the authors
based seamless handover in WLAN to improve seamless claim that M-SDN improves handover performance, such
mobility. Also, they proposed the converged SDN-based archi- as delay, latency, and packet loss.
tecture, including WLAN and LTE networks, and designed the The handover management framework given in [17] is
handover workflow, which supports to measure the seamless known as a HuMOR, which can create, validate, and evalu-
convergence during the seamless mobility. The paper [10] ate handover algorithms that preserve QoS. The authors
proposed a seamless handover method for heterogeneous wire- proposed ABRAHAM, a mAchine learning Backed multi-
less networks based on SDN architecture. In this method, the metRic proActive Handover AlgorithM. Although SDN
4 Wireless Communications and Mobile Computing

dominates handover procedures in the HuMOR, a machine address the problems related to these mechanisms, especially
learning algorithm called ABRAHAM supports to measure in highly dense and heterogeneous networks. The role of
all handover performance. SDN is addressing these issues discussed in [24]. In this
A mobile handover mechanism based on fuzzy logic work, the authors address the problems: (1) preservation of
proposed in [18] enhances user satisfaction with SDN session continuity and (2) scalability during handovers.
architecture and MPTCP. To implement seamless handover Three variants of SDN usage were presented, and it is
and improve handover performance, MPTCP can be used. observed that automatic management mechanism increases
It provides better throughput, achieves smoother handover, the robustness of the handover, and network resource utili-
and can be adjusted to lower energy consumption. Also, the zation is optimized.
MPTCP protocol can achieve true seamless handover and In [25], the authors discussed device to device communi-
ensure the quality of service for users. cation scenarios based on SDN principles. The authors
According to [19], the SDN-based mobility management observed that the signaling load is decreased significantly
(SDN-MM) scheme not only enhances handover perfor- in unicasting as well as multicasting if SDN-based architec-
mance but also outperforms the comparison schemes in data tures are used.
communication efficiency. Author shows the comparisons of A SDN-based framework is presented in [26] for cellular
different mobility management schemes for a basic under- networks. In addition to the framework, a mechanism is also
standing of the handover performance. proposed for management of QoS and non-QoS user traffic.
As an application of vehicular communication, the SDN- The proposed algorithm for load distribution in this work is
based handover approach in IEEE 802.11p and LTE net- quite efficient, and the authors claim up to 24% increase in
works improve the handover facilities efficiently. Although the average QoS user downlink throughput. It is further
efficient handover techniques are available, vehicular com- observed that with the proposed system, the desired QoS is
munication needs quick handover decisions in variable envi- achieved and handles the network congestion without any
ronmental conditions. Both LTE and 802.11p interfaces with substantial overhead. In [27], the solution for handover
SDN enhance the handover processing equipped in the failure problem is proposed using QAFT. The algorithm
vehicle. As the handover process happens when the SDN presented is tested with SD-LTE-RAN framework. The
controller monitors the movement of vehicles and cluster authors claim that it increases the throughput by 44% and
information to control the vehicular network. During the reduces the delay by 42% in prioritized handovers compared
handover, network performance may be degraded; to avoid to nonprioritized handover mechanisms. Moreover, it is also
this problem, SDN keeps the transport layer connection capable to maintain around 80% GBR satisfaction to all
unchanged, achieving a seamless handover [20]. networks UEs.
In [21], a horizontal handover algorithm considered A machine learning-based handover technique is pre-
with fuzzy logic control (FLC) enhances the benefits of the sented in [28] to reduce the signaling overhead during
smart handover scheme, which supports improving the handover process. The authors proposed a handover
QoS and QoE. In this scheme, FLC is implemented in the mechanism between LTE and mmWave using machine
centralized SDN controller which supports seamless mobil- learning algorithms to automate the procedure. Further-
ity challenge, such as reduced network complexity, granular more, a classification algorithm is also presented in this
network control, and improved scalability during the hand- work which is improved version of XGBoost. This algo-
over processing. rithm predicts handover success rate based on the channel
In [22], authors presented a handover decision algo- information collected through sampling window. It was
rithm based on LTE-SDN architecture which significantly observed that XGBoost-based handover achieves better
improved the performance of the network with respect results as compared to existing KNN-based handover.
to handover delays as well as number of handovers. The algo-
rithm used splits the handover procedure in two phases, i.e., 3. Proposed SDN-Based Handover Model
preparation and execution for its better management. The
results show a significant improvement range from 16% to To analyze and manage the handover in LTE network, an
24% on different parameters like handover delay, number SDN-based model is proposed. To study and compare the
of handovers, and signaling overheads while compromising data rate change and handover delay in traditional and
RSRQ value which decreased by 4%. SDN-based LTE networks, we developed a model which per-
A handover delay model is presented in [23], which forms the handover of the UE in traditional and SDN-based
measures the delay due to exchange of OpenFlow-related LTE networks. The following two scenarios A and B describe
messages in mobile SDN networks. In this model, two sys- the experimentation details of the traditional and SDN-
tems were analyzed, i.e., Priority Finite Buffering (PFB) based LTE networks to perform a UE handover from one
and single shared buffer without priority (model SFB) with eNB to other eNB. For simulation, we used Network
respect to minimal buffer capacity and total handover Simulator-3 (ns-3).
delay. With the help of simulated results, it was observed
that model PFB performs better than the model SFB when 3.1. Traditional LTE Handover Scenario A. We have created
numbers of users are significantly large. traditional LTE network topology in ns-3 as shown in
Mobility management mechanisms play an important Figure 1 which contains the EPC; Enb1, i.e., (eNB1); Enb2,
role in mobile networks, and it is significantly vital to i.e. (eNB2); and a UE1. The figure shows the handover of
Wireless Communications and Mobile Computing 5

