Technical University of Ilmenau
Faculty of Computer Science and Automation
Department of Integrated Communications Systems

Student research project
(Mobile IP in Source Routing Systems)

Supervisor: Dipl.-Inf. Florian Liers
Editor: Imad kailouh

Ilmenau, July 2010

1. Introduction ............................................................................................................. 4
1.1 Motivation .......................................................................................................... 4
1.2 Purpose of the work .......................................................................................... 4
2. Basics ..................................................................................................................... 5
2.1 Internet Protocol (IP) ......................................................................................... 5
2.1.1 IP Routing....................................................................................................... 5
2.1.1.1 What is routing ...................................................................................... 5
2.1.1.2 Source vs. Hop by Hop Routing ............................................................ 5
2.2 Mobile IP ........................................................................................................... 7
2.2.1 Terminology in Mobile IP ............................................................................. 7
2.2.2 How Mobile IP Works .................................................................................. 8
2.2.2.1 Agent Discovery .................................................................................... 9
2.2.2.2 Registration ......................................................................................... 11
2.2.2.3 Tunneling ............................................................................................ 11
2.3 Index-Based Forwarding ................................................................................. 13
2.3.1 Index-Based: Packet Forwarding Examples ............................................. 14
2.4 Naming, addressing, and routing..................................................................... 15
2.4.1 DNS - Domain Name System.................................................................... 16
2.4.2 Mobile IP Addressing ................................................................................ 17
2.4.3 Routing by Name ...................................................................................... 18
3. Architecture and processes .................................................................................. 18
3.1 Route replaces IP ............................................................................................ 18
3.2 Scenarios ........................................................................................................ 19
3.2.1 First scenario (MN and HA have a name) ................................................. 20
3.2.2 Second Scenario (reference route) ........................................................... 22
4. Implementation ..................................................................................................... 24
5. Evaluation ............................................................................................................. 26
6. Conclusion ............................................................................................................ 30
7. References ........................................................................................................... 32

1. Introduction
1.1 Motivation
„The Internet infrastructure is built on top of a collection of protocols, called the
TCP/IP protocol suite. Transmission Control Protocol (TCP) and Internet Protocol (IP) are the core protocols in this suite [1]”.
Each computer (host) is connected to the Internet should be uniquely identified by an
IP address. This raises one of the most important issues in mobility, because when a host moves to another location. This host should get a new IP address to be reachable from other partners on the network and this maybe leads to limitation problem in IP addresses.
So the motivation of our work is to use routes instead of IP addresses to reach the
Mobile Node. The computation of these routes will be depending on the name .i.e. of
(Mobile Node, Home Agent...).This new internet architecture will give us the Solution to face the limitation problem of IP addresses in currently internet architecture.

1.2 Purpose of the work
The new approach of next internet generation that we will talk about in this student research is based on the concept of source routing (index-based routing) and name based routing. All the nodes in the internet network will have the ability to get the routes that they need to reach their destination with help from the Route Service, this is responsible to compute these routes.
The basic idea behind this is to replace the IP addresses, which used by the Mobile
IP entities like Mobile Node, Home Agent, Foreign Agent to achieve an IP
Datagram’s routing operation, in other words to keep Correspondent Node and
Mobile Node in contact, with domain names, which will be resolved into Routes.
In this student research our work is to build Mobility scenarios, which will be represented by using Message Sequence Diagram. These scenarios will be only depends on the names like (Mobile Node, Home Agent...) which will be resolved into routes instead of IP addresses to allow the Correspondent Node(CN) to communicate with Mobile Node(MN).
The next step in this work is to simulate one of these scenarios by using Frogger
Simulator, which developed by the group of Integrated Communication in Technical
University of Ilmenau.

4

2. Basics
In this section we will take an overview about what is the Internet Protocol and what are the types of IP routing, what are terminologies that used in Mobile IP, and how does it work. In addition to this we will explain by example what Index-Based
Forwarding concept is and clear the difference between the names, addresses and paths. 2.1 Internet Protocol (IP)
The Internet Protocol (IP) is a protocol or method, by using this protocol data will be sent from one computer to another on the Internet network. Each computer (known as a host) on the Internet network has at least one IP address that uniquely identifies it from all other computers on the Internet.
The most widely used version of IP today is Internet Protocol Version 4 (IPv4).
However, IP Version 6 (IPv6) is also beginning to be supported [2].

2.1.1 IP Routing
2.1.1.1What is routing
In general routing is a process that forward traffic from its source to its destination across multiple networks by using a set of protocols (e.g. BGP, OSPF...), that determine the best path that data should follows. In IP network, routing based on the
IP addresses of source and destination.
Routing involves two basic activities: first finding out how to get from here to there, that’s mean determining the optimal routing path to destination and second actually getting from here to there (packet transporting); this can be done in many (very different) basic ways [3], [4].

