Dynamic Map and Diffserv Based AR Selection for Handoff in HMIPv6 Networks

International Journal of Research in Computer Science
eISSN 2249-8265 Volume 3 Issue 1 (2013) pp. 35- 45
www.ijorcs.org, A Unit of White Globe Publications
doi: 10.7815/ijorcs. 31.2013.059

DYNAMIC MAP AND DIFFSERV BASED AR
SELECTION FOR HANDOFF IN HMIPV6 NETWORKS
1
Shiva Prasad Kaleru, 2Damodaram Avula
1
Juniper Networks, Bangalore, INDIA
shivaprasadkaleruphd@gmail.com, shivakaleru@gmail.com
2
Professor of CSE & Director, Academic Audit Cell, JNTU, Hyderabad, INDIA
damodarama@rediffmail.com

Abstract: In HMIPv6 Networks, most of the existing outside the MAP domain in case of handoffs within the
handoff decision mechanisms deal mainly with the same domain and may improve handoff performance
selection of Mobility Anchor Point (MAP), ignoring reducing handoff latency and thus packet losses since
the selection of access router (AR) under each MAP. intra domain handcuffs are performed locally. This is
In this paper, we propose a new mechanism called the main advantage of HMIPV6.
“Dynamic MAP and Diffserv based ARs selection for
The MAP basically acts as a local Home Agent
Handoff in HMIPv6 networks” and it deals with
(HA) and, as such, it receives all packets on behalf of
selecting the MAP as well as ARs. MAP will be
the MNs it is serving. That is the MAP decapsulates
selected dynamically by checking load, session
and forwards the received packets to the MN’s current
mobility ratio (SMR), Binding update cost and
address. In HMIPv6 networks, an MN configures two
Location Rate. After selecting the best MAP, the
cars: a regional care-of-address (RCoA) and an on-line
Diffserv approach is used to select the AR under the
care-of-address (LCoA). The RCoA is an address in
MAP, based on its resource availability. The AR is
the MAP's subnet. When a mobile node enters into a
implemented at the edge router of Diffserv. DiffServ
new MAP domain, it registers with itself by obtaining
can be used to provide low-latency to critical network
a Regional Care-of-Address (RCoA) [3] [4]. The
traffic such as voice or streaming media while
RCoA is the address that the mobile node will use to
providing simple best-effort service to non-critical
inform its Home Agent and corresponding nodes about
services such as web traffic or file transfers. By using
its current location. Then, the packets would be sent to
this mechanism, we can assure that better resource
and intercepted by the MAP, acting as a proxy, and
utilization and throughput can be attained during
routed inside the domain to the on-link care-of-address
Handoff in HMIPv6 networks.
(LCoA) by the MAP. On the other hand, the LCoA is
Keywords: HMIPV6 Networks, Access Router (AR), an on-link CoA attributed to the MN's interface and it
Session Mobility Ratio (SMR). is based on the prefix information advertised by an
AR. After configuring the LCoA and RCoA, the MN
I. INTRODUCTION sends a BU message to the MAP, which then
maintains the binding information between the RCoA
To mitigate the high signaling overhead occurring and the LCoA (i.e., Local binding update).
in Mobile IPv6 networks when mobile nodes (MNs)
perform frequent handoffs, a Hierarchical Mobile IPv6 When the MN changes its current address within a
(HMIPv6) was proposed by Internet Engineering Task MAP domain, then it needs to register the new address
Force (IETF) [1] [2]. The Mobility Anchor Point (i.e., New LCoA) with the MAP. When it moves to
(MAP) was introduced in HMIPV6 to reduce the another MAP domain then only its RCoA will change
considerable number of the binding update (BU) and it does not change as long as the MN moves within
messages between the mobile node (MN), the the same MAP domain. This makes the MN’s mobility
correspondent node (CN), and the home agent (HA). transparent for the correspondent nodes (CNs). After
MAP is to handle the binding update (BU) procedures configuring the LCoA and RCoA, the MN sends a BU
due to handoffs within a MAP domain in a localized message to the MAP, which then maintains the binding
manner, which reduces the amount of network- wide information between the RCoA and the LCoA. Also,
signaling traffic for mobility. the MN sends a BU message containing the MN's
home address (HoA) and the RCoA to it’s HA and
Hierarchical Mobile IPv6 (HMIPv6) is an extension CNs. The MN's RCoA is not changed while the MN
of MIPv6 and it has been proposed to reduce the resides in the MAP domain and therefore the MN
signaling load and to improve the handover speed for
mobile connections. It reduces the signaling load

