OPTIMUM EFFICIENT MOBILITY MANAGEMENT SCHEME FOR IPv6

International Journal of Next-Generation Networks (IJNGN) Vol.4, No.3,September 2012
DOI : 10.5121/ijngn.2012.4305 63
OPTIMUM EFFICIENT MOBILITY
MANAGEMENT SCHEME FOR IPv6
Virender Kumar
Department of Electronics & Communication Engineering, HCTM Technical Campus,
Kaithal, India
gangotrahctm@gmail.com
ABSTRACT
Mobile IPv6 (MIPv6) and Hierarchical Mobile IPv6 (HMIPv6) both are the mobility management solutions
proposed by the Internet Engineering Task Force (IETF) to support IP Mobility. It’s been an important
issue, that upon certain condition, out of MIPv6 and HMIPv6 which one is better. In this paper an
Optimum Efficient Mobility Management (OEMM) scheme is described on the basis of analytical model
which shows that OEMM Scheme is better in terms of performance and applicability of MIPv6 and
HMIPv6. It shows that which one is better alternative between MIPv6 and HMIPv6 and if HMIPv6 is
adopted it chooses the best Mobility Anchor Point (MAP). Finally it is illustrated that OEMM scheme is
better than that of MIPv6 and HMIPv6.
KEYWORDS
Mobile IPv6; Hierarchical Mobile IPv6; Access Router; Regional Size; Mobility Anchor Point.
1. INTRODUCTION
With the fast increasing demand for the seamless mobility providers motivate to support seamless
connectivity to Mobile Nodes (MNs). To complete this aim, Internet Engineering Task Force
(IETF) proposed Mobile IP (MIP) protocol. MIPv4 and MIPv6 both are the mobility management
solution to maintain the on-going communication when one MN moves from one subnet to
another. MIPv6 become the next generation solution due to its several advantages over MIPv4.
HMIPv6 is another scheme which gives the several advantages over MIPv6. Although HMIPv6 is
the modification of the MIPv6 yet it can not outperform MIPv6 in all scenarios.
2. BACKGROUND
2.1 Mobile IPv6(MIPv6)
In Mobile IPv6 one MN is identified by two addresses: Home address and Care of address (CoA)
[1]. Home Address represents the permanent address of MN and Care of Address (CoA) is the
Temporary Address, representing the current location of MN. There is a mobility management
entity i.e. Home Agent (HA) which stores the binding information of the MN. Home Agent also
receives all the packets on behalf of the MN when the Correspondent Nodes do not know the
current location of the MN. In MIPv6 there is a process known as “Home Registration” in which
updated location is registered in HA when the MN roams in the visited networks. But in MIPv6 a
frequent handover by MN in a local region leads to a longer signaling delay. In Handover process
this longer signaling delay is the main problem of the MIPv6.