2.1.1.2 Source vs. Hop by Hop Routing
In Source Routing concept the Source has full information about the route should the
IP datagram pass through until it reach its destination. IP traffic get through specified routers (gateways), it can’t change its route.
The reason for using Source Routing might be to avoid certain networks for performance, or administrative reasons (such as privacy, access policy).

Figure (1) Source Routing
5

Figure (1) represent a simplified explanation of Source Routing concept, we assumed that each gateway have only one address. When the datagram is moved from the source host (A) into the Internet, the Source Address (SA), Destination address (DA),
Source Route List (SRL) and Source Route Pointer (SRP) fields take these values at this moment: (SA: A), (DA: B), (SRL: C, D, E), (SRP: 0).
As the datagram leave gateway B, the fields became: (SA: A), (DA: C), (SRL: B, D,
E), (SRP: 1).
At gateway C, the processing is similar. The datagram leaves gateway C with its fields appearing: (SA: A), (DA: D), (SRL: B, C, E), (SRP: 2).
Finally, at the gateway D, The datagram leaves gateway D with its fields appearing:
(SA: A), (DA: E), (SRL: B, C, D), (SRP: 3).
There are three IP options that support source routing. In each, the option includes a list of IP addresses specifying the routers that must be used, to reach the destination
[3], [4].
Strict Source routing: this means that the route specified in the option must be used exactly from datagram, in sequence, with no other routers allowed to handle the traffic at all [3], [4].
Loose source routing: the route is specified by a list of IP addresses that must be followed in sequence, but having hops that are out of the list in between the devices on the list is permitted [3], [4].
Record Route: this option allows the datagram’s get through the network to record their path within the header for the datagram itself. When a source device sends a datagram use this option in it, router’s which “process” the datagram adds their IP addresses to this datagram. The recipient can then extract the list of IP addresses to see the route taken by the datagram [3], [4].
By hop-by-hop routing, the source don’t have all the information about how to get to the Destination; it is only know how to get to the "next hop“. That’s means routing decisions are made at each router independently and locally.
Each router belongs to the network send to the other routers its information, so each router can extract from this information its Routing table which includes the addresses of all other nodes, therefore Hop by Hop routing is also referred to as destination based table driven routing [3], [4].
In other words we can say that each node has all information about the topology of the network and it can now determine which the better route to reach the destination is. 6

Figure (2) shows that routing table on each node contains the next hop node and
Data Packet only has the destination address.

Figure (2) Hop-by-Hop routing

2.2 Mobile IP
The idea behind the invention of Mobile IP is to give devices, which are attached to the network (e.g. mobile telephone, computer...) the ability to maintain Internet
Connectivity while moving from one Internet attachment point to another with their
Permanent IP addresses.

Figure (3) Traditional Routing for a Mobile Host
Figure (3) shows that’s, When the Host moving to another network, it became unreachable. 2.2.1 Terminology in Mobile IP
Mobile node: As examples for mobile nodes are (mobile telephone and laptop computers), these devices have the ability when they move from internet attachment point to another to be reachable via their home addresses.
To keep the Connection while the mobile node moves from one network to another, it needs a special support, because the traditional internet routing assumes that the

7

mobile node has the same IP address. Therefore, each node should have new IP
Address each time it connected to new network [5].
Home Agent (HA): is a router which plays as a connection point between the mobile node and its home network (when it moves to another network) [5]. Home agent always received information from mobile node about its new location (its new IP address). Foreign Agent (FA): is a router act as a point of attachment when the mobile node has been moved to a foreign network.
Foreign agent works in conjunction with home agent to support internet traffic routing for and from mobile node [5].
Correspondent Node (CN): it’s a device or node that communicates with mobile node (it’s out of the mobile node network) [5].
Care-of Address (COA): when a mobile node moves to new network, it’s obtaining a new IP address this IP address called CoA. It’s a temporary IP address that enables mobile Node to communicate with other nodes in other networks.
There are two types of Care-of Address:
Foreign Agent Care-of Address: This is the static IP address of a foreign Agent on a visited network.
Collocated Care-of Address: which is a temporary IP address assigned to the mobile node. 2.2.2 How Mobile IP Works

Figure (4) Mobile IP Scenario
8

When IP packet exchanged over the connection between correspondent node and mobile node the following operation occur:
Correspondent node send an IP packet intended for mobile node A, with A’s home address in IP header. The IP packet is forwarded to A's home network.
When the IP packet arrive the home network, the home agent intercept this packet encapsulate it within a new IP packet that has A’s care of address in the header and then resend it to mobile node A. IP packet encapsulation within other IP packet called
Tunneling.
When the IP packet arrive at the foreign network the foreign agent strip it from its outer IP header and then send it to mobile node A. Mobile node A sends directly the
IP traffic to correspondent node by using (CN) IP address.
To support the operations illustrated in Figure (1), Mobile IP includes three basic
Functions: Agent Discovery, Registration, and Tunneling [6], [7], [8].