UE1 from Enb2, i.e., (eNB2), to Enb1, i.e., (eNB1), using the 0.0, 0.0 28.0, 0.0 56.0, 0.0
random mobility model. UE1 was initially connected to
Enb2, i.e., (eNB2), and used a video streaming application.
UE1 starts moving from the coverage area of Enb2, i.e.,
(eNB2), towards Enb1, i.e., (eNB1) during its mobility. Once UE1
UE1 reaches in the overlapping coverage area of the two
Enbs, i.e., (eNBs), based on the Received Single Strength Enb2 Enb1
Indicator (RSSI) from both the Enbs, i.e., (eNBs), UE1 per-
forms traditional handover from Enb2, i.e., (eNB2) to 0.0, 34.0 28.0, 34.0 56.0, 34.0
Enb1, i.e., (eNB1). We noted down the change in the data
rate and effect of the handover on the delay faced by the
video streaming application during traditional LTE hand-
over. Later, these results will be discussed in the results sec-
tion and compared with the SDN-based LTE handover.

3.2. SDN-Based LTE Handover Scenario B. In experiment

scenario B, a SDN Controller (OFSWITCH13) is added in 0.0, 68.0 28.0, 68.0 56.0,
0 68.0
the topology of scenario A as depicted in Figure 2. The
Switching Gateways (S-GW) with, which both Enb1, i.e., Figure 1: Traditional LTE network topology. The figure is
(eNB1), and Enb2, i.e., (eNB2), were connected, are replaced reproduced from Emran [29].
with the OpenFlow-based switches Switch1 and Switch2,
respectively. The same experiment of scenario A is repeated
but this time the handover will be assisted or controlled by The delay is minor which is noted specifically only dur-
the SDN controller. For this purpose, an application is writ- ing the handover of the UE1 from Enb2, i.e., (eNB2) to
ten which runs on SDN controller to assist this handover. Enb1, i.e., (eNB1). At start of the handover the delay
The SDN controller collects the required information from observed was 0 sec, which was increased during the hand-
the UE1 and Enbs, i.e., (eNBs). It includes RSSI value on over till 7:575 × 10−8 secs. The net increase in the delay was
UE1, load on each Enb, i.e., (eNB), number of UEs con- 7:575 × 10−8 secs in case of traditional LTE handover. After
nected to each Enb, i.e., (eNB). Currently, the delay faced 3 secs, once the handover was completed, the delay is
by the video streaming application and its data rate is stabilized.
recorded in a file and compared to the scenario A delay Figure 5 shows the delay faced by the video streaming
and data rate. The comparison of the results of scenarios A application in scenario B while performing the SDN-based
and B is given in the experiment result analysis section. LTE handover. The vertical axis shows the delay time, and
The abstract view of the proposed model workflow is the horizontal axis shows the SimTime. The simulations
shown in Figure 3 which presents how the S-GW gets its for this scenario B were run from 1 to 6 secs. The SDN-
follow entries from the SDN controller whether the flow based LTE handover is performed during 2 to 4 secs.
table is missed or hit. In case of hit, the packet will be sent Initially, delay faced before the handover was very low
to destination as per available flow entry, whereas in case during simulation time 1 to 2 secs, i.e., 8:892 × 10−7 sec.
of miss, the flow entry will be updated from the SDN While there is very minor increase in the delay during the
controller. handover which reaches to 2:789 × 10−4 sec, the net increase
In future work, we will use the SDN controller to assist in the delay during the handover was 2:779 × 10−4 sec. More-
the handover in a way that it will check which eNB is over- over, it is evident that the net increase in delay during SDN-
loaded and which eNB is underloaded. It will facilitate the based handover is more as compared to the traditional LTE
handover of the UE to less overloaded eNB. Throughput of handover. One of the possible reasons is that initially the
the network will also be managed more properly by balan- flow table is missed due to the absence of the flow entries
cing the load on the eNBs. We can say that the SDN control- and S-GW contacts the SDN controller to fetch the flow
ler will manage the load balancing while performing the entries which is not the case in traditional handover. Once
handover. the handover is completed after 4 secs, the delay starts
decreasing. This is an additive advantage of using SDN tech-
4. Experiment Results and Analysis nology over the traditional LTE network, which becomes
stabilized even after the handover is completed. The decrease
The graph in Figure 4 shows the delay faced by the video in the delay in SDN-based LTE network after the completion
streaming application in scenario A while performing the of handover is due to the flow table hit for subsequent
traditional LTE handover. For the results presented in packets for the flow of same application.
Figure 4, the simulation time (SimTime) is shown on x To summarize the above presented results, we may con-
-axis and the delay time is taken on y-axis. All simulation clude that the reason for higher delay in SDN-based LTE
runs presented here are taken for the time from 1 to 4 sec- handover is due to the initial flow entries absence in the flow
onds (sec). It is observed that the handover started from tables of the OpenFlow switches (S-GW). When the first
the 2 sec and completed on 3 sec. packet of the application flow is received on the S-GW, the
6 Wireless Communications and Mobile Computing

0.0, 0.0 36.5, 0.0 73.0, 0.0 Delay time analysis

8 × 10–8 Delay
7 × 10–8
6 × 10–8
UE1
5 × 10–8

Delay time
Enb1 Enb2 4 × 10–8
3 × 10–8
0.0, 39.0
Switch1
36.5, 39.0 73.0, 39.0 2 × 10–8
Switch2
1 × 10–8
0
1 1.5 2 2.5 3 3.5 4
Controller Sim time

Figure 4: Traditional LTE handover delay.

