New Approach for Mobility Management in

Openflow/Software-Defined Networks

N. Omheni

, F. Zarai

, B. Sadoun

and M. S. Obaidat

3,*

New Technologies and Telecommunications Systems Research Unit, ENET’COM, Tunisia

College of Engineering, Al-Balqa Applied University, Jordan, ECE Department,

Nazarbayev University, Astana, Kazakhstan

Kazakhstan and King Abdullah II School of Information Technology, University of Jordan, Jordan

Keywords: Software-Defined Networking, Distributed Mobility Management, Proxy Mobile Ipv6, Media Independent

Handover, Openflow, SDN Controller, Signaling Cost.

Abstract: The Software-Defined Networking (SDN) paradigm predicts that the evolution of cellular and wireless

networks will bring a considerable increase of two factors, the densification of the Radio Access Network

(RAN) part and the progressive demand for traffic. This rapid evolution has led to the emergence of extremely

complicated systems, where a large number of logic modules must interact to lead to the desired behavior and

the desired quality of service. The key advantage of SDNs is the simplicity of networking and the deployment

of new mechanisms and applications. Furthermore, the programmable aspect on the traffic and devices in

SDNs makes them more efficient and flexible than traditional networks. In this context, Distributed Mobility

Management (DMM) has been recently presented as a new trend to solve the issues of the today’s mobility

management protocols. In this paper, we propose a partially distributed Mobility Management for

OpenFlow/SDN networks. According to simulation results, our approach guarantees a significant reduction

of the number of handover and the signaling cost.

1 INTRODUCTION

We are witnessing in recent years to an explosion of

mobile communications, together with an exponential

usage of Internet for all kinds of applications. It is

expected that by the year 2020, more than 50 billion

smart devices around the world should be connected

and the annual revenues of this new market will

exceed US $8.9 trillion (IoT, 2016). In this evolved

context, a wide range of wireless technologies has

gradually emerged indicating a great diversity in

communication needs in various application domains

and new standards have regularly been proposed.

Moreover, the massive volume of traffic and the

evolutionary aspect of the size of such systems make

any study carried out on these networks a very

difficult task. Thus, both mobile operators, industry

and research communities are trying to discover

innovative solutions and mechanisms to improve

their network efficiency and performance as well as

Fellow of IEEE and Fellow of SCS.

to moderate the costs of new service deployment and

network operation maintenance.

Software-Defined Networking (SDN) and

Network Function Virtualization (NFV) are two

important recent innovations that are expected to

offer such solutions. SDN is a new paradigm for

programmable networking and is adopted extensively

in enterprise networks and data centers. (Ahn, 2017).

Figure 1 depicts the SDN network. In such network,

control and data forwarding functions are managed

separately in order to permit centralized and

programmable network control (Kreutz, 2015).

The major components of the SDN architecture

comprise a data plane involving network resources

for data forwarding, a control plane including SDN

controllers permitting a centralized control of

network resources, and a management applications to

program network operations over a controller. The

interface between the data and the control forwarding

plans is called the southbound interface while the

Omheni, N., Zarai, F., Sadoun, B. and Obaidat, M.

New Approach for Mobility Management in Openﬂow/Software-Deﬁned Networks.

DOI: 10.5220/0006847800250033

In Proceedings of 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH 2018), pages 25-33

ISBN: 978-989-758-323-0

control-application interface is called the northbound

interface.

Mobility management plays a crucial role in such

environments, where mobiles can easily experience

numerous handovers during the same connectivity

session. Consequently, there is a real need to propose

novel mobility management approaches.

Figure 1: Overall main components of SDN architecture.

Mobility management approaches have to be

distributed to provide better reliability and avoid any

network bottleneck or single point of failure.

Here the DMM paradigm presents most of the

above-mentioned features, and specifically accounts

for distributed IP flow mobility managed via

distributed anchoring by separating data and control

planes to address the limitations of current centralized

mobility management such as scalability, reliability

and sub-optimal routing (Liu, 2014).