2.2.2.1 Agent Discovery
The first thing must the mobile node to do when it became active is to determine if it still at home or it has been moved. This can be done through a process called Agent discovery, which have three Basic Functions:
First Determine if the Mobile Node connected to foreign or home network.
Second Determine if Mobile Node moved from one network to another.
Finally obtain a care of address if it’s connected to foreign network.
There are two types of Messages are used by Agent Discovery: Agent
Advertisement, and Agent Solicitation [6], [7], [8].

Agent Solicitation

Figure (5) Agent Discovery Messages

9

Agent Advertisement message: it is similar to ICMP(Internet Control Message
Protocol) message, that is used for the regular IP router discovery process this message has one or more extensions added to ICMP message, and this extension contain Mobile-IP-specific information for mobile nodes.
As we see in Figure (5) Mobile Agent (Home agents and foreign agents) always send
Agent Advertisements messages, which are received by all Nodes on the network.
Mobile nodes listen to these Agent Advertisements and from the information that included in this message it determines whether they are stay connected to their home network or a foreign network.
If it discovers that it is stay connected to the home network, and then it behaves as any static node in this network, therefore, it does not use any mobile IP capabilities.
If it discovers that it is connected to the foreign network, then it gets a care of address from the Agent Advertisements Message [6], [7], [8].
Agent Solicitation message: When Mobile Node has been moved to the new network and has not received agent advertisements, it send Solicitation Message that forced the Foreign Agent to broadcast an agent advertisement Message as we see in Figure(5).
How Does a Mobile Node Determine That It Has Moved (Movement detection)
There are two ways to detect that the Mobile Node has been moved from its Home
Network to Foreign network.
First by using Lifetime field: in this method mobile node use the lifetime filed in agent advertisement message as a timer, when it does not receive an advertisement during this time from mobile agent, it assumes that it has moved and must try to find another network to register with.
Second by using Network Prefix: in this approach mobile node compare between its home address and Agent Advertisements’ source address, if the network prefix are the same for both of them, then the mobile node it still at home if not , that’s mean it has moved to a foreign network [6], [7], [8].

10

2.2.2.2 Registration

Figure (6) Registrations process
The purpose behind of registration process is to let mobile node to inform its Home agent about its current care of address, this address will be used by Home Agent to forward encapsulated traffic to the mobile node.
When registration about to expire mobile node renew this registration (another lifetime field), and when mobile node moves back to its home network deregistration from its care of address will be done [6], [7], [8].
The registration process includes four steps:
 The mobile node requests the forwarding service by sending a registration request to the foreign agent that the mobile node wants to use.
 The foreign agent rebroadcasts this request to the home agent of that mobile node.  The home agent either accepts or refuses the request and sends a registration reply to the foreign agent.
 The foreign agent rebroadcast this reply to the mobile node.
If the mobile node is using a collocated care-of address, then it registers directly with its home agent, rather than going through a foreign agent.

2.2.2.3 Tunneling

Figure (7) tunneling from home agent to foreign agent

11

When a mobile node has moved to foreign network and registered itself with its home agent, at this moment IP traffic sent to the mobile node will be intercepted by home agent. Home agent tells nodes on the home network that IP datagram’s to the mobile node should be delivered to home agent instead. Practically, the home agent steals the identity of the mobile node in order to capture datagram’s intended for that node that are transmitted across the home network. The home agent uses the Address
Resolution Protocol (ARP) as a possible mechanism to do this [6], [7], [8].
Address Resolution Protocol (ARP): IP addresses are virtual addresses, to send
IP traffic across the network we should translate this addresses to physical addresses this process called address resolution, therefore we use
ARP protocol which is transforming an IP address to a physical address that is acceptable in the local network.
To allow Home Agent to intercept mobile node’s traffic it should send a number of
ARP’s messages to update the ARP caches for the other nodes, which still have the original physical address for the Mobile Node in their caches. This physical address will be replaced with the physical address of the Home Agent, and then IP packets are routed to the mobile node’s care-of address via tunneling, each IP packet is encapsulated in an outer IP packet with a new header added [6], [7], [8].
Foreign Agent removes original packets from the tunnel and delivers the original packet to the mobile node over the foreign network.
If the mobile node wants to send packets, they will be sent directly to their destination without the need for tunneling. The foreign Agent acts as a router for all packets produced by mobile node.
Different types of Tunneling:
IP-within-IP: The whole original IP packet becomes the payload in a new IP packet.
Minimal encapsulation: A new header is inserted between the original IP header and the original IP payload. The original IP header is then modified to form a new outer IP header. This encapsulation technique need not be supported by all FAs and
HAs; it is optional [6], [7], [8].
Generic Routing Encapsulation (GRE): was designed to encapsulate a packet from one protocol into a network layer of a second protocol. This encapsulation technique need not be supported by all FAs and HAs; it is optional [6], [7], [8].

