Handover Prediction Technique for Network Based Network Mobility (N-NEMO)

International Journal of Computer Networking,

Wireless and Mobile Communications (IJCNWMC)

ISSN 2250-1568

Vol. 3, Issue 4, Oct 2013, 111-126

© TJPRC Pvt. Ltd.

HANDOVER PREDICTION TECHNIQUE FOR NETWORK BASED NETWORK

MOBILITY (N-NEMO)

RAVI V ANGADI1 & K. C SHET

2

1Research Scholar, Department of Computer Science & Engineering, National Institute of Technology, Surathkal,

Karnataka, India

2Professor, Department of Computer Science & Engineering, National Institute of Technology, Surathkal, Karnataka, India

ABSTRACT

Scalability and facilitating continuous Internet access to the users are the main design goals of network mobility.

However, latency and packet losses are the resultant factors in many handoff techniques of network mobility. In order to

provide seamless handover, in this paper, we propose a handover prediction technique for Network Based Network

Mobility (N-NEMO). The proposed technique defines a functional management unit (FMU) to manage and predict

handovers in the network. When a MR moves towards the overlay region, it triggers overlay region scanning technique.

After predicting the handover, it forwards the handover request to FMU. Priority for each network is assigned considering

the parameters bandwidth, battery power, received signal strength and link quality of MRs. Through overlay region

scanning technique significant computation cost is minimized. The technique is simulated in NS-2 and result shows the

efficiency.

KEYWORDS: Network Mobility (NEMO), Mobile Access Gateway (MAG), Quality of Service (QoS)

INTRODUCTION

Network Mobility (NEMO)

The mobility of an entire network that alters the sole attachment point to the Internet and transmission of entire

data packets to and from the mobile network through one or more selected mobile routers (MR) is termed as network

mobility. [1] [2] NEMO basic support protocol (BSP) is an extension of mobile IPv6 for network mobility. The mobile

node connections are offered by MR which acts as a gateway for the mobile networks. The mobile network node (MNN)

can be either fixed or mobile. The categories of mobile nodes which are supported by the MR are a follows

Local Mobile Nodes (LMN): These nodes are mobile and correspond to the mobile network as its home network.

Local Fixed Nodes (LFN): These nodes are fixed and connectivity is offered by MR. LFN possess the similar

home agent as MR.

Visiting Mobile Nodes (VMN): These nodes connect with mobile network temporarily and do not belong to the

mobile network [3].

NEMO BSP protocol facilitates the entire network to wander around different access networks devoid of

disturbing current sessions of network nodes and exclusive of necessitating any particular mobility capability in the hosts.

[4]

NEMO minimizes the number of handovers of individual hosts and power consumption of mobile host (MH)

which is its main advantage. [5].

112 Ravi V Angadi & K. C Shet

Mobility Management

Mobility management has been a significant issue for maintaining seamless communication commencing from

cellular networks. The techniques included in mobility management are Location management and handoff management.

The tracking and updation of current location of mobile nodes (MN) are performed by location management technique.

The process of upholding the active link when the MN changes its attachment point is performed using handoff

management scheme. As MN’s alter their point of attachment regularly and the network topology gets altered

unexpectedly, mobility management is vital for offering high-speed and seamless service for vehicular networks.

The mobility management technique must assure the reachability to correspondent nodes (CN) in the internet in

addition to global reachability to mobile nodes. [6]

The process of transmitting the traffic of on-going session(s) from one access point to another is termed as

handoff. [1] A mobile network is linked with internet through one or more MRs. MNNs behind MR possess transparent

movement. The home link includes an entity called Home Agent (HA) through which the mobile router register its care of

address (CoA) and prefix. HA interrupts packets on the home link intended to the mobile network and performs

encapsulation and tunneling of packets to the MR’s registered CoA, though mobile network is far apart from home. MRs

obtains network layer access to the global internet from AR through which the transportation of the packets from or to the

internet occurs at the foreign link. [7]

NEMO’s handover procedure does not vary from MIPv6’s handover scheme. Hence FMIPv6 and HMIPv6

protocols can be applied for MIPv6 handover progress.

