Chapter 8: Data Communications Protocols

A communications protocol defines the rules for sending blocks of data (each known as a Protocol
Data Unit (PDU)) from one node in a network to another node. Protocols are normally defined in a
layered manner and provide all or part of the services specified by a layer of the OSI reference
model. A protocol specification defines the operation of the protocol and may also suggest how the
protocol should be implemented. It consists of three parts:
1. Definition of Protocol Control Information (PCI) format which forms the PDU header
2. Definition of procedures for transmitting and receiving PDUs
3. Definition of services provided by the protocol layers

A Protocol Data Unit

The PDUs exchanged have two parts: a header (also known as the Protocol Control Information
(PCI)) and a payload (also known as a Service Data Unit (SDU)). The protocol does not define or
constrain the data carried in the payload part. It does specify the format of the PCI, defining the
fields which are present and the way in which the patterns of bits are to be interpreted.
A protocol also defines the procedures which determine how the PDU will be processed at the
transmit and receive nodes. The procedures specify the valid values for the PCI fields, and the
action be taken upon reception of each PCI value (usually based on stored control information).
Examples of procedures which are implemented in protocols include:
 error recovery (e.g. the checkpoint procedure, the go-back-n procedure)
 flow control
 segmentation
 service access point selection
 connection management
The documents which define a protocol procedures are usually large and are seldom concrete (i.e.
they may not be directly translated to executable programs). They typically define the actions to be
taken when a particular condition is detected, but not how the condition is to be detected. It has
been said that "Part of what makes a protocol mature is good implementation guidelines and
folklore". The success of the TCP/IP protocol suite is largely due to the "industrial strength" code
available in freely distributed reference implementations.
A protocol implemented by several processes (represented by circles) communicating using queues
of PDUs, a shared information area (shown as a rectangle) and function calls between the
processes (represented by arrows)
Protocols are usually implemented by writing a number of programs (processes) which
communicate with one another through queues and by function calls. One or more timers are also
usually required to ensure correct operation of the protocol. To start and stop timers, a protocol
normally uses an interface to the computer's operating system. This interface is also used to request
new (empty) buffers for received PDUs (or PDUs created by the layer) and to release buffers which
are no longer needed by the protocol.
Protocols are generally described using a layered architecture known as the OSI reference model.
Which abstracts the details of the protocol and allows a simple description of the service provided
by the protocol to the protocol layer above and the service required by protocol layer from the layer
below.
Examples of protocols include:
 Link Layer - HDLC, MAC, ARP
 Network Layer - IP, ICMP
 Transport Layer - UDP, TCP

Concepts & Terminology

�An example of protocol is the standard protocols developed by international agencies, such as
ITU�s HDLC and X.25.
�The simplest protocols have been established by common usage; such as a standard EIA-232
serial interface uses asynchronous protocol.
�This is universally adapted and compatible, meaning it can communicate with any other device
that is identically set up.
�Protocol Converters (or Gateway): communicates with both connecting protocols converting
their languages.
�Incompatible protocols can communicate through this.
�They communicate with DTE using that DTE�s won protocol, convert it to the network
protocol, and transport it to the distant DTE in its own language.
�compatibility of protocols is a huge problem for interconnectability of data networks.
�Data protocols can be implemented in firmware, such as a chip, software, or a combination of
both.
�Layered protocols allow developers to write software to a clearly defined interface.
�Each layer having its own function.
� An example of a layered protocol in action is in LAN standards, such as the OSI model.
�LAN standards further illustrate how the protocol is deployed.
�Interoffice Cards (NICs): Hardware vendors can build this network to connect to an of the
transmission media that LAN standards support, such as twisted-pair wire, fiber optics, coaxial
cable, and wireless.
�Firmware: The portion of the protocol that describes how the NIC communicates with the
transmission medium is implemented here.
� Logical Link Control (LLC): The card manufactures provide software drivers to enable the
functions in their cards to communicate with the network operating system across this protocol.

Some other protocols include:

� Address Resolution Protocol (ARP)
� Reverse Address Resolution Protocol (RARP)
� Routing Information Protocol: used by Unix-based computers for exchanging routing
information.
� Open Shortest Path First (OSPF) is now replacing this.

