Virtual ID: A Technique for Mobility, Multi-Homing, and Location Privacy in Next Generation Wireless Networks

Abstract

Cellular networking standards organizations such as the 3^rd Generation Partnership Project (3GPP) are currently developing System Architecture Evolution (SAE) as their core network architecture. SAE is all-IP based. However, IP-based networks face several known issues, such as mobility, multi-homing, location privacy, path preference, etc. Mobile IP (MIP) and its variants, such as Mobile IPv6 (MIPv6), Hierarchical MIP, and Proxy MIP, have been developed primarily to alleviate the mobility problem. These variation and extensions, however, still do not provide many of the features required in Next Generation Wireless Networks (NGWN). The limitations are especially due to the overloading of IP addresses as both node identity and locator. In this paper, we propose an extension to MIPv6 called Virtual ID. This concept applies the ID/Locator split idea into a Mobile IPv6 environment. Virtual ID and its extensions provide many features that would be desired in the NGWN. Since our proposed scheme is based on the standard MIPv6 and Proxy MIPv6, the scheme is fully compatible with the legacy MIPv6.

Full text

Abstract—Cellular networking standards organizations such
as the 3rd Generation Partnership Project (3GPP) are currently
developing System Architecture Evolution (SAE) as their core
network architecture. SAE is all-IP based. However, IP-based
networks face several known issues, such as mobility, multi-
homing, location privacy, path preference, etc. Mobile IP (MIP)
and its variants, such as Mobile IPv6 (MIPv6), Hierarchical MIP,
and Proxy MIP, have been developed primarily to alleviate the
mobility problem. These variation and extensions, however, still
do not provide many of the features required in Next Generation
Wireless Networks (NGWN). The limitations are especially due
to the overloading of IP addresses as both node identity and
locator. In this paper, we propose an extension to MIPv6 called
Virtual ID. This concept applies the ID/Locator split idea into a
Mobile IPv6 environment. Virtual ID and its extensions provide
many features that would be desired in the NGWN. Since our
proposed scheme is based on the standard MIPv6 and Proxy
MIPv6, the scheme is fully compatible with the legacy MIPv6.
Index Terms—Mobile IP, Virtual Identity, Mobility, Multi-
Homing, User Location Privacy, NGWN, ID/Locator Split.
I. INTRODUCTION
ystem Architecture Evolution (SAE) is the core
networking architecture being developed by the 3rd
Generation Partnership Project (3GPP) [1] for the next
generation of cellular wireless networks. SAE is all-IP based.
In this paper, we discuss issues that the next generation
wireless networks (NGWN) will face after SAE deployment.
We call these post-SAE networks. These networks, which will
result from converging wired and wireless networks, will
include a variety of wireless technologies such as cellular
networks (2G/3G/4G), wireless broadband networks (e.g.,
Mobile WiMAX and LTE), wireless sensor networks, and so
on. The interoperability among traditional wired networks,
such as Ethernet, and wireless technologies also needs to be
maintained. The applications on these networks may include
voice, video, TV broadcasting, online games, and data
services with a guaranteed quality of service (QoS).
With the emergence of billions of mobile users and wireless
devices, scalability and deployability issues arise and will
Manuscript received September 7, 2009, revised October 15, 2009.
Corresponding authors: C. So-In, R. Jain, S. Paul, and J. Pan are with the
Computer Science and Engineering Department, Washington University in St.
Louis, MO 63143 USA (e-mail: cs5, jain, paul, and jp10@cse.wustl.edu).
need to be considered when designing the next generation
networks. Compared to wired networks, the channel capacity
in wireless networks is not constant over time and distance. In
addition, mobile users may move from one location to another
at a high speed. As a result, disruptions may occur more
frequently. Therefore, mobility is clearly one of the key issues
in the NGWN.