12

2.3 Index-Based Forwarding
It’s important to understand the difference between how index based and IP forwarded datagram’s through the network. Hop by Hop and source routing is giving an obvious example of how IP routing deals with traffic and routed it across the network, this has been mentioned before as we see in section (1).
Index based is a new forwarding concept, in this concept the link interface IDs of each node will be ordered and then in this ordering each link will be represented by an index.

IP Address

Index

124.13.23.2

1

124.13.23.3

2

124.13.23.4

3

124.13.23.5

4

Figure (8) Example of How Interfaces are Indexed
As we see figure (8) give us an example how interfaces are indexed. It shows that the node have four links ID each one is represented by an index number.
After the node interfaces have been indexed the new link ID can now be efficiently encoded as binary code. In general a global hash function is used to encode the links
IDs [9].

Figure (9) Explanation of the Index-Based Encoding
13

Figure (9) shows how the links IDs will be encoded, the number of bits that used in binary form is depending on how many links the node had [9].
As result for this new concept paths between nodes can be specified as a concatenation of encoded index. Each node in the subnet knows the network topology and recognizes all links interface ID of other nodes.
So when all links ID are represented in binary form; this node can associate this code from source to destination as binary chains. This chain is guaranteed to be unique
(there is no collision probability) [9].
The concatenation operation used is an example of a reversible or perfect hash, so the next hop information can be extracted without needing to forward table.

2.3.1 Index-Based: Packet Forwarding Examples

Figure (10) Forwarding with the Index-based PathID
In this section will explain how the packets will be forwarded from its source to its destination by using index based concept. As shown above in figure (10) there are three upgraded node (S, 2, 3) these nodes have the ability to compute the path ID.
Each node build its index table, which associate between IP address and interface link ID.
Node 2 has two interface links, so the index table has only two entries and need only two bits to represent it in binary form (10=2) (2 is the interface link).
Node 3 has five interface links, so index table have five entries and we need here 3 bit to encode the link into binary code (110=3) (3 is the interface link). Node S
14

concatenates the previous interface links together to form the path ID; that the datagram will be used it to reach its destination [9].
Node S will send the packet to node 2; this node will shift the path ID two bit right and extract the appropriate interface link from that two bit then obtain the appropriate IP address from index table.
Node 6 let the packet pass because it is not upgraded, the packet will reach node 3, which made the same steps that node 2 achieve.
At the end the path ID will be zero and continue until it reach the final destination.
We will take another example that shows all the nodes are upgraded nodes, each node build its index table. Source node computes the complete route to destination

Figure (11) Forwarding with the Index-based PathID (All nodes are upgraded)
The index that the data packet should pass through its way to destination is (1, 2, and 1) these indexes are coded in binary form. When the datagram’s pass a node the route must be changed, part of the it will fall down, which represent the index that the packet already pass through, so at the destination it will be null.
In this student research project all nodes which are used in the scenarios are not upgraded nodes. There is a Routing Service entity which is responsible to compute the routes. So the nodes only send the name of its destination and the Routing
Service will compute the route.

2.4 Naming, addressing, and routing
There is always confusion when we discuss internet terms, name, address, and route because these terms are interpenetrating between each other.
In general they are taking these definitions, which are given by John F. Shoch’s [10]:
Name: a name identifies what we seek.
15

Address: an address indicates where it is.
Route: a route identifies a way to get there.
For more details we will describe those terms in different words: a name is string of characters, which is human readable and not need to be meaningful to all user, it indicate to resource, object .etc [10].
An address: In essence, addresses are just names but they are represented in binary form, which is understandable from machine. In practice addresses used to identify an object (it should unique throughout the domain) and as locater for this object [10].
A route: A route is the specific information that used to forward data to its destination address [10].

Figure (12) Name to address to path mapping
As it’s appearing in figure (12), there is interconnection between name, address and route. Each name (e.g. www.ilnenau.de), its bind to IP address (e.g.124.190.0.4)
(name→ address) and this address will be used to find the route to reach the destination (address→ route).