The two process involved in the FMIPv6 are as follows

Defining a New CoA: this is done to eliminate the latency owing to new prefix discovery following L2 handover

even when MN is present on the previous access routers’s (PAR) link.

Setting Up a Tunnel Between the Previous CoA and the New CoA: This is done to transmit the packet sent to

PAR to MN’s New AR (NAR) that further minimizes the packet loss.

HMIPv6 minimizes the handover latency by laying a hierarchical network structure and minimizing the signaling

overhead. However, HIMPv6 is more adaptable to micro-mobility, and its handover latency does not meet the needs of

performance-critical application [7].

Challenges in NEMO

In NEMO, the load balancing of multi-homing in nested mobile network is not considered widely. The handover

delay among MR causes issues in accessing failure recovery scheme. Also there is no real path evaluation metric

and algorithm available for selecting best path for load balance. [8]

As NEMO BSP depends on MIPv6, it takes over the drawbacks of MIPv6 that includes long signaling delay and

movement detection time. [9] Additionally some common issues such as sub-optimized routing and high overhead

for the packet transmission occurs in the nested NEMO. [10]

The QoS needs of vehicular applications create more challenges in mobility management design. The safety

application possesses higher priority when compared to non-safety applications and this priority should be assured

though the handoff is performed. The handoff latency needs to be minimized for multimedia applications. [6]

Handover Prediction Technique for Network Based Network Mobility (N-NEMO) 113

The process of using MIPv6 route optimization devoid of enhancement in NEMO generates problems.

o Since the CoA of an MNN in NEMO is linked with the home link of the MR, it does not offer the actual

geographic location of the MNN. Hence the MNN’s CoA cannot be utilized as a destination address in an IP

data packet for the sake of route optimization.

o When MR’s CoA is used by an MNN to perform RO, collapse of binding update occurs since entire MNN’s

requires updating their binding at HAs and CNs concurrently when an MR alters its CoA. [11]

When the entire network moves from one visited network to another, ITEF protocol is not appropriate. During this

case, Mobile IP protocol provides massive signaling. [7]

The packet loss and latency occurring in NEMO during handover, when the mobile network travels from previous

access router (AR) to its new AR, causes degradation in the on- going session performance. [7]

Problem Identification

Many researchers employed their potentials to consider mobility management and handover prediction techniques

only for Network Mobility (NEMO). However, the future technology in the Internet requires handover and mobility

management solutions for Network Based Network Mobility (N-NEMO). Further, scalability and facilitating continuous

Internet access to the users are the main design goals of network mobility. But, latency and packet losses are the resultant

factors in many handoff techniques of network mobility. In order to provide seamless handover and to manage mobility in

this paper we propose to employ handover prediction technique for Network Based Network Mobility (N-NEMO).

RELATED WORK

Yujia Zhai et al. [12] have modelled mobility support, which is based on distributed hash table (DHT). Their

model is structured in a way to lessen the effects of inconsistency on mobility management. They have designed the

mapping query through overlay network to a finite-horizon Markov decision process (MDP). They have also described a

reward function, which incorporates both physical layer information and application layer information. In this view of the

fact, they have introduced a Markov decision routing algorithm called MDR. Through this algorithm, they have enhanced

backward induction to obtain the global optimal strategy. Their algorithm has the characteristic of distributed computing.

Further, they have proposed a DHT and MDR-based mobility management scheme called DMDRMM.

Zhiwei Yan et al. [13] have proposed a novel network mobility supporting scheme (N-NEMO) in Proxy Mobile

IPv6 (PMIPv6) network. They have used a tunnel splitting scheme to differentiate the inter-Mobility Access Gateway

(MAG) and intra-MAG mobility. Moreover, the recursive binding state lookups in MAG is utilized with the help of the

extended “Dist” option in the LPBU (Localized Proxy Binding Update) registration message, which is a new extension of

the PBU message. Through their approach packet can be transmitted through the more optimized route in the intra-MAG

communication and the cost of encapsulation can be decreased significantly.