With the advance of networking technologies, multiple
networking interfaces with different combination of wired and
wireless technologies are becoming common. So the issue of
multi-homing, especially device and user multi-homing, will
play an important role in backup, load balancing, sharing, and
traffic engineering in future networks. The networks are
becoming more user-centric, that is, they will allow users to
make their own decisions. Network service providers may
only provide suggestions with inherent security. For example,
with multiple networking interfaces in a single mobile device,
the mobile users may choose their preferred paths for each
task, probably based on the price paid and on the quality of the
service offered by various service providers. The users may be
required to pay air-time charges, similar to a traditional
cellular phone system. In addition, users may want to keep
their location information private from their correspondent
users. This is the so-called user location privacy issue. Finally,
the security of data is always of concern to users.
The issues we have described above: mobility, multi-
homing, scalability, security, deployability, and user location
privacy, are key required for the design of next generation
networks. Since future networks are expected to be all-IP
based, the question is how to make them support these
features.
There have been many attempts to resolve some of these
key issues, especially in traditional all-IP based wired-
networks [2, 3, 4]. However, no clear consensus has been
reached. The problem is more serious within the mobile
wireless environment. In the current Internet, the main hurdle
in resolving the mobility and multi-homing issues is the
overloading of IP addresses as both identity and location [2,
3]. The techniques to resolve these problems are based on
redirection and indirection techniques [2, 3]. The main
differences among these techniques are their varying focuses
on the different protocol layers, on the introduction of new
naming spaces, on the required changes of a protocol stack,
and on the ways to separate a host’s identity from its locator.
We will briefly discuss the detailed concepts. Mobile IP [5 to
9] is another well known approach primarily designed to
Virtual ID: A Technique for Mobility, Multi-
Homing, and Location Privacy in Next
Generation Wireless Networks
Chakchai So-In, Student Member, IEEE, and Raj Jain, Fellow, IEEE
Subharthi Paul and Jianli Pan, Student Members, IEEE
S
U.S. Government work not protected by U.S. copyright
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE CCNC 2010 proceedings
Proceedings of 2nd IEEE International Workshop on Mobile IPv6 and Network-based Localized Mobility
Management (MobiWorld’10), In Conjuction with 7th Annual IEEE Consumer Communications and Networking
Conference (CCNC 2010), Las Vegas, Nevada, USA, January 9th, 2010

resolve the mobility issue. However, Mobile IP and its
extensions fail to fully support important other features for
NGWN.
In this paper, our focus is on a network layer approach to
mobility. A key advantage of this approach is that the
network-layer based solutions require no change in the higher
layers of the protocol stack, and so the solutions work for all
applications. We apply the ID/locator split idea explicitly into
the Mobile IP, especially for a mobile wireless environment.
We introduce Virtual ID as a node identity for the mobile user.
This add-on feature makes Mobile IPv6 to fully support
mobility, multi-homing, and user location privacy. This
concept and its extensions are built on the standard Mobile
IPv6 and Proxy MIPv6. Note that several other proposals
focus on scalability (e.g., the use of DNS and provider-
aggregatable addresses) and security, such as IPsec and secure
Mobile IP signaling. We do not handle these issues in this
paper.
This paper is organized as follows. In Section II, we briefly
describe the general concept of ID/locator split approaches as
well as the pros and cons of these approaches in terms of
mobility, multi-homing as well as user path preference, and
user location privacy. In Section III, we discuss Mobile IPv6
and its variants considering these criteria. Then, in Section IV
we introduce Virtual ID and its extensions by applying the
concept of the ID/locator split into Mobile IPv6, to fully
support user mobility, multi-homing as well as user path
preference, and user location privacy. We illustrate these
Virtual ID ideas with several detailed examples in Section V.
Finally, the conclusions are discussed in Section VI.
II. ID/LOCATOR SPLIT
The ID/locator split [5, 6] is a well-known approach used to
resolve both mobility and multi-homing issues. Basically, the
idea is to separate the functionality of the identity from that of
the locator. Each mobile node (MN) has its own unique
identity. When the node moves, its identity does not change,
but its locator does. The identity can be a string of characters
or digits. The locator represents the current point of
attachment to the network. In other words, the locator helps
decide where the packet should be routed.