Several DMM-oriented solutions have been

proposed in the literature as, for example, the

Distributed Mobility Anchoring (DMA) (Seite,

2013), Double NAT (D-NAT) (Liebsc, 2011), Inter-

domain DMM, and Local IP Access (LIPA)/Selected

IP Traffic Offload (SIPTO) (3GPP, 2015).

OpenFlow-based SDN architecture is another

candidate solution that supports DMM features and

outperforms existing solutions. Advantages provided

by OpenFlow, as the most joint communication

protocol used in SDN approach, would be an enabler

to reach the forwarding plane of OpenFlow switches

through the network and reconfigure it in line with the

necessities of applications and network services (The

OpenFlow, 2017).

In this paper, we present a new SDN/OpenFlow

based partially DMM. First, we introduce the

proposed architecture components. Then, we describe

our suggested mobility management procedures for

SDN/OpenFlow composed by two stages: (a)

preparation and registration stage and (b) handover

execution. Our approach belongs to the category of

partially distributed mobility management. The SDN

Controller handles the control plane in a centralized

way. While, the DMM-ARs (Distributed Mobility

management-Access Routers) manage the data plane

in a distributed manner.

The rest of this paper is organized as follows.

Section 2 presents the related work. In section 3, the

new Software Defined Networking (SDN)/OpenFlow

based partially DMM is detailed.OpenFlow

operations, which are executed in the DMM-AR are

detailed in section 4. Performance evaluation is

discussed in Section 5. Finally, Section 6 concludes

the paper.

2 RELATED WORK

2.1 Software-Defined Networking

(SDN)

In this section, we discuss different projects dealing

with wireless and cellular networks that integrate

SDN paradigm. In SoftCell (Jin, 2013), SDN

principles are incorporated in LTE (Long Term

Evolution) core networks. The aim of this framework

is to build simple programmable core switches inside

mobile network while the most of functionalities are

pushed to access switches in the radio access network.

This division between access and other core switches

creates real scalability.

Contrary to SoftCell, which is interested in the

dimensioning of the core network, SoftRAN project

makes use of SDN functionalities to remodel the LTE

radio access network (Gudipati, 2013). In traditional

radio access networks, interference and handovers are

handled using distributed protocols. While this

concept is acceptable in light networks, in ultra-dense

environment it leads to bad performance in terms of

interferences and latencies. In SoftRAN, the entire

LTE network is controlled in a centralized manner.

All e-Node B co-located in a geographical area are

managed as a virtual big-base station. Radio

resources are abstracted out as three-dimensional grid

of space, time and frequency slots. Periodically, all

radio interfaces send information about local network

state to SDN controller. The latter determines the way

SIMULTECH 2018 - 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

to assign radio resources block and transmit power for

each radio element.

Work in (Tomovic, 2017) was recently published.

The authors proposed a new architecture for IoT

(Internet of Thing) combining the two evolving

technologies: Fog computing and SDN. Their goal

was to support a high level of scalability, real-time

data delivery and mobility. Fog computing policy is

considered as the suitable platform for IoT thanks to

its capability to reduce latency for services that need

fast analysis and decision-making. While, SDN is

used due to its centralized aspect to manage control

plane, which permits the execution of sophisticated

techniques for resource management and traffic

control.

In (Ahmad, 2016), authors proposed a new

solution in the form of Software-Defined Mobile

Networks (SDMN) including SDN, NFV (Network

Function Virtualization) and cloud technologies. The

goal is to answer the issues confronted by current

mobile network architectures in integrating diverse

services.

Authors in (Knaesel, 2013) proposed an

OpenFlow based IEEE 802.21 Media Independent

Handover (MIH) to perform link connectivity

establishment in SDN. Nevertheless, 3GPP or ETSI

NFV architectures are not considered in their

framework. In this work, a MIH-enabled Evolved

Packet System was proposed to offer seamless

handovers between 3GPP and non-3GPP wireless

technologies. However, their work does not consider

programmable or virtualized mobile concepts.