2.4.1 DNS - Domain Name System
It is very hard to remember the IP addresses of all of the Web sites we access every day. We just are not that good at remembering chains of binary numbers. We are good at remembering words, and that is where domain names come in.DNS give us the allowance to use friendly names, such as www.yahoo.com, to easily locate computers and other resources on network [11].
Usually a domain name consists of two parts or more, which are separated with dot.
The first step after we type the domain name into the browser is to transform this name into IP address, and when we get this address it’s easy to request this page.
The browser ask its name server, whether has he the IP address of this domain, if not then it send this request to one of root name servers. The root server knows the
16

IP addresses for all top level domains, so the root server give us the IP address destined for COM server as example [11].
Name server asking the COM server, if he know sites IP address, COM server knows the IP address of the name servers handling the site name and when return it our name server can now contact the site name server and get the IP address intended for it [10].

Figure (13) Domain Name Hierarchy

2.4.2 Mobile IP Addressing
IP Structure: “An IP address consists of 4-bytes of data. A byte consists of 8 bits therefore we have a total of 32 bits for each IP address” [12]. The IP address can be represented using one of these three methods: Hexadecimal: C0.A8.0A.02
Binary: 11000000. 10101000. 00001010.00000010
Decimal: 192.168.10.2 to simplify things, the decimal representation is usually use.

Figure (14) Mobile IP data flow

17

As we mention before IP address is used as a locator (routing to) and identifier
(naming) for the communication end point.
In Mobile IP the addresses that used are: Home address, which is hosts home agent
IP address. Care-of address, IP address on the foreign link used to locate the host.

2.4.3 Routing by Name
In routing by name, Addresses are no more used, these addresses which are used in
IP routing as an artificial intermediaries between names and routes to perform packets forwarding and routing until they reach their destination. In figure (15) names, which means what we are seeking are mapping to routes, which means how we will reach there, packets forwarding and routing are based only on names.

Figure (15) Name to path mapping

3. Architecture and processes
The new approach of next internet generation that we will talk about in this section is based on the concept of source routing (index-based routing) and name based routing. In this internet architecture all the nodes in the internet network will have the ability to get the routes from the source to the destination with help from the Route Service, which is responsible to compute these routes.
New architecture should support mobility operations too, when a Mobile node move from its Home Agent to a foreign agent. Two different scenarios will be introduced in this section to explain how mobility problem will be solved.

3.1 Route replaces IP
In currently internet architecture the basic element that the whole network depends on is the IP address. As we said before each name (Domain Name) has a unique IP, this address will be included in the datagram’s (IP packets) and help it to reach its destination. 18

DNS (Domain Name System) play a great role and represent the first step should be taken in the routing operation, it help us to resolve each Domain Name to IP address, which is Machine-Understandable.

Figure (16) DNS resolve the name to IP address
In our future view to the internet architecture we will try to replace the IP address with
Route’s, that mean every node will get the route to reach its destination from the
Route Service depending on topology information that the Route Service hold such as(name ,index…). Route Service (RS) will replace Domain Name System, so when we write in our internet browser a name for web site the RS will give the route to the
Destination.

Figure (17) Route Service resolves the name to Route
No more IP addresses will be used in routing. Each node will be identifying bye its name, that will help to calculate the route should the data traffic go through to reach the final destination. No more limitation problem of IP addresses, that we faced in currently internet architecture. Removing of IP addresses will give us free space.
One of the challenges that we will face in the next internet generation that is based on route instead of IP addresses is how we can implement this new approach when a mobile node leaves its home Agent and moved to another network.
How the correspondent node will know the route to mobile node if it has moved. What is the information’s that we need to achieve this operation i.e. (home agent name, foreign agent name, etc...). These entire questions should be answered in the next section. 3.2 Scenarios
Figure (18) shows all the messages that transmitted between all entities, which are shared in mobile IP scenario in MSC (Message Sequence Chart) diagram. First the
Mobile Node and the Foreign Agent exchange Advertisement message and
Solicitation message then the Mobile Agent Register itself with the Home Agent.
Finally the Correspondent Node send the Datagram’s to Mobile Node.
19

Figure (18) MSC representation of normal Mobile IP Scenario
So how the Mobility operation will take under the using of name based routing. After the Mobile Node has been moved to a new network, our basic goals in this situation to keep the connection between the Correspondent Node and the Mobil Node, to do this job we should know the route that allow the Correspondent Node (CN) to reach the Mobile Node (MN) in its new position.
To bind Correspondent Node with Mobile Node we have suggested three scenarios.
In the first one, each of the Mobile Node (MN) and Home Agent (HA) have a name, the second one, only the MN has a name, but we used additionally a reference route, which is the route between the HA and MN. In the third scenario we have used the
HAMultiplexer (this entity will be explained in the third scenario) this HAMultiplexer will have no name.