Currently, there are two ways to implement an ID/locator
split: placing a split in the end host (e.g., HIP, SHIM6, and
MILSA, etc. [2]) or in the network (LISP [2]). The former
approach requires the insertion of a new ID sub-layer usually
between the transport and the network layers. Thus, the upper
layers are bound to an ID instead of locator. HIP and MILSA
introduce new secure naming spaces but SHIM6 uses one of
its current locators as the identity.
Note that although these splitting techniques can support
full mobility, multi-homing, and location privacy since the
identity is used instead of the node location, such indirection
mechanisms also require new naming and name resolution
mechanisms. In addition, there is no detailed discussion of the
path selection issue. The second set of splitting techniques
implements an ID/locator split in the network. The basic idea
is that there is no change to the end host. The routers take care
of the split. At the edge of the network, the IDs are resolved
into the locators needed for communication. This requires
changes to network infrastructure devices (routers).
III. MOBILE IPV6 AND ITS VARIANTS
Mobile IP (MIP) [5, 6] and its variants are well-known
techniques designed to resolve the mobility problem in
traditional wired and wireless networks. 3GPP has adopted
these concepts for System Architecture Evolution (SAE).
Most of the concepts discussed in this paper apply to both
IPv4 and IPv6. However, for simplicity, we limit our
discussion to IPv6 since it has sufficient address space and is
preferred for public wireless networks.
Consider mobility. If nodes change their networks and/or
locations, then their IP addresses also change. Consequently,
their TCP connections at the transport layer are broken. The
Mobile IPv6 is potentially used to maintain the connection
and/or session regardless of time and location with an IP-in-IP
encapsulation technique. In other words, the Mobile IPv6 is
used to preserve the connection.
Briefly, the Mobile IPv6 functions as follows: the node’s
home IP address is used as the node’s identity. When the node
moves from one network to another network, it informs its
home network (home agent, HA) about its new IP address
(care-of-address or CoA). In case a correspondent node (CN)
wants to contact this node, the CN sends packets to the home
network; the packet is intercepted by the home agent and
forwarded to the mobile node’s new address (CoA).
Several extensions of Mobile IP have been proposed to
mitigate the route-to-home network delay and/or hand-off
latency such as HAWAII, Cellular IP, and HMIP (Hierarchical
MIP) [7]. These approaches deploy several home agents in a
hierarchical manner, especially at the edge routers. With
HMIP, the binding update is sent to the local HA, which
decreases delay latency. However, these approaches require
synchronization among HAs and additional nodes.
Proxy-MIP [8] was originally introduced to improve the
deployability of MIP. The idea is to use the router or proxy
agent to act on behalf of the mobile node and to perform the
MIP functionality. In other words, with the Proxy-MIP, the
mobile node does not need to support the MIP.
Now let us consider multi-homing and user path preference.
Mobile IPv6 can’t support multi-homing because each single
mobile node is bound to only one IP address. Recently, some
have suggested allowing multiple care-of-addresses
registrations [9] to allow multi-homing. There is no detailed
discussion on user path selection issue.
Consider user location privacy. When nodes move away
from their home networks, Mobile IP implicitly supports
location privacy because the current location is no longer
bound to the home address. However, this scenario introduces
a triangular routing problem as indicated earlier. This problem
can be mitigated using HMIP to reduce the delay latency by
placing the home agent close to both the MN and the CN. In
Mobile IPv6, a route optimization feature was introduced to
resolve this triangular routing problem, that is, to allow the
MN and CN to communicate to each other directly. But again,
this direct communication introduces the user location privacy
issue for mobile users.