Functions
Every protocol is set up by rules.
� Data Protocols: dictate some of the same types of relationships as a Diplomatic Protocol
(which suggest who is seated next to whom, how officials of different ranks are to be addressed,
what kind of response is appropriate to another�s statement, who is introduced to whom, and other
such niceties that govern diplomatic affairs).
� In layered Protocols, functions are assigned to one layer, but the rules regarding this are
rigid.
� The internetworking protocol is likely to be either Transport Control Protocol/ Internet
Protocol (TCP/IP).
� IP is a higher-level protocol that uses an addressing scheme entirely different from that used
on the LAN.
A station has two addresses:
� A permanent firmware address
� An ad hoc IP address

Session Control

Data networks handle sessions in two distinct ways:

Connectionless: Data is launched into the network and delivered to the distant end based on its
address.
� Means that each packet or frame must contain the address of the sending and receiving
stations.
� This is what most LANs are
Connection-oriented: The devices have a physical or logical connection across the network; the
connection is set up at the start of the session and remains for its duration.
� Can be circuit-switched or virtual connection.
� Virtual connection: Defined in a software path that shares bandwidth with other sessions.
� The packets or frames typically contain a path identifier, but do not need the address of
either the sender or the receiver after the session is set up.

Communications Control
Protocols can be classified into two groups:
� Peer-to-Peer: Does not use a controller, so devices can communicate with one another at
will.
� Master-slave: In the latter protocol the master controls the functioning of the data link and
controls data transfer between the host and its terminals.
All communication goes between slaves goes through the master.
Link Management
After the session is set up, the protocol controls the flow of data across the data link.
Synchronizing
Modems exchange signal to determine the highest speed at which they can exchange data, falling
back to a lower speed if the circuit will not support the maximum.
Addressing
Every session requires an address to set up a connection if the protocol is connection-oriented or to
route packets if it is connectionless.
Routing
In data networks having multiple routes to the destination, the protocol determines the appropriate
route based on variables such as cost, congestion, distance, and type of facility.
Data Segmenting and Reassembly
A continuous data stream from the source is segmented into frames, cells, or packets as appropriate
and equipped with header and trailer records for transmission over the network.
Data Formatting
The bit stream may require conditioning before transmission and restoration after reception.
Supervision
The protocol establishes a connection, determines how the session will be started and ended, which
end will control termination of the session, how charging will be handled, and so on.
Flow Control
Protocols protect networks from congestion by sending signals to the source to halt or limit traffic
flow.
Error Detection and Correction
Protocols check for errors, acknowledge correctly received data blocks, and send repeat requests
when blocks contain an error.
� Most sophisticated protocols can acknowledge multiple packets using one or two types of
acknowledgment.
� Selective Repeat Acknowledgement: Enables the receiving device to request specific
packets to be repeated.
� Go-Back-N Method: The receiver instructs the sender to resend an errored packet and all
subsequent packets.
Failure Recovery
If the session terminates unexpectedly, the protocol determines how to prevent the application from
being corrupted.
Sequencing
If data blocks are received out of their original sequence, the protocol delivers them to the receiving
device in the correct order.
Setting Session Variables
The protocol determines such variables as whether the session will be half or full duplex, network
login and authentication, file transfer protocols that will be used, and so on.

Transport Control Protocol/Internet Protocol (TCP/IP)

TCP/IP is a collection of protocols that were developed beginning in the 1970s by the Department
of Defense as a way of providing interoperability among equipment manufacturers.� This is
designed for operation on the Internet, but it is equally adaptable to communication within a closed
network.
This is a true international standard that is administered through the Internet Engineering Task
Force (IETF), which is a voluntary body that distributes its recommendations through Internet
Requests for Comments that are open to anyone.
The Internet has four primary purposes:
� To provide electronic mail service to the users
� To support file transfer between hosts
� To permit users to log on to remote computers
� To provide users with access to information databases
The primary Application layer protocols that supports the four functions of the internet (these are all
in the same structure as VT, OSI, FTAM, and X.400):
� Simple Mail Transfer Protocol (SMTP) for e-mail
� File Transfer Protocol (FTP) for transfers
� TELNET allows users to log on a remote computer over the network and operate as if they
were directly attached
� Hypertext Transport Protocol (HTTP) is used to support database access through web
browsers.
It fits in a three-layer framework atop the physical and data-link layers.
Functions of TCP/IP
It has three-level hierarchy, lowest being:
� Subnetwork or segment: they linked together through switches or routers to comprise
domains.
� Domains: Linked by an enerprisewide internetwork.
An IP network has two types of nodes:
� Hosts: A source or destination of information such as a computer, printer, server, router, or
other addressable unit.
� Gateways: Select routes to a host based on the address, which is unique for every device.
� Today routers handle the gateway function in most networks.
� They have two types of Gateways:
� Core Gateways: have information about the structure of the network.
� Noncore Gateways: have incomplete routing information; they know the route to a core
gateway but have no knowledge of routing beyond the core.
TCP�s role is to ensure that any irregularities during the journey from source to destination are
corrected.
� All devices on the network are given addresses that correspond to their physical position in
the hierarchy.
� Open Shortest Path First (OSPF): the routing algorithm currently used in most networks in
which nodes need only know the shortest route to the destination.
� IP lacks end-to-end error checking and acknowledgment.
� TCP takes care of those functions.
� TCP has the function of disciplining an otherwise chaotic path through the Internet
� It sets up a connection at the start of a session, and terminates it at the end.
� It performs flow control by using a sliding window.