www.ijorcs.org

36 Shiva Prasad Kaleru, Damodaram Avula

need to send a local BU message only to the MAP (not BS: Base Station
to its HA) for a movement within the MAP domain. As CN: Corresponding Node
a result, HMIPv6 only reduces times needed to binding HA: Home Agent
update in the handover procedures. MS: Mobile Station
MAP: Mobility Anchor Point
In HMIPv6 networks, one or more MAPs can be
AR: Access Router
located within the same network hierarchy and a MAP
can exist at any level in the pecking order, including at Figure 1 illustrates the basic operations that are
the level of the AR, operate independently of one performed in HMIPv6 networks. The HMIPv6 is
another. Especially, when HMIPv6 is deployed in composed of a HA, a CN, Five MAPs, Seven ARs and
large-scale wireless/ mobile networks, multiple MAPs a MN. There are three Diffserv domains: domain A,
are deployed to provide scalable mobile services. It is domain B and domain C. Each domain has its own
very important for an MN to select the most suitable unique Diffserv domain number which is used to
MAP among the available MAPs, in order to reduce inform the MN which Diffserv domain the MN
the total cost. Here, the total cost, incurred by an MN belongs to. If the MN has received the information and
in a HMIPv6 network, consists of the binding update found that the Diffserv domain number was changed
cost and the packet deliver cost. from the present, it means that the MN has moved to
the boundary of new Diffserv domain and is required
An MN needs to consider several factors when
to configure the Diffserv configuration parameters
selecting the optimal MAP that minimizes the total
within itself in order to use the new Diffserv domain.
cost among the various MAPs available in a foreign
network. The advantage of having an appropriate MAP In this paper we propose a new mechanism called
selection that covers most of the MN’s mobility area “Dynamic MAP and Diffserv based ARs selection for
is, we can significantly reduce the binding updates Handoff in HMIPv6 networks” that deals with
(global binding update and local binding update) to the selecting the MAP as well as ARs. In Dynamic MAP
HA and further reduce the signaling cost and location selection we are considering the load, Session Mobility
update cost in HMIP. The global binding update is a Ration, Binding update cost and Location Rate, based
procedure in which MN registers its RCoA with the on these two we will select an optimal MAP. After
CNs and HA. On the other hand, a local binding selecting the best MAP, the Diffserv approach is used
update occurs when MN changes its current address to select the AR under the MAP, based on its resource
within a local MAP domain, it only needs to register availability. The AR is implemented as the edge router
the new address with the MAP. If a mobile node then of Diffserv. By using this mechanism, we can assure
performs a handoff between two access points within that better resource utilization and throughput can be
the same MAP domain only the MAP has to be attained during Handoff in HMIPv6 networks.
informed. Note, this does not imply any change to the
periodic BUs, a MN has to send to the HA, CNs and II. RELATED WORK
now additionally to the MAP.
Dong-Guen Kim et al, in [6] have proposed a
Hierarchical mobile IPv6 (HMIPv6) protocol is network-based handover approach for HMIPv6
proposed by employing a hierarchical network networks. If handover in HMIPv6 occurs while
structure to reduce handoff latency. HMIPv6 protocol handover process in IEEE 802.16e is in progress, then
suffers the long handoff delay and the high packet lost the total handover latency will be significantly
in the macro mobility. [1][2][6] reduced. This scheme uses the network-based
handover approach in which access routers (AR)
generate local binding update on behalf of MN. This
allows L3 handover and L2 handover to happen at the
same time. This scheme also minimizes packet losses.
By introducing additional management messages, AR
can buffer the packets destined to MN until total
handover process is finished. This technique does not
deal with the handover performance of macro mobility
in HMIPv6 networks.
Sangheon Pack et al., in [7] have proposed an
adaptive MAP selection scheme for HMIPv6
networks. In the adaptive MAP selection scheme, an
MN first estimates its session-to-mobility ratio (SMR).
Then, based on its SMR, the MN chooses a MAP that
minimizes the total cost, consisting of the binding
Figure 1: System Architecture