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE CCNC 2010 proceedings

In summary, a traditional Mobile IPv6 and its variants -
Proxy MIPv6 and Hierarchical MIPv6 - fail to provide a full
support for mobility, multi-homing, and user location privacy.
In next section, then we introduce the concept of Virtual ID
and its extensions to overcome these drawbacks of the
traditional Mobile IPv6 by applying the concept of ID/locator
split explicitly to a Mobile IPv6 environment.
IV. VIRTU AL ID AND ITS EXTENSIONS
In this section, we first describe the idea of Virtual Identity
(ID) applied to Mobile IPv6. We also discuss how to apply
this concept to solve two different problems: user location
privacy and multi-homing.
A. Virtual ID
In IPv6, a 128-bit address is used for both node identity and
locator which introduces many disadvantages, as indicated
earlier. In a mobile wireless environment, Mobile IPv6 also
mixes these functionalities. When the mobile node is in the
home network, a single IPv6 home address represents both
node identity and locator. But when the mobile node is outside
the home network, Mobile IPv6 can be treated as an
ID/Locator split scheme because another IP address, CoA, is
involved. This CoA can be treated as the node locator (the
indicator of where the node is). The node’s home address does
not change with its location and, therefore, serves as the
node’s identity.
To clearly separate the function of identity from that of
locator in Mobile IPv6, we introduce the concept of a “virtual
home address”. Similar to SHIM6, this 128-bit address format
is used to represent the node’s identity. However, we do not
use one of the node current addresses as its identity. Instead,
we use the virtual home address, called virtual ID.
The virtual ID is pre-defined and randomly assigned by the
service provider. This ID is permanent and thus no longer
bound to the home networks and/or locations. In other words,
the virtual ID is used even when the mobile node resides in the
home network. As in Mobile IPv6, the IP-in-IP encapsulation
is applied in that the nodes update their CoAs when they are in
different location/networks.
We use the 128-bit IPv6 address format to represent the
node’s identity. This allows backward compatibility since
legacy nodes (virtual ID unaware nodes) treat these identities
as addresses.
B. User Location Privacy
The concept of the virtual ID formed by separating the node
identity from its location helps to resolve the issue of user
location privacy in that the correspondent nodes do not know
the location of the mobile node; only the node identity. Note
that basically there are two levels of mapping: from node
name (FQDN) to node identity, and then from node identity to
node location. The result of the DNS resolution is the node’s
identity, not its location. The other mapping level can be done
at rendezvous servers. In a Mobile IPv6 environment, the
home agent does the second level mapping. With Virtual ID,
an additional mapping from the virtual home address to the
Mobile IPv6 home address is also required. Optional
additional mapping servers or extensions of the home agent
can do this mapping.
The correspondent nodes are required to send the packets
through the mobile node’s home network. Therefore, there is a
triangulation issue, as discussed earlier. To solve this problem,
we propose an add-on feature to Proxy Mobile IPv6 [9].
Traditionally, in Proxy MIPv6, a mobile access gateway or
proxy node is used to provide Mobile IP functionality on
behalf of Mobile-IP unaware nodes. In this add-on, the mobile
nodes no longer require the location information; instead the
proxy node does this work. The proxy can optionally rewrite
the address with its selected anonymity proxy address to hide
the exact location or CoA in case the Mobile IP functionality
is performed at the end node.
C. Multi-Homing
In NGWN, mobile users will want to exercise user path
selection because they will have to pay for their choices
according to bandwidth constraints and other quality of
service controls. For example, suppose Alice buys access
services from two different service providers: one services a
3G network accessible to her cellular phone; the other is over
WLAN. When she is at home or when WLAN is available,
Alice would prefer accessing the Internet service through
WLAN and also probably disable the 3G service, especially
when air-time charges are high.
To meet these requirements of multi-homing and user path
selection, again a 128 bits Virtual ID is used as the unique
identity. We do not change the identity regardless of the
networks and/or locations. Only the physical locations or care-
of-addresses can be changed with a change of locations.