2.2 Distributed Mobility Management

(DMM)

There are already many works that have treated

Distributed Mobility Management in the context of

SDN. In (Chen, 2016), a mobility management

approach called Mobility SDN (M-SDN) was

proposed to decrease the traffic pause time produced

by a host-initiated layer-2 handover. The key idea of

this work is that the handover preparation is

performed in parallel with layer-2 handover and

active flows are sent to each potential handover

target. In (Costa, 2013), the authors considered the 5G

as an application group and suggested that mobility

management can be treated as a service on top of the

SDN controller. In (Hampel, 2013), the authors

focused on forwarding data across networking layers

and discussed how to relate SDN to the Telecom

domain. Authors of (Jeon, 2013) presented a new

architecture for SDN based mobile networking with

two kinds of controller model: hierarchical and single

controller. In (Nguyen, 2016), authors proposed a

new DMM solution in the context of 5G networks

based on SDN architecture called S-DMM. In their

proposed architecture, DMM is delivered as a service

deployed on top of SDN controller. Mobile Access

Routers (MAR) are considered as a simple

forwarding hardware and do not necessitate any

mobility-related module. The centralized control

controller permits the operators to control the network

at a reduced complexity level.

In the rest of this subsection, we describe briefly

three main types of DMM solutions, which have been

published as extensions or improvements of already

published standards.

The first type of solutions is known as PMIPv6-

based DMM solution. It is based on amelioration of

classical IP mobility protocols and mainly PMIPv6

(Proxy Mobile IPv6). This protocol manages mobility

in a centralized way where the Local Mobility Anchor

(LMA) in the core network creates bidirectional

tunnels to Mobile Access Gateways (MAGs) fixed in

the radio access networks. More details about this

protocol can be founded in (Bernardos, 2014).

The second family of solutions is inspired by the

SDN context and named SDN-based DMM. In SDN,

network control and data forwarding functions are

managed separately to permit centralized and

programmable network control. With this paradigm,

network administrators are able to program the

comportment of the traffic and the network in a

centralized manner.

The last category of solutions is the routing-based

DMM. The main idea of this kind of solutions is to

eliminate any anchor point from the network

architecture; permitting all nodes in the network to re-

establish a new routing map when terminals change

their location. This family of solutions belongs to the

IP routing protocols.

3 SDN-BASED PARTIALLY

DISTRIBUTED MM (S-PDMM)

3.1 Proposed Architecture Components

In this section, we introduce in detail the proposed

system model for SDN-based partially distributed

MM and we describe the suggested mobility

management procedures. The proposed architecture

is shown in Figure 2. Our SDN mobility management

architecture relies on three main levels:

New Approach for Mobility Management in Openﬂow/Software-Deﬁned Networks

Figure 2: Proposed architecture.

the RANs (Radio Access Networks), the DMM-AR

and the SDN Controller. The proposed architecture is

shown in Figure 2. Our SDN mobility management

architecture relies on three main levels: the RANs

(Radio Access Networks), the DMM-AR and the

SDN Controller.

The first level is composed of SDN enabled WiFi

access points and 3GPP LTE radio access network.

RANs are the heterogeneous access networks, which

include different access technologies such as WiFi

and LTE.

These radio access networks can be

programmable and under the supervision of SDN

Controller in the core network.

The DMM-AR is an access router which not only

delivers connectivity to the RANs (default gateway),

it is also boosted with some specific DMM

functionalities. DMM-ARs act as OpenFlow switches

(OFS) and communicates with the SDN controller

with the southbound Application Program Interfaces

(APIs) to achieve the packet processing and

forwarding feature. The SDN controller with the use

of OpenFlow protocol can apply a set of actions such

as adding, deleting or updating flow entries to the

flow tables in the DMM-AR.