www.ijorcs.org

Dynamic Map and Diffserv Based AR Selection for Handoff in HMIPv6 Networks 37

update cost and packet delivery cost. In addition, the From the Previous works, we can say that the
MN calculates two threshold SMR values, which handoff decision mechanisms for HMIPv6 deal mainly
adaptively trigger a new MAP selection procedure. If with selection of MAP, ignoring the selection of access
the estimated SMR is larger (or smaller) than the upper routers (AR) under each MAP. Our mechanism
(or lower) threshold SMR value, the MN recalculates “Dynamic MAP and Diffserv based ARs selection for
the total cost and re-selects a MAP that minimizes the Handoff in HMIPv6 networks” deals with selecting the
total cost. The adaptive MAP selection scheme results MAP as well as ARs. MAP will be selected
in a degree of computation overhead at the side of dynamically by checking load and session mobility
MNs, since MNs should monitor the session arrival ratio (SMR) [6]. After selecting the best MAP, the
and mobility rates. Diffserv approach is used to select the AR under the
MAP, based on its resource availability. The AR is
Ying Hong Wang et al., in [8] have proposed a
implemented as the edge router of Diffserv.
dynamic MAP selection mechanism for the
Hierarchical Mobile IPv6 networks. According to the
HMIPv6 mechanism, the MAP of higher layer can III. DYNAMIC MAP AND DIFFSERV BASED AR
efficiently reduce the frequency of performing binding SELECTION
update; the higher loading of service is the bottleneck A. Overview of the Proposed Work
of the whole network. Because the bandwidth of the
MAP which can serve is finite, the whole network will Traffic Control and management are essential for
be crashed due to the overloading if the MAP serves as every mobile application due to the limited resources
the gateway at the same time. So the authors have of the mobile network. The main goal of our work is to
proposed a MAP selection mechanism that takes the enable the MN to choose the best MAP and the MAP
mobile node’s particular characteristics which include will choose the best AR before performing handover
the mobility velocity and quantity of communication and to provide quality of service for each application
services into consideration, the proposal can manage type of MN.
the MAPs efficiently. They also have designed a MAP We propose a dynamic MAP selection mechanism
load balancing mechanism to avoid the network crash based on the checking load, session mobility ratio
due to the overloaded MAP. This technique does not (SMR), Binding update cost and Location Rate of MN
predict the movement direction of the MN. that reduces the handoff latency and hence improving
WonSik Chung et al., in [9] have introduced a two location update and packet delivery. The main goals of
novel dynamic MAP selection schemes(LV-MAP and our framework are to enable MN to be able to choose
DV-MAP) for HMIPv6, that relieve overloaded MAPs the best AR before performing the handover and
as well as select a more suitable MAP according to quality of service of each application type on MN. An
each Mobile Node (MN)’s up-to date mobility towards MN determines its serving MAP based on the
reducing inter-domain handover, resulting in saving estimated session –to-mobility. By considering the
the overall signaling cost. LV-MAP scheme distributes SMR in the selection of MAP, the MN is able to select
load over multiple MAPs for an overloaded HMIPv6 a more appropriate MAP with respect to its own
network while DVMAP selects the furthest MAP mobility and session activity.
supporting MN’s velocity for less overloaded situation, After selecting the best MAP, we are using the
with the aim to reduce the frequency of inter-domain Differentiated Services (Diffserv) approach to select
handovers and distribute load over the MAPs the AR under the MAP, based on its resource
Yuh-Shyan Chen et al., in [10] presents a new availability. The AR is implemented at the edge router
cross-layer partner-assisted handoff mechanism based of Diffserv. We are adding two more parameters to the
on HMIPv6, termed as P_HMIPv6 protocol. The Diffserv to select the optimal AR and those two
P_HMIPv6 protocol is a cross-layer approach by the parameters are Signal Strength and Moving direction.
combination of layer 2 and layer 3. The partner station When the MN disconnect from old Access Router
(PS) is a new component with relay ability and (oAR), all packets will buffer at new Access router.
adopted by our protocol. With the assistance of the PS, After the MN connects to the new Access Router
care-of address (CoA) is pre-acquired and DAD (nAR), the nAR then forwards any buffered packets to
operation is pre-executed by the PS before the MS the MN. By using the Diffserv we will provide QoS
initiates the layer 2 handoff. The simulation results guarantees for mobile host in nAR before mobile host
show that P_HMIPv6 protocol actually achieves the handovers from the oAR, and enable MN choosing a
performance improvements in handoff delay time, nAR with the best match to the QoS request.
packet loss rate, and handoff delay jitter. This
B. Dynamic MAP Selection Mechanism
technique does not consider the security issue of cross-
layer partner-based handoff scheme for WiMax In the section, we propose a dynamic MAP
networks. selection mechanism that takes an MN’s load, Session