0.0, 78.0 36.5, 78.0 0 73.0, 78.0 have the flow entries in its flow table. So, we could say that
this higher delay is incurred while fetching the flow entries
Figure 2: SDN-based LTE network topology. The figure is
reproduced from Emran [29]. from the SDN controller and added in the overall delay of
the SDN-based LTE handover. Otherwise, the SDN-based
handover is having less delay as compared to the traditional
LTE handover overall.
Start
The graph in Figure 6 shows the data rate on the tradi-
tional LTE network while performing the handover in sce-
nario A. The x-axis represents the SimTime in secs where
UE y-axis represents the data rate in kilobits per second (kbps).
The simulation was run from 1 to 4 secs. The traditional
Packet sent handover is performed during 2 to 3 secs.
During simulation time 1 to 2 secs, the data rate was
More No increasing rapidly from 8.219 kbps to 16.437 kbps. Later,
Stop
packets when the traditional LTE handover is performed during
time stamp 2 to 3 secs, the data rate was still increasing but
Yes the increase in the data rate was relatively less as compared
to the initial increase. The value of the data rate increased
eNB - source from 16.437 kbps to 19.375 kbps. It is evident from these
results that the net increase in the data rate before the hand-
Flow entry over was 8.219 kbps whereas during the handover the net
updation increase in data rate is slow down to 2.938 kbps. Once the
S - GW handover is completed at 3 sec, the data rate is further slow
down with the net increase 1.523 kbps and expected to be
stabilized as shown in the graph of Figure 6 at some later
No
time point.
Flow table The graph in Figure 7 shows the data rate in SDN-based
SDN controller
updated ? LTE handover performed in scenario B. The horizontal axis
represents the SimTime in secs where vertical axis represents
Yes the data rate in kilobits per second (kbps). The simulation
was run from 1 to 6 secs. The SDN-based handover is per-
eNB - destination formed during 2 to 4 secs.
Initially, before the handover started, the data rate was
Figure 3: Proposed model workflow. increasing from 2.406 kbps to 4.812 kbps during 1 to 2 secs.
The net increase during this period was 2.406 kbps. More-
over, the data rate was further increasing while the handover
is being performed during simulation time 2 to 4 secs from
flow table is missed, and the first packet is sent to the SDN 4.812 kbps to 9.937 kbps. The net increase during the hand-
controller to fetch the flow entry from the SDN controller, over is 5.125 kbps. Once the handover is completed at 4 sec,
later for subsequent packets of the same application flow, the data rate starts increasing more rapidly till 6 sec; it
delay will be reduced as the OpenFlow-based S-GW will attained the value of 23.468 kbps. This shows that the
Wireless Communications and Mobile Computing 7

Delay time analysis Data rate analysis

0.0003 25
+ +
0.00025
20
+
0.0002 +

Data rate (Kbps)

+
Delay time

15
0.00015
10 +
0.0001
+
5 +
5 × 10
–5

0 0
1 2 3 4 5 6 1 2 3 4 5 6
Simtime Sim time
Delay + Data rate +

Figure 5: SDN-based LTE handover delay. Figure 7: SDN-based LTE handover data rate.

Data rate analysis (i) The SDN-Based Handovers in the Driverless Vehi-
22
cles. The use of SDN configuration supports pro-
20 vides many improvements in automation of
+ handover facilities [30–34]
18
Data rate (Kbps)

(ii) Mobility of Drones and the Interactions. Cell selec-

16 +
tion and handover measurements for drones’ inter-
actions depend on the efficient LTE network
14
emerged with SDN technology. The mobility and
12 handover measurements of drones vary with
heights and altitudes because these basic parameters
10 change the handover frequencies with varying envi-
ronmental and physical conditions [35–39]
8
1 1.5 2 2.5 3 3.5 4 (iii) High Handover Probability. Many moving objects
Sim time
depend on the multiple environmental conditions
which affect the accuracy of the handover measure-
Data rate + ments [40–42]
Figure 6: Traditional LTE handover data rate. These challenges may be addressed using our proposed
model without any major changes. It means legacy of the
proposed techniques will be same in these challenges.
SDN-based LTE handover has less affected the increase in
the data rate during handover. There is no decrease during 6. Conclusion
the SDN-based LTE handover as it was in the traditional
LTE handover (see Figure 6). The proposed solution presented in this article is an exten-
The results obtained from our experiments are promis- sion of the existing work in [43] for handover management
ing that handover technique used in the proposed model is in traditional and SDN-based WLAN network. Using pro-
efficient, and this model allows us to create some challenges posed and existing architectures, we calculated the delay
in the future use cases. and data rate in traditional and SDN-based LTE networks
during the handover of a UE from one eNB to another eNB.
According to results presented in Section 4 show that the
5. Challenges of SDN-Based Handovers delay during the handover in the traditional LTE and SDN-
based LTE networks is increased but this increment in SDN-
The handover techniques are evolving with emerging SDN- based LTE network is less than the traditional LTE network.
based technologies which enchase the efficiency and sensitiv- It is further observed from the presented results that the data
ity of the handover in autonomous devices such as driverless rate increased more rapidly in SDN-based LTE network as
vehicles. The following challenges may be considered with compared to traditional LTE network. This indicates that
our proposed results for future research directions. even if the initial delay is added in the total delay in SDN-
8 Wireless Communications and Mobile Computing