64
2.2 Hierarchical Mobile IPv6 (HMIPv6)
To solve this problem, Hierarchical Mobile IPv6 (HMIPv6) is introduced. In HMIPv6 [2] a new
entity is introduced known as Mobility Anchor Point (MAP) to act as a local Home Agent with in
a region. In HMIPv6 Mobility Anchor Point (MAP) have a number of Access Routers (ARs). The
number of ARs under a MAP is known as the Regional Size. In HMIPv6, there are two addresses:
Regional care of address (RCoA) representing the MN’s MAP & the on-Link CoA (LCoA)
representing the AR that the MN attaches to.There are two types of mobility in HMIPv6: micro-
mobility (handover with in a region) and macro-mobility (handover across the regions). In macro-
mobility, the MN gets two new addresses: RCoA, LCoA and it will initiate a regional registration
process to bind these two addresses. After having successful regional registration, the MN gives
its new update of having new RCoA to it’s HA i.e. there is a binding between its Home Address
and RCoA to the HA by a Home Registration. In micro-mobility there is only a regional
registration because there is no new RCoA of a MN within a region. Now, we see that in HMIPv6
when MN roams from one region to another, there is a double registration: regional registration
and home registration. So in HMIPv6 the handover latency is smaller than that of MIPv6 when
the MNs roam within the region but the handover latency is larger than that of MIPv6 when the
MNs roam inter- region. Besides, in double registration there is a MAP processing delay leading
to a longer packet delivery time because all the packets destined to MN are tunneled through
MAP. So, Double registration leads to a larger handover latency and longer packet delivery time.
So it is an interesting issue to select to find out the performance of MIPv6 and HMIPv6
depending upon certain conditions.
3. PERFORMANCE ANALYSIS OF MIPV6 AND HMIPV6
3.1 Relative Registration Cost [3]
Definition 1: (Relative Registration Cost): Relative registration cost (TR) is defined as the
average registration time saved by using HMIPv6 compared with MIPv6 [3]
TR may be positive or negative. TR > 0 means the average registration delay of MIPv6 is shorter
than that of HMIPv6, otherwise longer.
MAIN SYMBOLS IN REGISTRATION PERFORMANCE ANALYSES
Symbols Definitions
TR Average registration delay of MIPv6
TAM Average delay of delivering registration signaling over wireless link between AR
and MN
THA Average delay of delivering registration signaling between HA and AR
TH Average registration signal processing latency of HA
Tintra Average delay of a registration process in HMIPv6 during an intra-MAP handover
Tinter Average delay of a registration process in HMIPv6 during an inter-MAP handover
TMA Average delay of delivering registration signaling between MAP and AR
lMA Average distance between MAP and its reachable ARs
lHA Average distance between HA and AR
T Average dwell time that an MN stays in an AR
µ Unit distance signaling transmission cost of wired link
According to RFC3775[1] and RFC4140 [2] in MIPv6 there is only home registration but in
HMIPv6 there are to registrations: regional registration and home registration. Hence, TRM, Tintra
and Tinter can be calculated as:

65
TRM = 2TAM +2THA+TH (1)
Tintra= 2TAM +2TMA+TM (2)
Tinter= 4TAM +2TMA+2THA+TH+TM (3)
Let the MN needs mth
handover to move out of a region (m≥1).Then, in new region the MN will
enter at its mth
handover. So the total average delay (TIT ) that an MN spends for m handovers in
HMIPv6 and MIPv6 is [3]
TIT = (m−1) Tintra + Tinter (4)
TAT = mTRM (5)
Using definition 1 and equations (4) & (5), TR , can be calculated as
M A H A M H H
R
µ ( 2 θ + 2 m l - 2 l ( m - 1 ) + m ( T - T ) + T )
T = ( 6 )
m T
Where µ is unit distance signaling transmission cost of of wired link. We also suppose the
average signaling delivering delay of wireless link be θ.µ, where θ >1. From formulae (6) we can
say that if the nearer the distance between MN and MAP and the farther the distance between HA
and MN, then HMIPv6 gives higher average registration revenue. i.e. Only when, TR < 0,
HMIPv6 obtains the average registration revenue. Two theorems can be deduced.
Theorem 1: HMIPv6 outperforms MIPv6 in terms of registration revenue when an MN roams
within a region (intra-region) and the average registration revenue is |2µ.(lMA−lHA )/ T |. In micro-
mobility TR can be calculated as [4]
( )
R
2
T ( 7 )M A H A M Hl l T T
T
µ − + −
=
Theorem 2: TR lies on the regional size, K, when the MN roams across different regions (inter-
region). In this TR can be calculated as on certain conditions as [4]:
( )8
H H A
R
M A M H
( 2 µ . θ + T ) . ( 2 N - 2 K - 1 ) + 2 µ . l . ( 1 - 2 K )
T =
( 2 N - 2 ) . T
4 µ . ( N - 1 ) . l + 2 ( N - 1 ) . ( T - T )
+
( 2 N - 2 ) . T
3.2 Relative Packet Delivery Cost [3]
Definition 2: (Relative Packet Delivery Cost): Relative packet delivery cost (DP)[3] is defined
as the average time wasted by using HMIPv6 instead of MIPv6 to forward packets.
MAIN SYMBOLS IN PACKET DELIVERY PERFORMANCE ANALYSES
Symbols Definitions
DPM Average packet delivery delay of MIPv6
α. Average packet arrival rate