3.2.1 First scenario (MN and HA have a name)
In this scenario both of Mobile Node and Home Agent are identify by their names.
Figure (19) shows how the (MN), (HA), and (CN) will exchange the messages between of them to achieve Mobility Scenario.
First of all the nodes in the network should be registered with its names in the Route
Service, so it can easily compute the Routes between these entities. As example for that in our MSC scenario the Mobile Node register its name in the Route Service
(RS).
20

Figure (19) a MSC representation for Mobility Scenario, when both of MN and HA have a name
The second step should the Mobile Node take is to send a route request to the Route
Service to obtain the route between the Mobile Nodes Old Position and the Home
Agent. So when the MN get the route to the (HA), it send a message to the (HA), which can get and store from this message the return Route. This route will be called the OldRouteToMN.
When the mobile node has moved to another network it sends immediately a route request to the Route Service. This request includes the Home Agent Name, so the
Route Service computes the route between (MN) New Position and (HA) and sends it again to (MN).
When the Mobile Node gets this route it will be able to send a registration request to the Home Agent. The return route, which will be calculated through the journey of registration packet to the Home Agent will be stored in the (HA) and it will be used to reach the Mobile Node. This route will be called NewRouteToMN.
In other hand when randomly any correspondent node wants to communicate with mobile node. It sends the Mobile’s Node Name to the Route Service to get the route to the (MN), and then with this route can it reach to the MN in the Old Position. So
21

because of that the Home Agent should intercept CN Datagram’s and make a route mapping with stored return route (NewRouteToMN) to allow this Datagram to reach the (MN) in the new position.
This mapping operation will take place after a Comparison between the
RemainRoute, which is the route that allow CN Datagram’s to reach (MN) in Old
Position, and the OldRouteToMN, That we mentioned earlier.
If the two routes are equal the Home Agent will know that the (CN) Datagram’s want to reach (MN) in the New Position, so it will swap the RemainRoute with
NewRouteToMN and the (CN) Datagram’s will reach its destination.
In this scenario each Mobile Node should store the name of the Home Agent that’s means we need a more memory to do this. In other hand this scenario will be used with all network topology.

3.2.2 Second Scenario (reference route)
In this scenario only Mobile Node is identify by a name. Figure (20) shows how the
(MN), (HA), and (CN) will exchange the messages between them to achieve a
Mobility Scenario.
The question is how the Home Agent will get the name of the Mobile Node. This issue can be handled in several ways but these solutions will have the same problem, which is the need for large memory to allow the Home Agent to store the
MN names.
One of These solutions is that the Home Agent will be part of the registration process in the Routing Service for each MN belongs to it. When the MN wants to register itself in the Routing Service it sends first its name to the Home Agent which will hold this name for a short time.
The HA will use it to complete the MN registration process in RS (Route Service) and to get the reference route. RS will response to the registration message by sending a reference route to HA. HA will send this route to OldMN and at the end of this the HA can remove the MN name to reduce the size of used memory.
When the Mobile Node has moved to another network, it sends a route request
(include its name) to the Route Service to get this route between its new position and old position.
With some calculation based on the Reference Route the Mobile Node has the ability to compute the route to the Home Agent. After that the MN will be able to send a registration request to the Home Agent and the return route, which will be calculated through the journey of registration packet to the Home Agent. This route will be stored there as a (NewRouteToMN) and then used to reach the Mobile Node.
22

Figure (20) An MSC representation for Mobility Scenario with reference route
In other hand when any correspondent node wants to communicate with mobile node. It sends the Mobile’s Node Name to the Route Service to get the route to the
MN, and then with this route can it reach to the MN in the old position. So because of that the Home Agent should intercept (CN) Datagram’s and make a route mapping with stored return route (NewRouteToMN) to allow this Datagram to reach the (MN) in the new position.
This mapping operation will take place after a Comparison between the
RemainRoute, which is the remain route that allow (CN) Datagram’s to reach MN in
Old Position and the Reference Route, That we mentioned earlier. If the two routes are equal the Home Agent will know that the (CN) Datagram’s want to reach MN in the New Position. So it will swap the RemainRoute with NewRouteToMN and the
(CN) Datagram’s will reach to its destination.
In this scenario Mobile Node does not need to store the name of the Home Agent and this will be a good advantage. But the disadvantages will be that the Home
Agent will need more memory to store the names of mobile nodes and in addition to this, this scenario will be used only with specific network topology where the Home
Agent should be between the MN and RS.
23