based LTE network due to absence of the flow entries in the [9] L. Wang, Z. Lu, X. Wen, G. Cao, X. Xia, and L. Ma, “An SDN-
flow tables of the OpenFlow switches at the beginning, but based seamless convergence approach of WLAN and LTE net-
still the performance of the SDN-based LTE handover is bet- works,” in 2016 IEEE Information Technology, Networking,
ter than the traditional LTE handover in terms of overall Electronic and Automation Control Conference, pp. 944–947,
delay and data rate. IEEE, Chongqing, China, 2016.
In future, these scenarios may be extended further by [10] H. Tong, X. Liu, and C. Yin, “A FAHP and MPTCP based
incorporating the interwireless technology handover seamless handover method in heterogeneous SDN wireless
between Wi-Fi and LTE networks in the context of hand- networks,” in 2019 11th International Conference on Wireless
Communications and Signal Processing (WCSP), pp. 1–6,
over delay and data rate. The other directions in which
IEEE, Xi'an, China, 2019.
this work can be extended are 5G+ and 6G networks
[11] T. Nguyen-Duc and E. Kamioka, “An extended SDN control-
with/without the M-SDN to enhance the performance of ler for handover in heterogeneous wireless network,” in 2015
the wireless technologies in future communications. 21st Asia-Pacific Conference on Communications (APCC),
pp. 332–337, IEEE, Kyoto, Japan, 2015.
Data Availability [12] Y. Reddy, D. Krishnaswamy, and B. Manoj, “Cross-layer
switch handover in software defined wireless networks,” in
No specific data set is used for conducting this research 2015 IEEE International Conference on Advanced Networks
work. and Telecommuncations Systems (ANTS), pp. 1–6, IEEE, Kol-
kata, India, 2015.
[13] S. F. Hasan, “A discussion on software-defined handovers in
Conflicts of Interest hierarchical MIPv6 networks,” in 2015 IEEE 10th Conference
on Industrial Electronics and Applications (ICIEA), pp. 140–
The authors declare that they have no conflicts of interest. 144, IEEE, Auckland, New Zealand, 2015.
[14] J. Cao, M. Ma, Y. Fu, H. Li, and Y. Zhang, “CPPHA: capability-
References based privacy-protection handover authentication mechanism
for SDN-based 5G HetNets,” IEEE Transactions on Depend-
[1] Commsbrief and A Ghayas, What is the Difference between able and Secure Computing, vol. 18, no. 3, pp. 1182–1195,
NODE B, ENODEB, and GNB, 2019, https://commsbrief 2019.
.com/what-is-the-difference-between-node-b-enodeb-ng- [15] F. Meneses, C. Guimarães, D. Corujo, and R. L. Aguiar,
enb-and-gnb/. “Handover initiation comparison in virtualised SDN-based
[2] Commsbrief and A Ghayas, What Is a Mobile Core Network, flow mobility management,” in 2018 IEEE Symposium on
2019, https://commsbrief.com/what-is-a-mobile-core- Computers and Communications (ISCC), pp. 404–409, IEEE,
network/. Natal, Brazil, 2018.
[3] S. Anbalagan, D. Kumar, G. Raja, W. Ejaz, and A. K. Bashir, [16] C. Chen, Y.-T. Lin, L.-H. Yen, M.-C. Chan, and C.-C. Tseng,