Fabio Giust et al. [14] have presented the mobility architecture for MultimEDia transport for mobIlE Video

AppLications (MEDIEVAL) project. Some features of their mobility architecture are, (i) It supports a Distributed Mobility

Management (DMM) approach. That is mobility is fastened at the very edge of the network. (ii) Their architecture has

adopted a hybrid approach, such that it can operate at both network mobility management solution and client based

solutions. Finally, their technique supports multicast traffic delivery and content distribution, since the video-centric nature

of the project depends on it.


Huachun Zhou et al. [15] have detailed a global mobility management architecture and protocol procedure called

GPMIP. Their proposed architecture was based on Proxy Mobile IPv6. Instead of performing mobility management

through individual nodes, their GPMIP has utilized the network entity. One distinct characteristic of GPMIP is the

segregation of network address and identity of the mobile node. The mobile nodes are differentiated by a Network Access

Identifier (NAI), which has an associated set of information stored on the network, including a profile containing the home

prefix. The Mobile Access Gateway (MAG), located in the access router, retrieves the MN profile information from AAA

server and sends the customized Router Advertisements (RAs) to the MN, emulating the home network behavior.

Romain Kuntz et al. [16] have proposed an approach for discovering default router. Their approach permits the

cooperation of multiple mobile routers to improve the bandwidth, network coverage and reliability of a mobile network.

Further, a dynamic load sharing mechanism between all available mobile routers is supported by their approach. They have

focused on multiple MRs located in the same mobile subnet. These multiple MRs are associated to one or multiple home

agents and advertise the same mobile network prefix in the mobile subnet.

PROPOSED SOLUTION

Overview

In this paper, we propose a handover prediction technique for Network Based Network Mobility (N-NEMO). The

technique describes a new entity called Functional Management Unit (FMU) in N-NEMO architecture. FMU is defined to

predict and manage handovers in the network. It categorizes the mobile routers (MR’s) into three states as H_ON,

Swap_ON and Swap_OFF_HO. By default, all MR’s are assigned to be in Swap_OFF_HO state. When a MR moves

towards the overlay region, it triggers overlay region scanning technique. It predicts the handover and forwards handover

request to FMU. Now, the state of a MR is changed to H_ON and an another MR to selected to transmits the traffic of MR

that invoked handoff process. An optimal MR is selected considering criteria’s such as bandwidth, battery power, received

signal strength and link quality. The selected MR will be in Swap_ON state until the handover mechanism is completed.

Thus, the technique provides seamless data transmission during handovers.

Estimation of QoS Metrics

Estimation of Available Bandwidth (AB)

The available bandwidth (AB) at a link is defined as its unexploited capacity. At any time t, a link can be either

idle or transmitted packets at the maximum speed. Hence the definition of AB takes average unexploited bandwidth over

time interval T into consideration which is given by following Eq. (1)

ABi (t, T) =

tT

t

ii dttLCT

))((1

(1)

Where ABi (t, T) = Available bandwidth at link i at time t,

LCi = Link’s capacity

i = Traffic.

The available bandwidth along a path is the minimum available bandwidth of all traversed links. [17]

Estimation of Consumed Battery Power (E)

The power consumption in WiMAX networks are as follows. [18]

E = n* Pwx * L (1+ ) / Rwr (2)


where E = power degradation or consumed power

n= number of packets.

Pwx = transmit power

Rwr = received power

L = Length of the packet

= Rw / Pwx (3)

Estimation of Received Signal Strength (RSS)

The channel condition can be estimated based on the received signal strength (RSS) and signal to noise ratio

(SNR) at the receiver. RSS is estimated using Friis equation which is given using Eq: (4) [19]

RSS =

*)**4(

*****2

2

d

hhP rxtxtx (4)

Where,

Ptx = Transmission power

= Transmitter gain

= Receiver gain

htx = Transmitter height

hrx = Receiver height

= Wavelength

d = Distance between the transmitter and receiver.