www.ijorcs.org


Mobility Ratio (SMR), Binding update cost and ESMR (i+1) = α . ESMR (I) + (1- α) . CSMR (2)
location rate. The dynamic selection mechanism
consists of four procedures. One is the Checking Load,
Where ESMR is the estimated SMR and CSMR is the
Second is SMR, third is Binding update cost and fourth
current measured SMR and we will calculate the SMR
is Location rate. Detailed descriptions for each
according to the time interval. For each MI we
procedure will be elaborated in the following section.
calculate the SMR and we compare with the threshold
1. Checking Load: Initially, in the Multiple MAP values. Here α is a weighting parameter, where 0 < α
environments, a MN collects all RA messages sent < 1.
from the available MAPs in the foreign network.
3. Calculation of Binding Update Cost (BUC):
The MN can obtain information for each MAP
Binding update cost [7] is the cost of the update the
using the RA messages. We will get the network
message about the new care-of address of a mobile
load and hop distance. The network load will
node. It is measured as number of BU messages to
change according to time so to take the dynamic
the MAP and to the HA. Based on the regional care-
load we will consider the predefined time interval.
of-address, an on-line care-of-address (LCoA) the
From these RA messages we will get the Network
MAP BU message and the HA BU messages have
load so we will check the load on each MAP. Since
different effects. We are define an weighted binding
the time latency for each MAP to the MN is not
update cost as follows:
same, RA messages arrive at the MN at different
times. Therefore, the MN collects RA messages BU = α · NHA + β · NMAP (3)
during a predefined time interval (T). In equation 3, α and β are weight values for the HA,
2. Calculation of SMR: The MN estimates its Session MAP binding update. The hop distances between the
Mobility Ration (SMR) by calculating the number MN is directly proportional to these weight values and
of handoffs and session arrivals for each if they are not stated then we assign values like 10, 2.
Measurement Interval (MI). According to the time 4. Calculation of Location Rate (LR): In an unit time
interval the MN updates it SMR by comparing the how many different AR’s are attached to MN is
estimated SMR with the two SMR threshold values. called as Location Rate [13]. If the same AR
Session arrivial count N attached multiple times in an unit time we will list
SMR = = s
Mobility rate Nm only once. The visited AR’s addresses are recorded
(1) in list by the MN. The list is manages as follows: At
In the equation (1) Ns and Nm are the session arrival starting all AR lists are cleared, and the address of
count and mobility rate. In SMR Ns is the amount of the currently attached AR is added into the list as
ongoing session of the MN, which could be calculated the first entry. When MN changes its attached AR to
by MN itself during specific time duration (i.e. the new AR then it check the list for new AR address if
measurement interval). Nm is the mobility rate and that it is not in list then it will add to the list. At the end
is expressed as the AR’s coverage is divided by the of unit time we will calculate the Location rate
reside time which the MN within in the AR’s coverage. Number of entries
Location Rate = (4)
For each time interval (i.e. Measurement intervals), Unittime
the MN estimated its SMR by measuring the number of
5. Procedure for Dynamic MAP Selection: In
handoffs and session arrivals. At the same time, the
Dynamic MAP Selection first we collect the data
MN updates its SMR and compare with the estimated
about all about available MAPs. We will check the
SMR with the two SMR threshold values (th1 and th2).
load by sending the RA messages to all the MAP’s.
In order to select optimal MAP adaptively, we will
Calculate the SMR by calculating the eq 1. After
define two SMR threshold values: Minimum SMR
getting the values of load, SMR, BUC and Location
value and Maximum SMR value. In the MAP selection,
rate, we will compare with the threshold values and
if the new SMR is smaller than the lower SMR
if the values all are greater than the threshold values
threshold value, then we select that MAP because the
then we will select the optimal MAP otherwise we
mobility is relatively larger than the number of session
will start the process again. The process is
arrival, it will lead to the higher cost of a binding
described in following Flow Diagram.
update in this condition. Otherwise we will wait until
one time interval and again we will calculate the SMR Procedure for MAP Selection
of that MAP. The MN selects the MAP that minimizes
Define the values of min and max threshold value
the total cost.
1. For each MAP Mi
In order to avoid the sudden changes of session
activity or mobility rate, in this paper we are using the
Exponentially weighted moving average (EWMA) [7]
for estimation of the SMR.