4. Implementation
In this section we will discuss how we can take advantage of the scenarios mentioned above and turn them into an Applicable scenario. This implementation scenario will be a detailed design for the first and second scenario. In this scenario we are going to suggest using of HomeAgentMultiplexer.
So what HomeAgentMultiplexer means. This entity is derived from type (gates), which is defined in Frogger simulator classes.
The major task of this entity is to handle the packets, which are sending from the
Mobile Node and the Correspondent Node. The Home Agent Multiplexer makes a comparison and through this comparison it can decide to any Mobile Node should the data packets reach.
So if this Mobile Node is not at home the HomeAgentMultiplexer adds a route to the
Correspondent Node packets that helps it to reach MN in its new position. This process will be explained more clearly in figure (21).

Figure (21) a MSC representation for implementation Scenario with (Using
HAMultiplexer)
24

Figure (21) show us how the entities in the implementation scenario with (Using
HAMultiplexer) will exchange the messages between them to allow the
Correspondent Node packets to reach the Mobile Node in its new position.
First all the nodes should register their names in the Routing Service. Second when the Mobile Node reach to its new position it send a request to the Routing Service asking for the route to its old position (it’s simply send its name to RS).
The Routing Service will send back the route to (MN), then the Mobile Node sends
Registration Packet to Old Mobile Node and this packet will be intercept by the
HAMultiplexer which will get the return route to (MN) and use it as new route to reach
(MN).
When a Correspondent Node wants to communicate with the Mobile Node, simply it sends the name of this Mobile Node to the Routing Service and gets the route to it.
The HomeAgentMultiplexer will intercept all the packets are coming from the CN and add to them the new route that’s he got it from the Registration Packet. All CN packets will use the new route that they got it from the HAMultiplexer to reach MN in its new position.
The software structure of this implementation scenario will consist of these entities
Mobile Node, Old Mobile Node, Correspondent Node, Home Agent, and Home Agent
Multiplexer.
The first four entities the Mobile Node, Old Mobile Node and Correspondent Node and the Home Agent are derived from the Application class in the frogger simulator.
The Mobile Node Application will be responsible for sending registration packet, which will be intercepted by the Home Agent Multiplexer.
The Correspondent Node will be responsible of sending datagram packet to the
Mobile Node and this packet will be intercepted by Home Agent Multiplexer.
Home Agent Multiplexer is the important part of our software, which is represented by frogger simulator as gate; it’s derived from the Multiplexer Gate. It’s responsible of handle the registration packet and correspondent Datagram and then forwards the correspondent packet to the Mobile Node in its new position.
In addition to Mobile Node, Old Mobile Node, Correspondent Node, etc…, there is a registration packet, which is derived from the Packet class in frogger simulator.
To achieve our scenario we have made some changes to the frogger simulator especially to connect the Home Agent Multiplexer with the Home Agent, because of this we add some methods. Some changes are made too in the multiplexer gate to allow the packets to be forwarded to the Home Agent Multiplexer, which there will be handled. 25

5. Evaluation

Figure (22) Frogger GUI
In this section we will talk about the Simulator that we used in this paper. Frogger simulator is software that developed by the group of integrated communication in
Technical University of Ilmenau. Route Service is one of the important parts of frogger, which is responsible of routes computation.
Figure (22) show us how our mobility scenario looks like all the entities are connected to each other through bus objects, which are the yellow dots.
The white dots represent the Mobile Node, Correspondent Node and Home Agent, and the Old Mobile Node. The blue dot represents the Global Routing Service, which is explained in figure (23) clearly.
Each node in the routing service picture is identifying by a number generated randomly. This number will be registered in the Routing Service to help it to compute the routes. The arrows which are connecting between these nodes have an index number. Theses index are interconnected to form the routes between the nodes such
Mobile Node, Home Agent, etc….
26

Figure (23) Frogger GUI GlobalRS

Figure (24) Frogger GUI HA
27

Figure(24) shows how the Home agent look like .it consist of central multiplexer connected to interfaces ,buses which allow him to communicate with other entities in our scenario.