One of the issues that can be observed in the first

level is the coexistence of various complementary

technologies (cellular vs WiFi), in terms of

connection type, shared/dedicated medium and QoS

requirements. In order to closer incorporate QoS

signaling and mobility for these various technologies,

we combined the IEEE 802.21 MIH functions with

the proposed SDN-based partially distributed MM.

The reason of this mapping between MIH and SDN is

that WiFi RANs allow handover only to the same

network types and behind the same domain.

Furthermore, mobility management at upper layers

does not tolerate significant delays to get technology-

independent handover triggers and to select the best

candidate AP.

The aims of MIH standard are to facilitate

mobility management by delivering a technology-

independent interface under the network layer. The

IEEE 802.21 proposes a MIH specification for

achieving seamless handover for mobile users in the

same or in different networks.

The main functionality offered by MIH is a

seamless connection to different RATs. The control

messages are relayed by the Function (MIHF) located

in the protocol stack between layer 2 wireless

technologies and IP layer. This type of deployment

makes it easy for mobile nodes to move between

different Points of Attachment (PoAs).

The third level is composed of SDN controller

which is the key entity in our architecture. The role of

this controller is the management and the

configuration of DMM-ARs through the Southbound

common application programming interface as

mentioned earlier. The northbound API is reserved

for communications between the SDN controller and

the network applications. The Mobility Management

Entity (MME) side core network is integrated in the

SDN controller. The latter interacts with the AAA

server to get the MN’s profile for authentication.

Furthermore, SDN controller supports others control

plane features related to LTE networks such as

Serving Gateway (S-GW), Home Subscriber System

(HSS) and others nodes.

3.2 SDN-based Partially Distributed

MM Procedures

The suggested mobility management procedures are

divided into two stages: (a) The preparation and

registration stage and (b) the handover execution

stage. In the first step, we take advantage of the MIH

standard in SDN context for the optimization of the

MN attachment detection.

In the second stage, the proposed mobility

management approach uses the PMIPv6 protocol for

handover execution. The SDN controller will play the

role of LMA in the PMIPv6 context.

SIMULTECH 2018 - 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

Figure 3: Mobility management operations.

The MAG is replaced by the DMM-AR. Figure 3

shows the proposed mobility management operations.

3.2.1 Preparation and Registration Stage

For the preparation stage, the DMM-AR must be able

to be aware of the existing RANs status. These

actions can be achieved by applying

MIH_Link_Get_Parameters primitives between the

MIHF and its users. MIH_Get_Information message

primitives can ask Handover policies and network

context information from the SDN and the MIIS at

the core network. Upper layers can also command

specific actions for a given radio interface such as

Link Power Up or Link Power down operation.

When a MN moves inside an LTE RAN network,

the radio interface is switched on by running MIH

Link Power Up primitive. This will trigger the

network attachment of the LTE interface to the

DMM-AR and an IP route will be established to the

MN. Registration occurs when the MN changes its

attachment from a previous DMM-AR (p-DMM-AR)

to a new DMM-AR (n-DMM-AR) to preserve

connectivity with the correspondent node (CN). Two

types of messages are exchanged between the MME,

which is integrated in the SDN controller, and the

new DMM-AR. The latter acts as the HA (Home

Agent) of the MN in the PMIPv6 domain during this

process; proxy binding update (PBU) and proxy

binding acknowledgement (PBA) processes.

A DMM-AR incorporates either mobility

anchoring features and is able to forward MN’IP

flows without disruption before moving to a new

DMM-AR. The SDN controller will play the role of

LMA (Local Mobility Anchor) in the PMIPv6 context

and maintain the mobile node’s (MN) reachability

when it is away from home. The MN sends a Router

Solicitation message to its DMM-AR. The latter

exchanges PBU and PBA messages with the SDN

controller. The MN attachment detection is achieved

and new IPv6 prefix is assigned to the MN.