Unlike SHIM6, we do not change the protocol stack, but
instead we apply the concept of multiple CoAs registrations at
the home agent to support multi-homing feature [9].
In NGWN, users should be able to choose both their own
ingress and egress paths, based on the price paid and the
quality of service constraints. For simplicity, we use a weight
factor along with the CoA registration when the nodes update
the address to the home agent. In a more general case, the
users could specify a set of connection rules. The home agent
will forward the packets to the node according to pre-selected
user path rules.
V. VIRTU AL ID AND ITS EXTENSIONS: EXAMPLES
In this section, we provide detailed examples for the Virtual
ID concept and its extension. For user location privacy, we
show that the concept of Virtual ID can protect the user
location information. We also show that home agent chaining
can be used to support user and/or device multi-homing as
well as how to support a user path selection feature.
A. Virtual ID Example
Fig. 1 shows an example of Virtual ID. In this figure,
Alice’s node’s name is Alice.xyz.com registered at a domain
name server or DNS. The service provider (SP) allocates a
virtual address (Virtual ID), ::10.2.1.2, as Alice’s identity.
Suppose the SP networks are ::10.x.x.x with ::10.3.x.x and
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE CCNC 2010 proceedings

::10.4.x.x sub-networks assigned into different physical
regions. The virtual addresses ::10.2.x.x are specifically
dedicated as the virtual ID. Only the SP knows the mapping
between the virtual ID or node identity (::10.2.1.2) and the
physical address (::10.3.1.2) or current IP address. This
mapping can be stored at rendezvous servers.
Alice.xyz.com
Virtual ID, ::10.2.1.2
DNS
Alice.xyz.com
::10.2.1.2
::10.2.1.2 →::10.3.1.2
Mapping Server
HoA (Virtual ID):
Assigned regardless of location
::10.3.x.x
::10.4.x.x
SP1 ::10.x.x.x
Fig. 1. Virtual ID Example
B. User Location Privacy Example
This section describes two main scenarios that use Virtual
ID to achieve a user location privacy requirement in NGWN:
when the correspondent code (CN) resides either out of the
home network or inside the home network.
Alice.xyz.com
Virtual ID, ::10.2.1.2
Liza.abc.com
::11.1.1.2
::10.2.1.2 →::10.3.1.2
Mapping Server
SP 1
SP 2
DNS
Liza.abc.com ::11.1.1.2
Alice.xyz.com ::10.2.1.2
Fig. 2. Virtual ID with User Location Privacy Example
The first scenario is when the node is in a different network.
Fig. 2 shows the correspondent node or Liza.abc.com in SP2
contacting Alice.xyz.com, which is in SP1. First, Liza retrieves
Alice’s identity, ::10.2.1.2, from a DNS resolution process and
uses that ID to route packets to Liza’s home network. Since
Alice’s ID is used instead of her physical attached address,
::10.3.1.2, Alice’s location privacy can be maintained. Notice
that if Alice is in foreign networks, her user location privacy is
implicitly maintained. This scenario is similar to a traditional
Mobile IPv6 because the permanent home address is different
from the virtual ID.
The other scenario is when Liza is within the same network,
say in an SP1 network, with ::10.x.x.x networks. Suppose
Liza’s address is ::10.4.1.2 and again Alice is at ::10.3.1.2,
within her home network. With a traditional Mobile IPv6, Liza
knows where the current location of Alice is. However, with
Virtual ID, Alice’s identity is used instead, ::10.2.1.2;
therefore, Liza no longer knows Alice’s location information.
C. Proxy-assisted User Location Privacy Example
Fig. 3 shows an example of Proxy-assisted Mobile IPv6 and
Virtual ID providing user location privacy. In this figure, Liza,
a correspondent node, wants to contact Alice, who is not in her
own home network. Note that the proxy will do both the
binding update and the Mobile IP functionality on behalf of
the mobile nodes. The binding update refers to a pairing of the
virtual ID and the proxy locator: ::9.1.1.2 and ::11.5.1.1 for
Liza and ::10.1.1.2 and ::12.5.1.1 for Alice. Liza does not
know Alice’s location and vice versa. Notice that each proxy
has local node location information so that the proxy can
forward the packets to the correct final destination.
