Address Resolution Problem

Two types physical address

• Ethernet:- Large,fixed physical address ,Address resolution is difficult
• proNET :-small easily configured

Resolution through direct mapping
proNET
When installing an interface board in a computer,

• proNEt use small integers for physical address
• Allow users to choose hardware address

One has the freedom to choose both IP and physical address
One can assign an IP address like 172.5.48.3 with the host id portion 3
One assign IP address with the host ID portion equal to 1,2,3 and so on
When installing network interface hardware selects a physical address that corresponds to the IP address
eg:- 192.5.48.3
The system administrator would select physical address 3 for computer with 192.5.48.3 becuase it is a class c address with the host id portion 3 ,Mapping is easy to maintain ,because it can be performed without reference to any external data.
Extracting physical address is extracting host portion of IP
A new computer can be added to the network without changing existing assignments or compiling code
resolving IP address IA  means computing
Pa =f(IA) 

Resolution through Dynamic Binding
Address Resolution is difficult for some networks like Ethernet technology
here 48 bit physical address is assigned during manufacturing the device ,when hardware fails and required to replace the Ethernet interface the machine address changes more over it is 48 bit address ,so no hope it can be enconded to 32-bit address
Designers of TCP/Ip protocol found a creative solution to Address resolution problem for network like Ethernet having broadcast capability
Solution allow new host or routers to be added to the network with out recompiling the code and does not require to maintain centralized database

Address Resolution Protocol 
Designers chose to use a low level protocol to bind address dynamically.termed ARP
this also provide both reasonably efficient and easy to maintain
Working of Dynamic ARP
When A host want to resolve IP address IB
It broadcast a special packet asking the host IB to respond with its physical address PB

All host receives the packet  including B
But only host B recognize its IP address and sends a reply that contain its physical address PB

When A receive the reply ,it uses the physical address to send the internet packet directly to B

ARP CACHE
ARP cache time out
To reduce communication cost,computer use ARP maintain a cache of recently required IP to physical address binding
When transmitting packet it looks in its cache for binding, before sending ARP request

SOFT STATE
A technique used in network protocol ,information can be fried out without warning,because ETHERNET does not have a guaranteed delivery ,it normally set to 20 minutes expiry time
ARP with other protocol
ARP as part of the physical system and not as part of the internet protocol

ARP IMPLEMENTATION

ARP is divided into two
first part

• maps IP to physical address

Second part

• Answer request from other machine

Steps

• 1.check if IP and physical adress are in its cache
• 2.if not send ARP request (set n timer)
• 3.reply may delyed
• 4.reply may destroy
• 5.target host may crashes
• 6.send same packet after time expires if reply is not arrived
• 7.extract IP and physical address from both request and reply
• 8.Check the request containing target host if then prepare a reply

ARP encapsulation and identification

Frame header contain a field to identify ARP message type
Frame data contain ARP message which contain ARP request or ARp reply ,physical address IP etc

ARP PROTOCOL FORMAT

ARP messages does not have a fixed format header .
The design allow ARP to map arbitrary high level protocol address to an arbitrary network address.
In practice ARP is only used to map 32 bit IPV4 address to 48 bit ETHERNET address

• Operation type field contain:- ARP req ARP reply RARP req RARP reply
• Hardware type filed:- I for ethernet
• Protocol address field:- 080016 for IP address ,type of higher level protocol address
• HLEN and PLEN allow ARP to be used arbitrary networks

when making a request it swap the sender and target ( IP , hardware )pair and operation field to reply

Q.II.Summarize the address resolution through dynamic binding using ARP. Outline the
steps involved in an ARP process.(15 marks)[B. Tech. Degree VII Semester Examination November 2012 CS/IT 703 ADVANCED COMPUTER NETWORKS (2006 Scheme)]