66
DH Average packet processing latency of HA
DCH Average delay of forwarding packets from CN to HA
DHA Average delay of forwarding packets from HA to AR
DAM Average delay of forwarding packets from AR to MN
DPH Average packet delivery delay of HMIPv6
DM Average packet processing delay of MAP
DHM Average delay of forwarding packets from HA to MAP
lHM Average distance between HA and MA
According to [1] and [2], the average latency of forwarding packets from a CN to the MN in
MIPv6 and HMIPv6 are
DPM = α.( DH + DCH + DHA +DAM ) (9)
DPH = α.( DH + DM +DCH +DHM + DMA + DAM ) (10)
According to definition 2, the average packet delivery cost is given by
DP = DPH - DPM = α.( DM + DHM + DMA - DHA ) (11)
We assume that δ is the average delay of encapsulating a packet in MAP, so DM can be calculated
as:
DM = A.wK + B.lg.K + δ (12)
Where A and B are positive coefficients
Assume that the average packet delivery delay of wired link is proportional to the number of hops
that the packets travel with the proportionality constant η. Then Equation becomes [4]
DP = α.( A.wK + B.lg.K +δ+η.(lHM +lMA-lHA )) (13)
Where α is average packet arrival time, δ is the average delay of encapsulating a packet in MAP,
A & B are coefficients, w.k is the average no. ARs in a region with assuming that an AR can
serve w MNs on average and lg is the logarithmic function.
Equation (13) leads to the conclusion that average packet delivery cost is positive on certain
condition [4]. When DP >0, it means average packet delivery delay of HMIPv6 is longer than that
of MIPv6.
3.3 Relative Cost [3]
Definition 3: (Total Cost Function): Total cost function denoted as CT gives the overall
performance of HMIPv6 against MIPv6 in terms of registration and packet delivery cost [3].
CT = n1.TR + n2.DP (14)
Where n1 > 0 and n2 > 0 are the coefficients.
As per the eq. (14), when CT > 0, MIPv6 will be more applicable than HMIPv6 otherwise
HMIPv6 is adopted.

67
3.4 The IMS Scheme [4]
When regional size K increases, HMIPv6 may gain more average registration revenue while
paying more average packet delivery cost. However, K cannot increase indefinitely due to the
processing bottleneck of the MAP [7]. The total average packet processing latency of the MAP is
given by α⋅(A.wK+B.lg.K+δ), which depends on its load. Thus, a proper K that minimizes CT will
optimize the overall performance of HMIPv6 against MIPv6. Denote such K as Kopt, which can
be solved as follows [4]
min CT(K )
α⋅ (A.wK + B.lg K+δ) < ψ (15)
Where ψ is a constant restricting the total packet processing latency of the MAP.
Definition 4: Cost function of HMIPv6: called CHMIPv6 formulates the absolute performance of
HMIPv6 in terms of the average registration and packet delivery delay. It is given by
intra inter
HMIPv6 1 2 PH
(m -1)T + T
C = n . + n .D
mT
(16)
Where n1 and n2 are the same as in Definition 3.
4. THE OEMM SCHEME
In this section we propose an Optimum Efficient Mobility Management (OEMM) scheme in
which Firstly; we employ an algorithm to evaluate the performance of MIPv6 and HMIPv6
against some key parameters using an analytical model. The algorithm of the impact of
parameters on the applicability of MIPv6 and HMIPv6 has been described in this section.
Secondly, we employ an algorithm which shows that how cost changes with α and T.
4.1 Algorithm for the performance of MIPv6 & HMIPv6 against Key parameters
1. Enter the x different values of Regional Size (K) and a fixed number of Access Routers N.
2. Enter the fixed values of T, α and lMA
3. For (i = 1 to x)
IF (Regional size [i] ≥ N (no. of different access routers)
Compute TR according to eq. (7).
Compute DP according to eq. (13).
Compute CT according to eq. (14).
Plot a graph between CT and Regional size
ELSE
Compute TR according to eq. (8).
Compute DP according to eq. (13).
Compute CT according to eq. (14).
Plot a graph between CT and Regional size
4. Change T to see the impact of T on CT for a fixed value of regional size and go to step 3.
5. Change α to see the impact of α on CT for a fixed value of regional size and go to step 3.
6. Change lMA to see the impact of lMA on CT for a fixed value of regional size and go to
step (3).
7. Exit.