“SDN-assisted efficient LTE-WiFi aggregation in next genera- “Mobility management for low-latency handover in SDN-
tion IoT networks,” Future Generation Computer Systems, based enterprise networks,” in 2016 IEEE wireless communica-
vol. 107, pp. 898–908, 2020. tions and networking conference, pp. 1–6, IEEE, Doha, Qatar,
[4] P. Karimi, W. Lehr, I. Seskar, and D. Raychaudhuri, “SMAP: a 2016.
scalable and distributed architecture for dynamic spectrum [17] E. Zeljković, N. Slamnik-Kriještorac, S. Latré, and J. M.
management,” in 2018 IEEE International Symposium on Marquez-Barja, “ABRAHAM: machine learning backed pro-
Dynamic Spectrum Access Networks (DySPAN), pp. 1–10, active handover algorithm using SDN,” IEEE Transactions
Seoul, Korea (South), 2018. on Network and Service Management, vol. 16, no. 4,
[5] J. Vijayshree and T. Palanivelu, “Vertical handover triggering pp. 1522–1536, 2019.
between WLAN and WIMAX using sip,” in 2014 IEEE Inter- [18] D. Yao, X. Su, B. Liu, and J. Zeng, “A mobile handover mech-
national Conference on Advanced Communications, Control anism based on fuzzy logic and MPTCP protocol under SDN
and Computing Technologies, pp. 769–774, IEEE, Ramanatha- architecture,” in 2018 18th International Symposium on Com-
puram, India, 2014. munications and Information Technologies (ISCIT), pp. 141–
[6] S. Suboonsan and S. Pattaramalai, “Performance evaluation 146, IEEE, Bangkok, Thailand, 2018.
of vertical handover on Bangkok mass transit system,” in [19] Y. Bi, G. Han, C. Lin, M. Guizani, and X. Wang, “Mobility
International Conference on Frontiers of Communications, management for intro/inter domain handover in software-
Networks and Applications (ICFCNA 2014-Malaysia)., Kuala defined networks,” IEEE Journal on Selected Areas in Commu-
Lumpur, 2014. nications, vol. 37, no. 8, pp. 1739–1754, 2019.
[7] R. Bencel, K. Košt'ál, I. Kotuliak, and M. Ries, “Common SDN [20] R. Duo, C. Wu, T. Yoshinaga, and Y. Ji, “SDN-based handover
control channel for seamless handover in 802.11,” in 2018 approach in IEEE 802.11 p and LTE hybrid vehicular net-
Wireless Days (WD), pp. 34–36, IEEE, Dubai, United Arab works,” in 2018 IEEE SmartWorld, Ubiquitous Intelligence &
Emirates, 2018. Computing, Advanced & Trusted Computing, Scalable Com-
[8] H. T. Larasati, F. H. Ilma, B. Nuhamara, A. Mustafa, puting & Communications, Cloud & Big Data Computing,
R. Hakimi, and E. Mulyana, “Performance evaluation of hand- Internet of People and Smart City Innovation (SmartWorld/
over association mechanisms in SDN-based wireless network,” SCALCOM/UIC/ATC/CBDCom/IOP/SCI), pp. 1870–1875,
in 2017 3rd International Conference on Wireless and Tele- IEEE, Guangzhou, China, 2018.
matics (ICWT), pp. 103–108, IEEE, Palembang, Indonesia, [21] O. Aldhaibani, F. Bouhafs, M. Makay, and A. Raschellà, “An
2017. SDN-based architecture for smart handover to improve QoE
Wireless Communications and Mobile Computing 9