= System loss

SNR is computed from the received signal power (RSS) using Eq: (5)

SNR = log10 (Ptx) – log10 (Prx) dB (5)

Where Prx = Receiver Power

Link Quality Estimation (EETTi)

The links holding highest link quality are permitted to transmit more packets when the poor quality links are still

waiting to link improvement. This perhaps enhances the link quality. If the link performs normally then the poor quality

link could try to communicate. The Exclusive expected transmission time (EETT) metric is utilized to estimate the link

quality. It offers a better evaluation of a multi-channel path.

Consider N-hop path with M channels.

For a given link z, its interference set ( ) is defined as the set of links that has interference with it. A link’s

interference set also includes the link itself. Then the link z’s EETT is defined as:


EETTi = )(zlinki

iETT

(6)

where (z) is the interference set of link z.

Network Architecture

Our previous paper [20] describes the architecture of N-NEMO comprehensively. The core functional entities in

the N-NEMO are the local mobility anchor (LMA) and the mobile access gateway (MAG).

LMA: It is responsible for maintaining the MN’s reachability state and is the topological anchor point for the

MN’s Home Network Prefix (HNP).

MAG: It is the entity that performs the mobility management on behalf of the MN and it resides on the access link

where the MN is anchored. The MAG is responsible for detecting the MN’s movement to and from the access link and for

initiating binding registrations to the MN’s LMA.

There exist two tunnels in the architecture.

Global Tunnel: This is established between local mobility anchor (LMA) and mobile access gateway (MAG).

Local Tunnel: This is established between mobile router (MR) and MAG.

The MR includes mobile network nodes which are categorized into three groups such as local fixed nodes (LFNs),

local mobile nodes (LMNs), and visiting mobile nodes (VMNs).

To enhance the performance of network and as an extension to our previous paper, this paper introduces an

additional unit in our existing architecture termed as a Functional Management Unit (FMU). The sketch of an enhanced

architectural diagram is given in figure-1.

Figure 1: Architecture for N-NEMO

Functional Management Unit (FMU)

An objective functionality of FMU is to predict handovers in the network considering constraints such as

available bandwidth, consumed battery power, received signal strength (RSS), and link quality estimation. FMU includes a


set of MAG’s and on the other hand MAG’s take in the subset of access points (AP’s). To keep track of entire network,

FMU comprises of two tables namely K_MAG (Knowledge of MAG) table and K_MR (Knowledge of mobile router)

table. The former K_MAG comprises the information about MAG like MAG’s address, bandwidth, battery power, RSS,

identity of AP’s, boundary overlapping area and link quality. The information included in this table is periodically

transmitted to FMU by the corresponding MAG. The table format is shown below in table-1,

Table 1: K_Mag Table

MAG’s

address Bandwidth

Battery

Power RSS

ID of

Access

Point

Boundary

Overlapping

Area

Link

Quality

Each MR is well-defined to be in any of the following three states as H_ON, Swap_ON and Swap_OFF_HO.

H_ON: This state specifies that the corresponding MR carries out handover.

Swap_ON: MR in this state swaps and forwards other MR’s traffic. Here MR can permit intra domain handoff

rather than inter domain handoff to ensure that it keeps the current state until the other MR finishes its handoff.

Swap OFF_HO: In this state MR may perform inter domain handoff and it can not transfer the traffic of any other

MR.

K_MR table incorporates the home address, current state, address of currently connected MAG, and ST_info.

Here, ST_info represents another MAG address whose traffic is carried by the current MAG. The format of K_MR table is

shown in table-2.

Table 2: K_MR Table

Home

Address Current State

Address of currently

connected MAG ST_info

Overlay Region

Consider C1 and C2 as the coverage areas of MAG1 and MAG2 respectively. The intersection region of C1 and C2

is termed as overlay region (OR). This overlay region may include one or more AP’s that belongs to both MAG1 and

MAG2. In this technique, OR region is employed to predict the mobility of MR’s.