www.ijorcs.org


1.1. Check the load of Mi extending Differentiated Services (Diffserv) to the
1.2. Calculate the SMR of Mi HMIPv6 architecture. DiffServ is a computer
1.3. Calculate the BUC of Mi networking architecture that specifies a
1.4. Calculate the LR of Mi simple, scalable and coarse-grained mechanism for
1.4.1. If (SMR(Mi) < th1 and load(Mi) <th2 and classifying and managing network traffic and
BUC(Mi)<th3 and LR(Mi)<th4) providing quality of service (QoS) on
MN will register with that Mi modern IP networks. DiffServ can be used to
Else provide low-latency to critical network traffic such
Wait for one time interval as voice or streaming media while providing
Endif simple best-effort service to non-critical services such
as web traffic or file transfers. By using this
2. End For mechanism, we can assure that better resource
utilization and throughput can be attained during
Handoff in HMIPv6 networks.
Based on DiffServ Mechanism, AR will classify
and mark packets. We are adding two new ICMPv6
options used by AR to advertise information to MN;
and then MN will use this information as criteria for
choosing the new access router in the handover
procedure.
1. AR’s Router Advertisement Message: The Router
Advertisement message is sent out periodically with
the type 3 option and that options are used to
specify the prefixes that are used for address auto
configuration. Using the available resources the
MN has to choose best nAR. For neighbor discovery
message of IPv6 we are implementing two new
additional ICMPv6 options. These two options are
type 9 and type 10 options, which are used for
advertising the remaining resources and Diffserv
QoS configuration parameters on certain Access
Routers.
It must attach type 9 and 10 options to the message,
when the Access Router sends out Router
Advertisement with type 3 options. MN uses
information about type 9 option to select the best
Access Router for handover. Type 9 option includes
current available bandwidth and dropping percentage
information of each class of service on Access Router
and Signal Strength. Type 10 option informs MN the
pre-defined Diffserv service class in which the
information for each class consisting of class
bandwidth and list of applications within each class
and Moving direction. We categorize the application
Figure 2: Flow Diagram for Adaptive MAP Selection by using protocol and port number. The information in
type 10 option is used when mobile determine to move
C. AR Selection into a new Diffserv domain. The option formats of
type 9 and 10 options are as depicted in Table 1.
After selecting the best MAP, the Diffserv approach
[11] is used to select the AR under the MAP, based on 2. Handover Procedure: Handover on HMIPv6 can be
its resource availability. The AR is implemented at the classified into two types. One is intra-MAP
edge router of Diffserv. We are adding two more handover and the second one is inter-MAP
parameters to the Diffserv to select an optimal AR and handover. Figure 4 describes the steps of Intra-
those two parameters are Signal Strength and Moving MAP handover procedure.
direction.
For AR selection, we propose a mechanism to
control network resources on mobile access routers by

www.ijorcs.org


Table 1: Router Advisement Message with Type 9, 10 Options

Available DiffServ
%Dropped Signal
Bandwidth Domain
Type 9 packets Strength
Width (Kbps) Number

TYPE
DiffServ
DiffServ DiffServ Moving
Domain
Type 10 Domain Class Domain Port directin
Protocol
Next port Next class ……

In Intra-MAP handover procedure at step 1, 2 and Steps of Inter-MAP handover procedure are shown
3, MN sends application data to CN via oAR and in Figure 5. This procedure is parallel to the Intra-Map
MAP. Each AR forward packets to the next hop and in handover, but it is different at step 6, 7, 8 and 9. In
that packet header a Differentiated Service Code Point case of Inter-MAP handover, MN has to update its
(DSCP) value is pinned. At step 4, MN moves into an LCOA and RCOA. The oMAP will get the data from
overlapping area of oAR, nAR1. The MN will receive nMAP and nMAP is selected using dynamic MAP
RtAvt (type 8 and 9) from both nARs. In this situation selection. After that, MN will inform nMAP the local
MN chooses nAR1, because it best matches resources binding update message that contains new LCOA and
available for MN. At step 5 and 6, MN makes Local RCOA and then perform Binding update with new
Binding Update (LBU) process with MAP. After that at RCOA to HA and CN.
step 8, 9 and 10, MN continues sending application
data to CN via nAR1.

Figure 3: Intra-MAP Handover Procedure

www.ijorcs.org


Figure 4: Inter-MAP Handover Procedure

In handover procedure there is no nARn so the process
was changed
3. Example of Handover Procedure: In the example
consider the one MAP with three AR’s and on CN.
The following procedure describes how MN will
communicate with MAP, AR’s and CN.
i. Intra-MAP Handover Procedure: In Intra-MAP
handover procedure at step 1, 2 and 3, MN sends
application data to CN via oAR and MAP. Figure
5 is explained step-by-step procedure.