68
4.2 Algorithm for the performance of MIPv6 & HMIPv6 to show cost changes with
α and T
1. Start
2. Input the value of OC, M & all parameters.
3. Input the value of Kopt(i) of MAP(i). i.e.(i=1,2………M).
4. If MN gets the information about parameters on handover.
5. Check for α, T or MAP.
6. If α, T, MAP changes then go to step 7, otherwise go to step 13.
7. Calculate Ct (i) of MAP (i) i.e.(i=1,2………M).
8. OC= Min.{Ct(i)} i.e.(i=1,2………M).
9. OKopt=K Arg min.{Ct(i)} i.e.(i=1,2………M).
10. If OC ≥ 0, then MN adopts MIPv6.
11. Else, OC < 0, then MN adopts HMIPv6 and MN chooses the MAP whose sequence no.
is OC.
12 The chosen MAP regional size OKopt
13 There is no information about Handover (MN will stay in same AR).
14 END

69
4.3 Flowchart for OEMM scheme
5. NUMERICAL RESULTS AND DISCUSSION
In this section, we demonstrate the performance of MIPv6 and HMIPv6 against some key
parameters. We employ numerical analysis to show the results. The parameters used in the
simulation process are taken from the various existing literature. The estimating value of α can be
found in [7,8] while T can be computed by the method introduced in [6]. In addition the value of
w and N are from [10]. The estimating value of lHM and lMA are taken from [9].
Start
MIPv6
Set count =1
Input K (Count)
IS
Count=M
Count=Count+1
No
yes
MN gets the
information about
all parameters
Check
Is, α=α′
Is, T=T′
Is, MAP no.=Map′ no
There is no
handover
yes
No
Set Temp = 1
Calculate CT (temp)
Is
CT (temp)<OC
NoIS
Temp=M
Temp=Temp+1
No
Set OC = CT (temp)
OM=Temp
OKopt=Kopt (temp)
OC≥0
yes
No
HMIPv6
Read OC,M
& all parameters
yes
yes
Choose MAP whose
sequence no. is OM & its
regional size is OK(opt)

70
Figure.1 Impact of T on CT
In figure 1 impact of T on CT has been shown. In this case the values of lMA=7 and α=.06. In this
figure we see that for K ≤ 13, CT is greater than 0 and increases with the decreasing of T.
In figure 2 the impact of α on CT has been shown. In this case the values of lMA=7 and T=110.
Since as α increases the average packet delivery cost is increased which leads to increase in CT.
0 5 10 15 20 25 30
-0.02
0
0.02
0.04
0.06
0.08
0.1
Regional size,K
TotalCostFunction,Ct
Alpha=.06
Alpha=.9
Alpha=1.6
Figure 2. Impact of α on CT
Figure 3 shows the impact of the distance (lMA) between AR and MAP on CT. In this case the
values of T=110 and α=.06 is taken. In this figure we see that as lMA is increased, the CT is also
increased. This is due to the fact that both average packet delivery delay and the average
registration delay is increased in HMIPv6 when the distance lMA increases, leading in the increase
in CT.
0 5 10 15 20 25 30
-0.02
-0.015
-0.01
-0.005
0
0.005
0.01
0.015
0.02
0.025
Regional size,K
T=110
T=510
T=1510