Figure (25) Home Agent Multiplexer attached with two MN

Figure (26) Home Agent Multiplexer after the MN1 move
The Home Agent Multiplexer is the most important part of our scenario. This gate will be responsible to handle the packets which are coming through the Home Agent to reach the Mobile Node in its old position.
Figures (25) and (26) represent the Home Agent Multiplexer connected with two mobile nodes one of them will moves to another network. So when MN1 in its new position sends a registration message this message will intercept by the Home Agent
Multiplexer.
The HAM will verify that this message is come from MN1 through a comparison operation between the routes ID that it holds (which is in our example is 17 for MN1) and the routes IDs which the packet hold. When the comparison is true the HAM get the return route from the registration packet and store it to use it later to reach MN1 in its new position.
HAM will intercept the packets sends from any correspondent node by using the same comparison method and it will add the stored route to this packet to allow it reach the MN1.
28

Figure (27) Messages Exchange
Figure (27) shows the messages exchange in our implemented scenario. The Mobile node sends from its new position a message to Routing Service this message is called Route Request to get the route to its old position.
This Route Request message includes the name of the MN and as a result of it the
MN will obtain the route to its old position from the RS. The second thing that the MN do is sending a registration message by using the route he got from RS.
This message will reach through the Home Agent to the Home Agent Multiplexer, which will handle this message and get from it the return route to reach MN in its new position. When any Correspondent Node wants to communicate with MN it send the name of
MN to the Routing Service to get the route but this route will be the route to the old position of MN. The packet which is sent from the Correspondent Node will reach the
MN Home Agent and it will be intercepted by the Home Agent Multiplexer.

29

The Home Agent Multiplexer will make a mapping by adding the route it got from the registration packet to the Correspondent Node packet to allow him to reach the MN in its new position.

6. Conclusion
There are many methods to implement Mobility scenarios; in the new internet architecture which based on the route instead of IP addresses, table (1) and (2) shows us the differences between these two methods and the Mobile IP.

Mobile IP

Mobility based on route
(Home Agent)

CN send the data traffic to MN, which is identifying by the IP address of its
Foreign Agent

CN sends its data traffic to MN, which is identifying by its Name.

In the registration process the MN send its CoA to the HA, which will intercept the
Datagram’s and redirect it to MN

In the registration process the MN send the route to reach its new position to the
HA ,which will redirect the Datagram’s to
MN

The HA is identify by its IP address

The HA should be identify by its global name, that’s mean all nodes which attached to this HA should know the HA names Table (1) difference between Mobile IP and Mobility based on route (Home Agent)

30

Mobile IP

Mobility based on route
(Home Agent Multiplexer)

CN send the data traffic to MN, which is identifying by the IP address of its
Foreign Agent

CN sends its data traffic to MN, which is identifying by its Name.

In the registration process the MN send its CoA to the HA, which will intercept the
Datagram’s and redirect it to MN

In the registration process the Home
Agent Multiplexer will intercept the registration packet which is sent from MN in its new position to MN in its old position and get from this packet the return route to MN in its new position

The Home Agent Multiplexer is identify by the return route to MN in its new position, each packet to old MN will be intercepted and return route will be added to it to reach MN in its new position The HA is identify by its IP address

Table (2) difference between Mobile IP and Mobility based on route (Home Agent
Multiplexer)

31

7. References
1. Debalina Ghosh “Mobile IP” http://www.acm.org/crossroads/xrds72/mobileip.html. 2. http://searchunifiedcommunications.techtarget.com/sDefinition/0,,sid186_gci21
4031,00.html
3. David M. Piscitello, Bellcore and A. Lyman Chapin, BBN “Introduction to
Routing” Reprinted with permission from Connexions Magazine (Volue 7, No.
9, September 1993), with permission. Connexions are a Softbank Forums publication. 4. http://www.inetdaemon.com/tutorials/internet/ip/routing/
5. Chakchai So-In, s_chakchai@yahoo.com “Mobile IP Survey Mobile IP
Terminology”.
6. http://ciscosystems.com/en/US/docs/ios/solutions_docs/mobile_ip/mobil_ip.pdf
7. http://tlclab.unipv.it/downloads/Sistemi%20di%20Trasmissione%20Radio/2008
0528/20_MobileIP_WAP.pdf “Mobile IP and Wireless Application Protocol”.
8. Charles E. Perkins Sun Microsystems “Mobile Networking through Mobile IP”. http://www.cs.colorado.edu/~rhan/CSCI_7143_002_Fall_2001/Papers/Perkins _Tutorial_Mobile_IP.htm.
9. Hema Tahilramani Kaur, Shivkumar Kalyanaraman, Andreas Weiss, Shifalika
Kanwar, Ayesha Gandhi, “BANANAS: An Evolutionary Framework for Explicit and Multipath Routing in the Internet”.
10. John F. Shoch, Inter-Network Naming, Addressing, and Routing, pp. 7279, Seventeenth
IEEE
Conference on Computer
Communication
Networks, Washington, D.C., 1978.
11. http://www.howstuffworks.com/dns.htm “How Domain Name Servers Work”
12. http://www.trainsignaltraining.com.” IP Addressing and Routing Part 1:
Invasion of IP Addresses”, Posted on November 5th, 2007 by Stelios
Antoniou.

32