− MN will send the RA messages to oAR to get the
details of which domain it is and under the
which MAP Figure 5: Intra-MAP Handover Procedure
− The oAR send the Diff Server pin code value to
MAP ii. Inter-MAP Handover Procedure: Steps of Inter-
− The MAP will send the data to CN MAP handover procedure are shown in Figure 6.
This procedure is parallel to the Intra-Map
− The new AR will send the type 9 and type 10
handover, but it is different at step 6, 7, 8 and 9.
values to MN
− The MN will do Local Binding with MAP − MN will send the RA messages to oAR to get the
− MAP will send the RA messages to MN to details of which domain it is and under the
choose the AR which MAP
− The MN will choose the New AR and send the − The oAR send the Diff Server pin code value to
data oMAP
− nAR1 send the data to MAP − The oAP will send the data to CN
− MAP will send data to CN which AR has − The new AR1 will send the type 9 and type 10
selected values to MN

www.ijorcs.org


− The MN will do Local Binding with nMAP and
the new MAP will select by using the dynamic
MAP selection
− nMAP will send the RA messages to MN to
choose the AR
− The MN sends the local binding update message
that contains new LCOA and RCOA and then
performs Binding updates with new RCOA to
HA.
− The MN sends the local binding update message
that contains new LCOA and RCOA and then
performs Binding updates with new RCOA to
CN.
− The MN will choose the New AR and send the
data
− nAR1 send the data to nMAP
− nMAP will send data to CN which AR has
selected
Figure 7: Simulation Topology
Table 2: Simulation Settings
No. of Mobile Nodes 5
No. of APs 5
Area Size 600 X 600
Mac 802.11
Simulation Time 50 sec
Traffic Source CBR
Packet Size 512
Speed 25 m/s
Transmission range 250m
Routing Protocol AODV
Figure 6: Inter-MAP Handover Procedure
B. Performance Metrics
IV. SIMULATION RESULTS We compare our proposed DMDA with P_HMIPv6
A. Simulation Setup scheme. We mainly evaluate the performance
according to the following metrics:
We use NS2 [12] to simulate our proposed
Dynamic MAP and Diffserv based AR (DMDA) Packet Delivery Ratio: It is the ratio of number of
method with P_HMIPv6 architecture. In our packets received successfully into total number of
simulation, the channel capacity of mobile hosts is set packets sent.
to the same value: 2 Mbps. We use the distributed Throughput: It is the average number of packets
coordination function (DCF) of IEEE 802.11 for received
wireless LANs as the MAC layer protocol. It has the Packet Drop: It is the average number of packets
functionality to notify the network layer about link dropped in the mobile hosts.
breakage. The following table (Table.1) summarizes Delay: It is average the end-to-end delay
the simulation settings. The CBR traffic is established
from CN to MS, and the bandwidth and latency for The performance results are presented in the next
every link between every two components are also section.
specified in this scenario. The topology in figure4 is
C. Results
used in our simulation. We have simulated both
horizontal (layer3) and vertical handoff in our system. Case 1: Complete Transmission
Figure 7 shows the screenshot of Network 1. Varying Rate: In our first experiment we vary the
Animator (NAM) output of simulation topology. rate as 50,100,150,200 and 250kb.

www.ijorcs.org


From figure 9, we can see that the packet drop of
Rate Vs Delay(Com plete)
the proposed DMDA is less than the existing
P_HMIPv6 method.
10
Delay(Sec)

8 From figure 10, we can see that the throughput of
6 DMDA
our proposed DMDA is higher than the existing
4 P_HMIPv6
2
P_HMIPV6 method.
0
50 100 150 200 250
2. Based on Time: In our second experiment we vary
the time as 0,2,4,6,8,10 and 12 sec
Rate(Kb)

Time Vs Bandwidth(Complete)
Figure 7: Rate vs. Delay
0.25
Rate Vs DeliveryRatio(Com plete) 0.2
DMDA

Mb/s
0.15
0.5 0.1 P_HMIPv6
DeliveryRatio

0.4 0.05
0.3 DMDA 0
0.2 0 2 4 6 8 10 12
P_HMIPv6
0.1 Tim e(Sec)
0
50 100 150 200 250
Figure 11: Time vs. Bandwidth
Rate(Kb)