in IEEE 802.11 WLANs,” in 2018 32nd International Confer- [36] S. Singh, D. Kedia, N. Rastogi, T. Velmurugan, and
ence on Advanced Information Networking and Applications P. Prakasam, “P2P mobility management for seamless hand-
Workshops (WAINA), pp. 287–292, IEEE, Krakow, Poland, over using D2D communication in B5G wireless technology,”
2018. Peer-to-Peer Networking and Applications, vol. 14, no. 4,
[22] K. Ahmadi, S. P. Miralavy, and M. Ghassemian, “Software- pp. 1988–1997, 2021.
defined networking to improve handover in mobile edge [37] L. D. Manh, N. V. Hoai, and Q. V. Khanh, “Advanced hand-
networks,” International Journal of Communication Systems, over techniques in 5G LTE-A networks,” International Jour-
vol. 33, no. 14, article e4510, 2020. nal, vol. 9, no. 3, 2021.
[23] S. Panev and P. Latkoski, “Performance analysis of hand- [38] A. Fakhreddine, C. Bettstetter, S. Hayat, R. Muzaffar, and
over delay and buffer capacity in mobile OpenFlow- based D. Emini, “Handover challenges for cellular-connected
networks,” International Journal of Communication Systems, drones,” in Proceedings of the 5th Workshop on Micro Aerial
vol. 33, no. 15, article e4529, 2020. Vehicle Networks, Systems, and Applications, pp. 9–14, New
[24] S. Kukliński, Y. Li, and K. T. Dinh, “Handover management in York, 2019.
SDN-based mobile networks,” in 2014 IEEE Globeecom Work- [39] J. Angjo, I. Shayea, M. Ergen, H. Mohamad, A. Alhammadi,
shops (GC Wkshps), pp. 194–200, IEEE, Austin, TX, USA, and Y. I. Daradkeh, “Handover management of drones in
2014. future mobile networks: 6G technologies,” IEEE Access,
[25] H. B. Valiveti and C. Duggineni, “Software defined device to vol. 9, pp. 12803–12823, 2021.
device communication handover-latest advancements,” in [40] S. M. Kala, V. Sathya, E. Yamatsuta, H. Yamaguchi, and
2021 6th International Conference on Inventive Computation T. Higashino, “Operator data driven cell-selection in LTE-
Technologies (ICICT), pp. 1079–1083, IEEE, Coimbatore, LAA coexistence networks,” in International Conference on
India, 2021. Distributed Computing and Networking 2021, pp. 206–214,
[26] F. H. Khan and M. Portmann, “Joint QoS-control and New York, 2021.
handover optimization in backhaul aware SDN-based LTE [41] W. Huang, H. Zhang, and M. Zhou, “Analysis of handover
networks,” Wireless Networks, vol. 26, no. 4, pp. 2707– probability based on equivalent model for 3D UAV networks,”
2729, 2020. in 2019 IEEE 30th Annual International Symposium on Per-