3.2.2 Handover Execution

PMIPv6, in the original version, suffers from several

limitations such as packet loss and long handover

latencies. For that reason, our approach takes benefit

from the wireless link layer triggers to anticipate the

next handover when the MN changes its attachment.

IP flows are directly forwarded to the new DMM-AR

in a pre-established tunnel between the two DMM-

ARs before the new attachment.

In the following, we give the different steps of

handover execution. After the registration stage, the

SDN Controller creates a new entry for the MN. After

receiving the PBU message from the current DMM-

AR, the MN updates its location with the new DMM-

AR. As mentioned previously, IP flows are directly

forwarded to the new DMM-AR in a pre-established

tunnel between the two DMM-AR before the new

attachment. After that, DMM-ARs act as OpenFlow

switches (OFS) and communicates with the SDN

controller with the APIs to achieve the packet

New Approach for Mobility Management in Openﬂow/Software-Deﬁned Networks

processing and forwarding feature and the OpenFlow

rules in the new DMM-AR are configured.

The final step of handover is achieved when rules

on DMM-AR are translated and forwarded. The

visited DMM-AR changes the IP destination address

with the last MN’s IP address to forward traffic in the

new MN’s domain.

Our approach belongs to the category of partially

distributed mobility management. The SDN

Controller handles the control plane in a centralized

way. While, the DMM-ARs manage the data plane in

a distributed manner.

4 OPENFLOW OPERATIONS

DMM-ARs act as OpenFlow switches (OFS). The

comportment of OFS is wholly determined by the

contents of Flow Tables, whose entries identify flows

and the way to analyze packets of these flows. Each

entry comprises three features:

(1) Rules: to match incoming packets to current

flows. Each entry must contain a set of header

values. These are used to look for the flows to

which incoming packets belong.

(2) Actions: This entry determines how to handle

packets belonging to the flows. A packet can be

either forwarded, dropped, or modified.

(3) Statistics: This feature covers statistical data,

which can be used by the SDN controller to adjust

policies.

The flow-table format is illustrated in the Figure

4. Counters field indicates some statistics such as

number of received packets, received bytes and

duration. Three actions are defined depending on the

case (forward the packet, encapsulate and forward the

packet or drop the packet. The priority field in our

approach gives priority of the used radio network.

Figure 4: Flow-table format.

The OpenFlow protocol allows handling different

nodes in the network in right time by executing

configuration changes in the data plane. In OFS, the

latency experienced by data plane features to treat

packets increases due to the transmission delay

between OpenFlow switch and the SDN Controller,

the performance of the latter in terms of processing

speed and finally the receptiveness of OpenFlow

switches in creating new flow and updating their

tables after receiving information from the SDN

Controller (He, 2016).

DMM-ARs check the SDN controller for the first

packet of each new flow. Later all packets belonging

to the same flow will be processed in line with the

flow-table entry (rules) fixed after receiving the first

packet. The traffic between the SDN controller and

DMM-ARs is then considerably reduced by flow-

based traffic.

We consider that the OFSs adopt a pipeline

treatment of the received packets as mentioned in

(Javed, 2017). Three components are defined in this

pipeline treatment. A flow-table to manage and

forward packets, a safe channel to connect a DMM-

AR and the SDN Controller and finally the OpenFlow

protocol, an open standard for communication

between the SDN Controller and DMM-ARs.

Figure 5 illustrates the different steps that the

DMM-AR takes to process a packet. When a packet

Figure 5: Packet processing steps.

SIMULTECH 2018 - 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

arrives at DMM-AR, which acts as an OpenFlow

switch, it checks whether there is an entry matchin its

flow-table. In case of correspondence, one of these

three actions will be excuted (forward the packet,

encapsulate and forward the packet or drop the

packet). If not, the packet will be sent to the SDN

Controller to update the flow-table.

5 SIMULATIONS RESULTS

5.1 Simulation Setup

The proposed approach is evaluated through

simulation. The implementation of the proposed

framework is performed in OMNeT++ 5.0. The

simulation scenarios are as follows: one eNodeB, one

DMM-AR and UEs, where the UEs are dynamically

moving from one place to another.