Tim e Vs PacketsReceived(Com plete)
Figure 8: Rate vs. Delivery Ratio
400
Rate Vs Drop(Com plete) 300
DMDA
Pkts

200
P_HMIPv6
400 100
300 0
DMDA
Pkts

200 0 2 4 6 8 10 12
P_HMIPv6
100 Tim e
0
50 100 150 200 250 Figure 12: Time vs. Packets Received
Rate(Kb)
From figure 11, we can see that the received
bandwidth of our proposed DMDA is higher than the
Figure 9: Rate vs. Drop existing P_HMIPv6 method.
Rate Vs Throughput(Com plete) From figure 12, we can see that the packets
received ratio of our proposed DMDA is higher than
1000 the existing P_HMIPV6 method
Throughput

800
600 DMDA
Case-2: Inter Transmission
400 P_HMIPv6
200 1. Based on Rate: In this experiment we vary the rate
0 as 50,100,150,200 and 250kb.
50 100 150 200 250
Rate(Kb) Rate Vs Delay(Inter)

Figure 10: Rate vs. Throughput 15
Delay(Sec)

From figure 7, we can see that the delay of our 10 DMDA
proposed DMDA is less than the existing P_HMIPv6 5 P_HMIPv6
method.
0
From figure 8, we can see that the delivery ratio of 50 100 150 200 250
Rate(Kb)
P_HMIPv6 method.
Figure 13: Rate vs. Delay

www.ijorcs.org


Rate Vs DeliveryRatio(Inter) Rate Vs Delay(Intra)

0.5 6
DeliveryRatio

Delay(Sec)
0.4
DMDA 4 DMDA
0.3
0.2 P_HMIPv6 2 P_HMIPv6
0.1
0 0
50 100 150 200 250 50 100 150 200 250

Rate(Kb) Rate(Kb)

Figure 14: Rate vs. Delivery Ratio Figure 17: Rate vs. Delay

Rate Vs DeliveryRatio(Intra)
Rate Vs Drop(Inter)

0.42

DeliveryRatio
300 0.4
250 0.38 DMDA
200 0.36
DMDA P_HMIPv6
Pkts

0.34
150 0.32
100 P_HMIPv6
0.3
50 50 100 150 200 250
0
Rate(Kb)
50 100 150 200 250
Rate(Kb)
Figure 18: Rate vs. Delivery Ratio

Figure 15: Rate vs. Drop Rate Vs Drop(Intra)

Rate Vs Throughput(Inter) 30

20 DMDA
Pkts

400
10 P_HMIPv6
Throughput

300
DMDA 0
200
P_HMIPv6 50 100 150 200 250
100 Rate(Kb)
0
50 100 150 200 250 Figure 19: Rate vs. Drop
Rate(Kb)
Rate Vs Throughput(Intra)
Figure 16: Rate vs. Throughput
600
Throughput

From figure 13, we can see that the delay of our
400 DMDA
proposed DMDA is less than the existing P_HMIPv6
method. 200 P_HMIPv6

From figure 14, we can see that the delivery ratio of 0
50 100 150 200 250
P_HMIPv6 method. Rate(Kb0

From figure 15, we can see that the packet drop of Figure 20: Rate vs. Throughput
our proposed DMDA is less than the existing
P_HMIPv6 method. From figure 17, we can see that the delay of our
proposed DMDA is less than the existing P_HMIPv6
From figure 16, we can see that the throughput of method.
P_HMIPV6 method From figure 18, we can see that the delivery ratio of
Case-3: Intra Transmission P_HMIPv6 method.
1. Based on Rate: In this experiment we vary the rate From figure 19, we can see that the packet drop of
as 50,100,150,200 and 250kb. our proposed DMDA is less than the existing
P_HMIPv6 method.