[27] A. K. Rangisetti and V. Sathya, “QoS aware and fault toler- sonal, Indoor and Mobile Radio Communications (PIMRC),
ant handovers in software defined LTE networks,” Wireless pp. 1–6, IEEE, Istanbul, Turkey, 2019.
Networks, vol. 26, no. 6, pp. 4249–4267, 2020. [42] G. Gaur, T. Velmurugan, P. Prakasam, and S. Nandakumar,
[28] N. Nayakwadi and R. Fatima, “Automatic handover execution “Application specific thresholding scheme for handover
technique using machine learning algorithm for heteroge- reduction in 5G ultra dense networks,” Telecommunication
neous wireless networks,” International Journal of Information Systems, vol. 76, no. 1, pp. 97–113, 2021.
Technology, vol. 13, no. 4, pp. 1431–1439, 2021. [43] M. Emran and I. Kotuliak, “Performance analysis of traditional
[29] M. Emran, “Wireless devices handover in SDN based Wi-Fi and SDN based handovers in wireless LAN networks,” in 2020
and LTE networks,” Information Sciences and Technologies IEEE International IOT, Electronics and Mechatronics Confer-
Bulletin of the ACM Slovakia, vol. 13, no. 1, pp. 1–6, 2021. ence (IEMTRONICS), pp. 1–6, IEEE, Vancouver, BC, Canada,
2020.
[30] C. R. Storck, E. D. Efrem, G. D. Guilherme, R. A. Mini, and
F. Duarte-Figueiredo, “FiVH: a solution of inter-V-cell
handover decision for connected vehicles in ultra-dense
5G networks,” Vehicular Communications, vol. 28, article
100307, 2021.
[31] D.-T. Dao, C.-M. Huang, M.-S. Chiang, and V.-T. Nguyen, “A
load-considered handover control scheme for distributed
mobility management (DMM) using software defined network
(SDN) in the vehicular environment,” in 2020 IEEE Eighth
International Conference on Communications and Electronics
(ICCE), pp. 70–74, IEEE, Phu Quoc Island, Vietnam, 2021.
[32] M. Cicioğlu, “Multi-criteria handover management using
entropy-based SAW method for SDN-based 5G small cells,”
Wireless Networks, vol. 27, no. 4, pp. 2947–2959, 2021.
[33] M. Alotaibi and A. Nayak, “Linking handover delay to load
balancing in SDN-based heterogeneous networks,” Computer
Communications, vol. 173, pp. 170–182, 2021.
[34] A. Abdulghaffar, A. Mahmoud, M. Abu-Amara, and
T. Sheltami, “Modeling and evaluation of software defined net-
working based 5G core network architecture,” IEEE Access,
vol. 9, pp. 10179–10198, 2021.
[35] H. Tong, T. Wang, Y. Zhu, X. Liu, S. Wang, and C. Yin,
“Mobility-aware seamless handover with MPTCP in
software-defined HetNets,” IEEE Transactions on Network
and Service Management, vol. 18, no. 1, pp. 498–510, 2021.