We conduct the simulation in the area of about

2000 m x 2000 m. To simplify the simulation, we

assume that control messages are of the same size of

a unit length and the distance between the SDN

controller and the DMM-AR is set to 1 hop. Table 1

shows the simulations parameters.

Table 1: Simulations parameters.

Parameters Values

Simulation area

Mobility model

UE distribution

Transmission rate

Distance between eNB and

DMM-AR (hops)

Radius of eNB coverage

Velocity of UEs

2000 m x 2000 m

random walk model

Uniform randomly

100 Mbps

500 m

2 m/s

The performance of our proposed S-PDMM

framework is evaluated based on a comparison of

signaling cost and number of handovers with the

work in (Javed, 2017)

The Signaling Cost (SC) represents the cost of

mobility-related signaling messages for a mobile

node per unit time. The Signaling Cost (SC) can be

calculated as:

 = 

〖





〗

×







Where Message j is the size (in bytes) of a

signaling message j, and Hops j are the average

number of hops traversed by message j. The number

of hops changes for each type of transmission based

on the path selected. Here, one hop is assumed to be

set as the distance between the SDN Controller and

the DMM-AR.The Overall handover indicates the

number of networks that are involved in handover.

5.2 Simulation Results

5.2.1 Signaling Cost

Figure 6 shows the performance results of signaling

cost in the network over the time interval. We can

observe that when the MN moves far away from its

DMM-AR, the signaling cost is notably increased

because it is proportional to the number of switches

along the path from the source to the destination.

We can clearly see that our proposed S-PDMM

framework outperforms the other existing system in

terms of signaling cost. Compared with the (Javed,

2017), signaling cost is approximately reduced by

around 10%.

Figure 6: Signaling cost.

Our approach takes benefit from the wireless link

layer triggers to anticipate the next handover when

the MN changes its attachment.

IP flows are directly forwarded to the new DMM-

AR in a pre-established tunnel between the two

DMM-ARs before the new attachment and without

recourse to additional signaling messages exchange.

This act significantly reduces the signaling cost. In

conclusion, the low signaling cost makes the

proposed scheme more scalable in comparison with

other competing approaches.

5.2.2 Overall Handovers

Overall handover indicates the number of networks

that are involved in handover. Figure 7 shows the

simulation results for the overall handover process in

the network. It can be observed that our proposed

scheme reduces the overall unnecessary handover.

New Approach for Mobility Management in Openﬂow/Software-Deﬁned Networks

Figure 7: Overall Handover.

Even in the cases of increasing load (number of

UEs); the overall handover has been minimized. This

result backs up the innate advantages of the combined

the IEEE 802.21 MIH functions with the proposed

SDN-based partially distributed MM, implied in our

framework.

6 CONCLUSION

In this paper, we have presented an SDN-based

partially distributed mobility management in cellular

and wireless networks. Our solution benefits from the

scalability and the flexibility provided by SDN and

DMM schemes. The proposed solution belongs to the

category of partially distributed mobility

management. The SDN Controller handles the control

plane in a centralized way, while, the DMM-ARs

manage the data plane in a distributed manner.

Moreover, we combined the IEEE 802.21 MIH

functions with the proposed SDN-based partially

distributed MM in order to incorporate QoS signaling

and mobility for the various technologies in the

network. Simulation results prove that our scheme

can guarantee a significant reduction of the number

of handovers and the signaling cost. As a future work,

we will introduce the OpenFlow v1.3.1 software in a

real testbed to better study the performance of the

proposed approach.

REFERENCES

3GPP Technical Specification 23.829, 2015. Local IP

Access and Selected IP Traffic Offload (LIPA-SIPTO)

(Release10). Available: http://www.3gpp.org/Dyna

Report/23829.htm. Last accessed 2017.