www.ijorcs.org


From figure 20, we can see that the throughput of [4] Xavier P´erez-Costa and Marc Torrent-Moreno, “A
our proposed DMDA is higher than the existing Performance Study of Hierarchical Mobile IPv6 from a
P_HMIPV6 method System Perspective”, IEEE International Conference on
Communications, 2003. ICC '03. doi:
10.1109/ICC.2003.1204221
V. CONCLUSION
[5] M.H. Habaebi, “Macro/micro-mobility fast handover in
In this paper, we have proposed a Dynamic MAP hierarchical mobile IPv6”, Computer Communications,
and Diffserv based ARs selection for Handoff in vol. 29, issue 5, Elsevier, 2004. doi: 10.1016
HMIPv6 networks. The dynamic MAP will be selected /j.comcom.2004.12.004
based on the values of load, Session Mobility Ratio, [6] Dong-Guen Kim, Ho-Jin Shin and Dong-Ryeol Shin, “
Binding update cost and Location rate. By considering A Network-based Handover Scheme for Hierarchical
the SMR in the selection of MAP, the MN is able to Mobile IPv6 over IEEE 802.16e”, 10th International
Conference on Advanced Communication Technology,
select a more appropriate MAP with respect to its own
ICACT 2008. doi: 10.1109/ICACT.2008.4493804
mobility and session activity. After selecting the best
[7] Sangheon Pack, Minji Nam, Taekyoung Kwon,
MAP, we are using the Differentiated Services
Yanghee Choi, “An adaptive mobility anchor point
(Diffserv) approach to select the AR under the MAP,
selection scheme in Hierarchical Mobile IPv6
based on its resource availability. The AR is networks”, Computer Communications, Volume 29,
implemented at the edge router of Diffserv. The Issue 16, 2006. doi: 10.1016/j.comcom.2005.11.004
advantage of using the Diffserv is to enable MN to be [8] Ying-Hong Wang, Chih-Peng Hsu and Chien-Shan
able to choose the best AR before performing the Kuo, “Adaptive MAP Selection with Load Balancing
handover and quality of service of each application Mechanism for the Hierarchical Mobile IPv6”, 2009.
type on MN. As a future work, we wish to analyze the [9] WonSik Chung and SuKyoung Lee, “Cost-Effective
effects of losses and provide undisturbed transmission MAP Selection in HMIPv6 Networks”, IEEE
of real time traffic by suitable transport layer International Conference on Communications, ICC '07.
mechanisms. 10.1109/ICC.2007.998
[10] Yuh-Shyan Chen and Kun-Lin Wu, “A cross-layer
VI. REFERENCES partner-assisted handoff scheme for hierarchical mobile
IPv6 in IEEE 802.16e systems”, Wireless
[1] Sapna Gambhir, M.N.Doja, Moinuddin and Mohit Communications and Mobile Computing, 2009. doi:
Gambhir, “A Novel Approach to Reduce Signaling 10.1002/wcm.844
Delay in HMIPv6 mobile Networks”, International
[11] Sophon Mongkolluksamee, Vasaka Visoottiviseth,
Journal of Engineering and Technology Vol. 1, No. 1,
April, 2009. “Diffserv Conditionalized Handover for HMIPv6”, 10th
International Conference on Advanced Communication
[2] Tarik Taleb, Abbas Jamalipour, Yoshiaki Nemoto, and Technology, ICACT 2008. doi:
Nei Kato, “DEMAPS: A Load-Transition-Based 10.1109/ICACT.2008.4494026
Mobility Management Scheme for an Efficient
[12] Network Simulator: http:///www.isi.edu/nsnam/ns
Selection of MAP in Mobile IPv6 Networks”, IEEE
Transactions on Vehicular Technology, Vol. 58, No. 2, [13] Lusheng Wang, Brahim Gaabab, David Binet and
Feb. 2009. doi: 10.1109/TVT.2008.927037 Daniel Kofman, “Novel MAP Selection Scheme Using
[3] Indra Vivaldi, Mohd Hadi Habaebi1, Borhan Mohd Ali
Location History in Hierarchical MIPv6 Networks”,
and V. Prakash, “Fast Handover Algorithm For IEEE Wireless Communications and Networking
Hierarchical Mobile Ipv6 Macro-Mobility Conference, WCNC 2008. doi: 10.1109/WCNC.
Management”, The 9th Asia-Pacific Conference on 2008.426
Communications, APCC 2003. doi: 10.1109/APCC.
2003.1274434

How to cite
Shiva Prasad Kaleru, Damodaram Avula, "Dynamic Map and Diffserv Based AR Selection for Handoff in
HMIPv6 Networks". International Journal of Research in Computer Science, 3 (1): pp. 35-45, January 2013.
doi: 10.7815/ijorcs. 31.2013.059

www.ijorcs.org

Dynamic Map and Diffserv Based AR Selection for Handoff in HMIPv6 Networks

More Related Content

What's hot (20)

Similar to Dynamic Map and Diffserv Based AR Selection for Handoff in HMIPv6 Networks (20)

More from IJORCS (20)

Recently uploaded (20)

Dynamic Map and Diffserv Based AR Selection for Handoff in HMIPv6 Networks