71
The comparison of the performance of OEMM, HMIPv6 (the regional size is 7 or 13) and MIPv6
in figures 4 and 5 using cost as the metric. The cost of the OEMM is taken from the formula (16).
Figures 4 & 5 shows that how cost changes with the changing of α and T. We also observe that
the cost of OEMM is minimum w.r.t.MIPv6 and HMIPv6.
0 5 10 15 20 25 30
-0.02
-0.01
0
0.01
0.02
0.03
0.04
0.05
Regional size,K
Lma=7
Lma=11
Lma=15
Figure 3. Impact of lMA on CT
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
Avg. Packet Arrival Rate
TotalCostfunction,Ct
Ct vs. Alpha
OEMM
HMIPv6,K=13
HMIPv6,K=7
MIPv6
Figure 4. Cost vs. α

72
0 10 20 30 40 50 60 70 80 90
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Avg. Dwell Time
Totalcostfunction,Ct
Ct vs. T
OEMM
HMIPv6,K=13
HMIPv6,K=7
MIPv6
Figure 5. Cost vs. T
6. CONCLUSION
In this paper we evaluated a performance of MIPv6 and HMIPv6 in the form of impact of certain
parameters on the total cost function. It is also evaluated that OEMM scheme shows the minimum
cost when it is compared with MIPv6 and HMIPv6. Finally, the performance of MIPv6 and
HMIPv6 has been simulated in this paper.
REFERENCES
[1] D. Johnson, C. Perkins, and J. Arkko, “Mobility Support in IPv6,” RFC3775, June 2004.
[2] H. Soliman, C. Castelluccia, K. El Malki, and L. Bellier, “Hierarchical Mobile IPv6 Mobility
Management (HMIPv6),” RFC4140, 2005.
[3] Shengling Wang, Yong Cui, Sajal K. Das, Wei Li, and Jianping Wu, “Mobility in IPv6: Whether and
How to Hierarchize the Network?” 1045-9219/11/$26.00 , 2011 IEEE
[4] Shengling Wang, Yong Cui, Sajal K.Das “Intelligent Mobility support for IPv6”. 978-1- 4244-2413-
9/08/$25.00 ©2008 IEEE
[5] H. Tzeng and T. Przygienda. “On Fast Address-Lookup Algorithms”. IEEE J. Selected Areas in
Communications, vol. 17, no. 6, pp. 1067-1082, 1999.
[6] Y. Chen and M. Huang. “A Novel MAP Selection Scheme by Using Abstraction Node in Hierarchical
MIPv6”. Proc. IEEE International Conference on Communications, 2006. pp:5408-5413
[7] H. Xie, S. Tabbane, and D.J. Goodman. “Dynamic Location Area Management and Performance
Analysis”. Proc. 43rd
IEEE Vehicular Technology Conference, 1993. pp:536-539
[8] M. Yabusaki. “Mobility/Traffic Adaptive Location Management”. Proc. IEEE 56th Vehicular
Technology Conference, Vancouver, 2002. pp:1011-1015
[9] W.R. Stevens, TCP/IP Illustrated, Volume 1: The Protocols. Addison Wesley Longman, Inc., 1994.
[10] J Xie. and I.F.Akyildiz. “A Novel Distributed Dynamic Location Management Scheme for
Minimizing Signaling Costs in Mobile IP”. IEEE Transactions on Mobile Computing, vol. 1, no. 3,
pp.163-175,2002.

OPTIMUM EFFICIENT MOBILITY MANAGEMENT SCHEME FOR IPv6

More Related Content

What's hot (20)

Similar to OPTIMUM EFFICIENT MOBILITY MANAGEMENT SCHEME FOR IPv6 (20)

More from ijngnjournal (20)

Recently uploaded (20)

OPTIMUM EFFICIENT MOBILITY MANAGEMENT SCHEME FOR IPv6