Mobility Prediction Technique

Figure 2: Coverage Area of Two MAG’s


Consider the sketch given in figure-2 to understand the mobility prediction algorithm. MAG 1 includes a set of

AP’s namely AP 1, AP 2, AP 3 and AP 4. On the other hand, MAG 2 comprises of three AP’s such as AP 5, AP 6 and AP

7. MR 1 and MR 2 are two mobile routers that move with the speed v.

Initially, all MR’s will be in the state Swap_OFF_HO. According to figure-2, when MR 1 moves towards overlay

region it predicts handoff by triggering overlay region scanning technique. This technique scans for the AP’s that are only

in overlay region. Further, this scanning approach limits significant computation cost that occurs due to frequent scanning

of the entire network.

Once the handover is predicted by MR 1, it forwards HOR (HandOverRequest) message to FMU. By receiving

HOR, the FMU changes the state of MR 1 from Swap_OFF_HO to H_ON. The message HOR contains the MAC address

of AP’s that are in overlay region.

Now, FMU has to select another MR to transmit the traffic of MR 1 until it finishes its handoff. It selects the MR

considering the parameters such as bandwidth, battery power, received signal strength and link quality. These values are

stored in K_MAG and K_MR Tables. According to the priority, FMU chooses an ideal MR for transmitting the traffic of

MR 1. The selection process is as follows,

Algorithm-1

Let Li represents the link

Let APi represent the existing mobile access gateway. (i=1, 2, 3,…n)

If (RSS (Li) =high) && (EETT (APi) = high)

Then

APi is added in the Priority list (PL) which is built based on respective user’s QoS preferences

Else

Particular APi is eliminated

End if

Estimate AB and E related with each Li with respect to each of the available APi in PL.

The priority list is then sorted in descending order of combined values of AB and E such that the best suitable

AP appears first in the list.

Then, the sorted list is sent to the core network along with Handover request (HOR).

MN )(HORPL Core network

If (AB andE) < Th

Then

MN finds the combination of connections of APi with max (AB andE).

MN eliminates the APi with min (EETT)


MN sorts the remaining APi as per user preference

Repeat step 4.

End if

In figure-2, the FMU selects MR 2 as the best MR for transmitting traffic of MR 1. After choosing an appropriate

MR, FMW sends back two ACK messages. ACK 1 message is forwarded to MR 1. It is a handoff response message and it

contains the MAC address and home address of selected MR.

ACK 2 is transmitted to the currently connected MAG. It is a handover packet transfer message. By receiving this

message, the corresponding MAG generates a tunnel between itself and selected MR to transmit traffic addressed to MR 1.

Once handoff is completed by MR 1, it transmits proxy binding update message to local proxy and handover

complete message to FMU. The handover complete message contains modified addresses of MAG and AP. Finally, the

MR 1 returns to Swap_OFF_HO.

SIMULATION RESULTS

We have used the Network Simulator (NS2) to implement the proposed technique and obtained the results to

prove the proposed technique is efficient in terms of delay, bandwidth and drop.

Validation of results done by comparing the results of our proposed approach with an existing approach by

simulating the concept in NS2. By varying different parameters such as the speed of the mobile node, power level and

bandwidth in various scenarios of NEMO, we have conducted various experiments. The results were obtained by mining

the simulation trace files using AWK or PERL scripts and graph were plotted using the xgraph utility of NS2.

The NS2 version is 2.28 with Mobiwan patch included. The simulation topology is given in Figure 3. It consists

of 4 mobile routers MR1, MR2, MR3 and MR4 which are connected to the access routers AR1,AR2,AR3 and AR4,

respectively. MR1 and MR2 are connected to MAG1 whereas MR3 and MR4 are connected to MAG2.

We compared the proposed QOSMP-N-NEMO scheme with the existing N-NEMO scheme.