Alice.xyz.com
Virtual ID, ::10.1.2.2
Locator, ::12.3.1.2
DNS
Liza.abc.com ::9.1.1.2
Alice.xyz.com ::10.1.1.2
Proxy1
Locator, ::11.5.1.1 Proxy2
Locator, ::12.5.1.1
Liza.abc.com
Virtual ID, ::9.1.1.2
Locator, ::11.3.1.1
Alice Home Network
::10.x.x.x
Fig. 3. Proxy-assisted User Location Privacy Example
D. Multi-Homing Example
In this section, we provide the details of how to incorporate
the multi-homing feature into NGWN. Our proposal is based
on home agent chaining among service providers.
We consider two main scenarios: when the multi-homing
attachments are either to the same service provider or to
different service providers.
SP1 ::12.x.x.x
Alice.xyz.com
Virtual ID, ::12.3.1.2
::12.2.1.2
::12.1.1.2
HA
Alice.xyz.com
Virtual ID, ::12.3.1.2
CoAs, ::12.2.1.2 w=1
::12.1.1.2 w=2
If move/change preference
→ update CoAs
Fig. 4. Multiple CoAs Registration Example
The first scenario, Fig. 4 shows the process of multiple
CoAs registrations with the preferred path selection when both
networking attachment points are with the same service
provider (SP1). In this figure, Alice has two access
technologies with the same service provider (::12.x.x.x):
toward cellular networks and toward WLAN on her single
mobile device. Alice’s virtual ID is::12.3.1.2, and the two
physical locators or CoAs are ::12.2.1.2 (on a 3G network),
and ::12.1.1.2 (on WLAN). When Alice is at home, she can
send the update to her home agent to set a higher priority
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE CCNC 2010 proceedings

toward the WLAN interface so that the inbound traffic can be
forwarded toward this WLAN interface.
The other scenario is when mobile users have multiple
access services from different network providers. Fig. 5a
shows this configuration. As shown, Alice has two access
services from two different service providers: SP1 cellular
networks and SP2 WLAN. Since there are different SPs, Alice
can acquire two different virtual IDs. Alice can send the
update to the DNS server with her preferred path selections
(with different weights). In this scenario, Alice is at home and
she prefers the WLAN path (with a higher weight, or higher
priority), which is toward SP2 or ::11.x.x.x networks.
SP 1
::10.x.x.x
SP 2
::11.x.x.x
Alice.xyz.com
Virtual ID, ::10.3.1.2
::11.3.1.2
::10.1.1.2
::11.1.1.2
DNS
Alice.xyz.com
::10.3.1.2 w=1
::11.3.1.2 w=2
HA1 HA2 Mapping
::11.3.1.2 →::11.1.1.2
Mapping
::10.3.1.2 →::10.1.1.2
(a)
SP 1
::10.x.x.x
SP 2
::11.x.x.x
DNS
Alice.xyz.com
::10.3.1.1 w=1
::11.3.1.1 w=2
HA1 HA2
::10.1.1.2
::11.1.1.2
Liza.abc.com
::10.5.1.2
Alice.xyz.com
Virtual ID, ::10.3.1.2
::11.3.1.2
(b)
Fig. 5. Multi-homing Feature in Mobile IPv6 with Virtual ID Example
Note that the CoA address of Alice on the SP2 network is
::11.1.1.2, not the Virtual ID ::11.3.1.2. In this scenario, the
packets are sent only towards WLAN as long as Alice does
not update her preferred path on its DNS. There are no
requirements for cooperation and interaction between two
service providers.