Ahmad I. et al, 2016. New concepts for traffic, resource and

mobility management in software-defined mobile

networks. 12th Annual Conference on Wireless On-

demand Network Systems and Services (WONS),

Cortina d'Ampezzo. pp. 1-8.

Ahn C-W. and Chung S-H., 2017. SDN-Based Mobile Data

Offloading Scheme Using a Femtocell and WiFi

Networks. Mobile Information Systems journal. 1 (1),

pp. 1–15..

Bernardos C. J., De La Oliva A., and Giust F., 2014. A

PMIPv6-Based Solution for Distributed Mobility

Management. IETF Internet Draft.

Chen C et al , 2016. Mobility management for low-latency

handover in SDN-based enterprise networks. IEEE

Wireless Communications and Networking Conference,

Doha. pp. 1-6.

Costa J. et al, 2013. Software Defined 5G Mobile Backhaul.

1st International Conference on 5G for Ubiquitous

Connectivity. pp. 258-263.

Gudipati A., Perry D., Li L. and Katti S., 2013. SoftRAN:

Software dened radio access network. ACM SIGCOMM

HotSDN workshop. pp. 25–30.

Guimaraes C., Corujo D., Aguiar R., 2013. Empowering

software defined wireless networks through media

independent handover management. IEEE Global

Communications Conference (GLOBECOM). pp.

2204–2209.

Hampel G., Steiner M., and Bu T., 2013. Applying

Software-Defined Networking to the Telecom Domain.

IEEE Conference on Computer Communications

Workshops (INFOCOM WKSHPS). pp. 133-138.

He K. et al, 2016. Measuring control plane latency in

sdn-enabled switches”, The 1st ACM SIGCOMM

Symposium on Software Defined Networking. pp. 1-6.

Internet of Things (IoT) 2013 to 2020 Market Analysis:

Billions of Things, Trillions of Dollars, IDC, 2016.

Available: www.marketsandmarkets.com/Internet-of/

Things-Market. Last accessed 20th mars 2017.

Javed U. et al, 2017. Stochastic Model for Transit Latency

in OpenFlow SDNs. Computer Networks Journal.

113(1), pp. 18–229.

Jeon S., Guimaraes C., and Aguiar R., 2013. SDN-based

Mobile Networking for Cellular Operators. 9th ACM

Workshop on Mobility in the Evolving Internet

Architecture (MobiArch’ 14). 13-18.

Jin X. et al., 2013. Softcell: Scalable and exible cellular core

network architecture. ACM conference on emerging

networking experiments and technologies (CoNEXT).

pp. 163–174.

Knaesel F., Neves P., and Sargento S., 2013. IEEE 802.21

MIH-enabled evolved packet system architecture.

International Conference on Mobile Networks and

Management. pp. 61-75.

Kreutz D., Ramos F., Verissimo P. and Uhlig S., 2015.

Software-Defined Networking: A Comprehensive

Survey. Proceedings of the IEEE. 13 (1), pp. 14–76.

Liebsch M.. et al, 2011. Per-Host Locators for Distributed

Mobility Management. IETF Internet draft.

Liu D. et al, 2014. Distributed Mobility Management:

Current practices and gap analysis. IETF Internet Draft.

Nguyen T., Bonnet C. and Harri J., 2016. SDN-Based

Distributed Mobility Management for 5G Networks.

SIMULTECH 2018 - 8th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

IEEE Wireless Communications and Networking

Conference. pp. 1-7.

Seite P. and Bertin P., 2013. Distributed Mobility

Anchoring. IETF Internet draft.

The OpenFlow Switch Specification, Version

K.15.05.5001, 2017. Available:) http://archive.open

flow.org. Last accessed March 2017.

Tomovic S., Yoshigoe K., Maljevic I., and Radusinovic I.,

2017. Software-Defined Fog Network Architecture for

IoT. International Journal of Wireless Personal

Communications, 2(1), pp. 181-196.

New Approach for Mobility Management in Openﬂow/Software-Deﬁned Networks