In the first handoff scenario, VMN1 from MR1 is handoff to MR2 (intra-domain handoff-refer figure 3), both are

located in the same MAG1.In the second handoff scenario, VMN1 from MR2 is handoff to MR3 which is at MAG2

(inter-domain handoff-Refer Figure 4.)

Figure 3: Intra-Domain Handoff Topology


Figure 4: Inter-Domain Handoff Topology

We measure the received bandwidth of mobile node at various time intervals using UDP data flows.

Based on Rate

In our first experiment we vary the load value as 500, 750, 1000 and 1250 Kb.

Rate Vs Bandwidth(MR-1)

0

2

4

6

8

10

500 750 1000 1250

Rate(Kb)

Mb

/s N-NEMO

QOSMP

Figure 5: Rate vs Bandwidth

Rate Vs Bandwidth(MR-2)

0

2

4

6

8

500 750 1000 1250

Rate(Kb)

Mb

/s N-NEMO

QOSMP

Figure 6: Rate vs Bandwidth

Rate Vs Delay(MR-1)

0

50

100

150

500 750 1000 1250

Rate(Kb)

Dela

y(S

ec)

N-NEMO

QOSMP

Figure 7: Rate vs Delay


Rate Vs Delay(MR-2)

0

50

100

150

500 750 1000 1250

Rate(Kb)D

ela

y(S

ec)

N-NEMO

QOSMP

Figure 8: Rate vs Delay

Rate Vs Drop(MR-1)

0

1000

2000

3000

500 750 1000 1250

Rate(kb)

Pkts

N-NEMO

QOSMP

Figure 9: Rate vs Drop

Rate Vs Drop(MR-2)

0

500

1000

1500

500 750 1000 1250

Rate(Kb)

Pkts

N-NEMO

QOSMP

Figure 10: Rate vs Drop

From figure 5 and 6, we can see that the bandwidth of our proposed QOSMP is higher than the existing N-NEMO

technique.

From figure 7 and 8, we can see that the delay of our proposed QOSMP is less than the existing N-NEMO

technique.

From figure 9 and 10, we can see that the packet drop of our proposed QOSMP is less than the existing N-NEMO

technique.

Based on Speed

In our second experiment we vary the mobile speed as 10,20,30,40 and 50m/s.


Speed Vs Bandwidth(MR-1)

0

2

4

6

8

10 20 30 40 50

Speed(m/s)M

b/s N-NEMO

QOSMP

Figure 11: Speed vs Bandwidth

Speed Vs Bandwidth(MR-2)

0

2

4

6

10 20 30 40 50

Speed(m/s)

Mb

/s N-NEMO

QOSMP

Figure 12: Speed vs Bandwidth

Speed Vs Delay(MR-1)

0

50

100

150

10 20 30 40 50

Speed(m/s)

Dela

y(S

ec)

N-NEMO

QOSMP

Figure 13: Speed vs Delay

Speed Vs Delay(MR-2)

0

50

100

150

10 20 30 40 50

Speed(m/s)

Dela

y(S

ec)

N-NEMO

QOSMP

Figure 14: Speed vs Delay


Speed Vs Drop(MR-1)

0

500

1000

1500

2000

2500

10 20 30 40 50

Speed(m/s)P

kts

N-NEMO

QOSMP

Figure 15: Speed vs Drop

Speed Vs Drop(MR-2)

0

5000

10000

15000

10 20 30 40 50

Speed(m/s)

Pkts

N-NEMO

QOSMP

Figure 16: Speed vs Drop

From figure 11 and 12, we can see that the bandwidth of our proposed QOSMP is higher than the existing N-

NEMO technique.

From figure 13 and 14, we can see that the delay of our proposed QOSMP is less than the existing N-NEMO

technique.

From figure 15 and 16, we can see that the packet drop of our proposed QOSMP is less than the existing N-

NEMO technique.

Based on Time

In our third experiment we analysis the metrics for the different set of time intervals.