When WLAN is not working, as shown in Fig. 5b, either
SP2 or Alice can detect the disconnection. Without the
interaction between the SPs, the packets can continue to flow
to SP2 until Alice sends the update to the DNS server.
Therefore, an additional operation is required. We recommend
the concept of home agent chaining. Similar to the cellular
phone system roaming mechanism, both service providers
should have an agreement based on their user roaming policy
to provide a packet forwarding mechanism. In this example,
suppose SP1 and SP2 have a roaming agreement, and mobile
users agree to pay for the additional cost of roaming.
During the disconnection, the steps in Fig. 5 are as follows:
Liza, ::10.5.1.2, originally sends her packets to Alice through
SP2. Due to a link failure, the WLAN interface of Alice is
unreachable. After the link failure detection, HA2 (from SP2)
redirects all packets with Alice indicated as the destination to
HA1 in order to reach Alice. Again, this redirection is based
on a roaming policy. Whenever Alice sends the update to the
DNS server to withdraw the disconnected path and/or to set a
lower preference, this redirection will be terminated. Note that
this example shows two service providers; however, the
chaining concept can still apply with more service providers
with multiple networking interfaces.
VI. CONCLUSIONS
Next Generation Wireless Networks, or NGWN, will be a
cloud of all IP-based networks. The main features of these
networks will be to fully support user mobility, multi-homing,
user location privacy, and so on. Mobile IP and its variants
have been introduced to resolve some of these issues. These
proposals focus on a network layer technique. Some of these
techniques have been selected by 3GPP for the System
Evolution Architecture standard. However, these techniques
have several limitations, especially due to the problem of
identity and locator overloading.
In this paper, we discussed Mobile IP and its variants and
also pointed out several drawbacks. In addition, we introduced
a new technique called Virtual ID and its extensions, to make
the Mobile IP fully support mobility, multi-homing, and user
location privacy as well as user path selection. These add-ons
are based on the standard Mobile IPv6 and its extensions and
are therefore easy to be deployed along with Mobile IPv6.
REFERENCES
[1] 3GGPP TS 23.402 V8.0.0 3rd Generation Partnership Project; Technical
Specification Group Service and System Aspects; Architecture
enhancements for nono-3GPP accesses, Dec. 2007, 131 pp.
[2] R. Jain, “Internet 3.0: Ten Problems with Current Internet Architecture
and Solutions for the Next Generation,” in Proc. IEEE Military Comm.
Conf., 2006, pp. 1-9.
[3] C. So-In, R. Jain, J. Pan, and S. Paul, “Next Generation Wireless
Networks: key issues and survey,” Submitted to EUSASIP Journal on
Wireless Communicaiton and Networking, Oct. 2009.
[4] S. Paul, J. Pan, and R. Jain, “A Survey of Naming Systems:
Classification and Analysis of the Current Schemes Using a New
Naming Reference Model,” To appear in Computer Communicatoin,
May 2010.
[5] C. Perkins, Ed., “IP Mobility Support for IPv4,” RFC 3220, Jan. 2002.
[6] D. Johonson, C. Perkins, and J. Arkko, “Mobility Support in IPv6,” RFC
3775, June 2004
[7] A.T. Campbell, J. Gomez, K. Sanghyo, W. Chieh-Yih, Z.R. Turanyi,
and A.G. Valko, “Comparison of IP micromobility protocols,” IEEE
Wireless Comm. Mag., vol. 9, no. 1, pp. 72-82, Feb. 2002.
[8] S. Gundavelli, Ed. K. Leung, V. Devarapalli, K. Chowdhury, and B.
Patil, “Proxy Mobile IPv6,” RFC 5213, Aug. 2008.
[9] R. Wakikawa, V. Devarapalli, G. Tsirtsis, T. Ernst, and K. Nagami,
“Multiple Care-of Addresses Registration,” Internet-Draft, draft-ietf-
monami6-multiplecoa-14.txt, May 2009.
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE CCNC 2010 proceedings