Time Vs Bandwidth(MR-1)

0

0.1

0.2

0.3

0.4

0.5

1 7 13 19 25 31

Time(Sec)

Mb

/s N-NEMO

QOSMP

Figure 17: Time vs Bandwidth


Time Vs Bandwidth(MR-2)

0

0.1

0.2

0.3

0.4

0.5

35 41 47 53 59 65

Time(Sec)M

b/s N-NEMO

QOSMP

Figure 18: Time vs Bandwidth

Time Vs Delay(MR-1)

0

10

20

30

40

1 7 13 19 25 31

Time(Sec)

Dela

y(S

ec)

N-NEMO

QOSMP

Figure 19: Time vs Delay

Time Vs Delay(MR-2)

0

50

100

150

200

35 41 47 53 59 65

Time(Sec)

Dela

y(S

ec)

N-NEMO

QOSMP

Figure 20: Time vs Delay

Time Vs Drop(MR-1)

0

20

40

60

80

100

1 7 13 19 25 31

Time(Sec)

Pkts

N-NEMO

QOSMP

Figure 21: Time vs Drop


Time Vs Drop(MR-2)

0

1000

2000

3000

35 41 47 53 59 65

Time(Sec)

Pkts

N-NEMO

QOSMP

Figure 22: Time vs Drop

From figure 17 and 18, we can see that the bandwidth of our proposed QOSMP is higher than the existing N-

NEMO technique. From figure 19 and 20, we can see that the delay of our proposed QOSMP is less than the existing N-

NEMO technique. From figure 21 and 22, we can see that the packet drop of our proposed QOSMP is less than the existing

N-NEMO technique.

CONCLUSIONS

In this paper, we have proposed a handover prediction technique for Network Based Network Mobility (N-

NEMO). The proposed technique defines a functional management unit (FMU) to manage and predict handovers in the

network. The technique categorizes the mobile routers (MRs) into three states as H_ON, Swap_ON and Swap_OFF_HO.

When a MR moves towards the overlay region, it triggers overlay region scanning technique. It predicts the handover and

forwards handover request to FMU. Priority for each network is assigned considering bandwidth, battery power, received

signal strength and link quality of MRs. Through overlay region scanning technique significant computation cost is

minimized. The technique is simulated in NS-2 and result shows the efficiency.

REFERENCES

1. Cheng-Wei Lee, Meng Chang Chen, Yeali S. Sun, “A Network Mobility Management Scheme for Fast QoS

Handover”, Proceedings of the 4th Annual International Conference on Wireless Internet, WICON '08, 2008

2. Rituparna Chaki, Nandini Guha Niyogi, “An Improved Scheme towards Fast Handover In a Hierarchial Mobile

Network”, IEEE General Assembly and Scientific Symposium, 2011

3. P. Calduwel Newton, L. Arockiam, “Route Optimization Mechanisms for Internet Applications in Mobile

Networks: A Survey”, International Journal of Future Generation Communication and Networking, Vol. 3, No. 3,

Sept. 2010.

4. María Calderón, Carlos J. Bernardos, Marcelo Bagnulo, Ignacio Soto, and Antonio de la Oliva, “Design and

Experimental Evaluation of a Route Optimization Solution for NEMO”, IEEE Journal On Selected Areas In

Communications, Vol. 24, No. 9, Sept 2006.

5. Cheng-Wei Lee, Yeali S. Sun and Meng Chang Chen, “HiMIP-NEMO: Combining Cross-layer Network Mobility

Management and Resource Allocation for Fast QoS-Handovers”, IEEE Vehicular Technology Conference, 2008.

6. Kun Zhu, Dusit Niyato, Ping Wang, Ekram Hossain, and Dong In Kim, “Mobility and Handoff Management in

Vehicular Networks: A Survey”, Wireless communications and mobile computing, 2009.


7. Hai Lin and Houda Labiod, “Hybrid Handover Optimization for Multiple Mobile Routers-based Multihomed

NEMO Networks”, IEEE International Conference on Pervasive Services, pp-136-144, 2007

8. Huang Songhua, Wu Xiaobing, Lu Yin, Huang Hao and Chen Guihai, “SMART: A Strengthened Model of

Access Router Tree for Multihoming in Nested Mobile Network", Chinese Journal of Electronics, Vol.18, No.1,

Jan. 2009.

9. Hyo-Beom Lee, Youn-Hee Han and Sung-Gi Min, “Network Mobility Support Scheme on PMIPv6 Networks”,

International Journal of Computer Networks & Communications (IJCNC) Vol.2, No.5, Sept. 2010.

10. Zhiwei Yan, Huachun Zhou and Ilsun You, “N-NEMO: A Comprehensive Network Mobility Solution in Proxy

Mobile IPv6 Network”, Journal of Wireless Mobile Networks, Ubiquitous Computing and Applications, pp-52-

70, 2010

11. Ved P. KAFLE, Eiji KAMIOKA, and Shigeki YAMADA, “MoRaRo: Mobile Router-Assisted Route

Optimization for Network Mobility (NEMO) Support”, The Institute of Electronics, Information and

Communication Engineers, pp.158-170, 2006.

12. Yujia Zhai, Yue Wang, Jian Yuan, Yong Ren, Xiuming Shan and Ilsun You, “A DHT and MDP-based Mobility

Management Scheme for Large-Scale Mobile Internet”, IEEE Conference on Computer Communications

Workshops (INFOCOM WKSHPS), pp- 379- 384, 2011

13. Zhiwei Yan, Huachun Zhou, Hongke Zhang, Sidong Zhang and Ilsun You, “Network Mobility Support in

PMIPv6 Network”, Proceedings of the 6th International Wireless Communications and Mobile Computing

Conference, (IWCMC '10), pp- 890-894, 2010

14. Fabio Giust, Carlos J. Bernardos, S´ergio Figueiredo, Pedro Neves and Telemaco Melia, “A Hybrid MIPv6 and

PMIPv6 Distributed Mobility Management: the MEDIEVAL approach”, IEEE Symposium on Computers and

Communications (ISCC), pp- 25-30, 2011

15. Huachun Zhou, Hongke Zhang, Yajuan Qin, Hwang-Cheng Wang and Han-Chieh Chao, “A Proxy Mobile IPv6

Based Global Mobility Management Architecture and Protocol”, Journal of Mobile Networks and Applications,

Volume 15 Issue 4, pp-530 – 542, August 2010

16. Romain Kuntz, Julien Montavont and Thomas Noel, “Multiple Mobile Routers in NEMO: How Neighbor

Discovery Can Assist Default Router Selection”, IEEE 19th International Symposium on Personal, Indoor and

Mobile Radio Communications, (PIMRC), pp-1-6, 2008.

17. Kiran Ahuja, Rajesh Khanna, and Brahmjit Singh, “ Available bandwidth estimation: A criterion for network

selection in 3G and WLAN heterogeneous environment”, International Journal of Research and Reviews in

Computer Science (IJRRCS) Vol. 2, No. 3, June 2011.

18. Malak Zareif Habeib, Hussein Abd Elaty Elsayed, Salwa Hussein Elramly and Magdy Mahmoud Ibrahim,

“Battery Based Vertical Handover between WiMAX and WLAN Technologies”, Proceeding IEEE, 2011.

19. S. J. K. Jagadeesh Kumar and T. Purusothaman, “ An Enhanced Scheduling Scheme for QoS Guarantee using

Channel State Information in WiMAX Networks”, European Journal of Scientific Research, Vol.64, No.2, pp.

285-292, 2011

20. Ravi Angadi, Dr K.C Shet “QoS Based Handover Technique for Network Based Network Mobility (N-NEMO)”.

Handover Prediction Technique for Network Based Network Mobility (N-NEMO)

Documents

Transcript of Handover Prediction Technique for Network Based Network Mobility (N-NEMO)