Internet Protocol (IP)

This is a packet protocol that segments data so it fits within the packet-size limitations, and
launches the packets into the network as datagrams; it routes information between devices.
� Datagrams: a connectionless and unacknowledged packet.
� Internet Control Message Protocol (ICMP): reports on delivery of IP datagrams.
� It warns users of when a destination is unreachable, and reports on how long it takes to
reach a host or when datagrams exceed their time-to-live parameters.
� Internet Group Management Protocol (IGMP): allows devices to be added and removed
from address groups that use class D addresses.
� It is unreliable, connectionless, best-effort, datagram protocol that delivers data across an
Internet.
� It routes packets and defines the rules under which host and gateways handle packets.
� It defines the basic PDU of traffic passing across an Internet, which is an IP datagram.
� This is a simple PDU that contains a 28-octet header plus a data area that can be up to
65,535 octets long.
� Maximum Transfer Unit (MTU): the maximum length permitted by a network.
� This is a laissez-faire protocol, for the Internet lacks flow control and has no way of
detecting duplicate, out-of-sequence, or lost packets.
� Therefore, every packet has a time-to-live field that has a maximum value of 255.
� Internet Control Message Protocol (ICMP): enables gateways and host to send messages
over the Internet to other gateways and hosts to do functions that lower levels handle in the OSI
model.

IP Addressing
This is both the strength and the weakness of the TCP/IP protocol. Addressing is composed of three
parts:
� Class
� Network Portion
� Host
Each IP address is made up of four classes:
� Class A Addresses: has 24 bits for hosts and 8 bits for networks, 1 bit of which identifies the
network class, leaving 7 bits for network number.
� Class B Addresses: Allocates 14 bits for host addresses and 16 bits for networks.
� Class C Addresses: Has 8 bits for hosts and 21 bits for the network address.
� Class D Addresses: Are multicast addresses.
IPv6 (IP version 6): Is an expanded addressing method that has been approved.
Internet Network Information Center (InterNIC): This is how Internet addresses are assigned.
Anyone using TCP/IP is advised to obtain an address from InterNIC.

Transport Control Protocol (TCP)

This is a connection-oriented guaranteed delivery protocol.
� It provides reliable end-to-end data delivery, and is al so responsible for sequencing, flow
control, deleting duplicate packets, and arranging delivery of missing packets.
� It is analogous to OSI�s transport layer.
� It sets up a connection at the start of a session, and terminates it at the end.
� Provides positive acknowledgment of the receipt of packets from the distant end.
� It uses a sliding window protocol to control the session between two hosts.
Address Resolution
When computers share a physical network, they are assigned a network address.
� Hosts and gateways must have a method of mapping IP addresses to physical addresses in
order to send data across an Internet.
� Address Resolution Protocol (ARP): when a host needs to know the physical address of a
station, it sends a broadcast message request in the physical address of a station with a given IP
address.
� Reverse Address Resolution Protocol (RARP): At start up, stations broadcast their physical
address, asking hosts to respond with their IP address, which is contained in a table.� Once this
initial transaction is completed, the workstation is able to respond to ARP messages.

Dynamic Address Assignment

Bootp: the TCP/IP model specifies this protocol that is intended to enable diskless PCs to request an
IP address assignment from a server.
� This eliminates the need to visit each node to enter an IP address.
� Drawback: Once it assigns an address, it has no way of unassigning it if it is not used.

Gateways
These are the key to a host�s finding its way through the Internet.
� They contain routing tables that the programmer enters or that the gateway builds by
querying neighboring gateways.
� It has detailed routing information for all directly attached networks and knowledge of
where to send traffic for remote networks.

Point-To-Point Protocol (PPP)

PPP was designed by the Internet Engineering
Task Force to route multiple protocols over dial-up and dedicated point-to-point links.
� Most dial-up Internet programs support PPP in addition to the les-effective synchronous-line
interface protocol (SLIP).
� It permits interoperability of hosts, routes, and bridges over serial links.
� The data link protocol is HDLC.
� Operating above HDLC are two higher-level protocols:
� Link Control Protocol (LCP): responsible for negotiating link options and authenticating the
link between devices.
� This can include frame compression, adjustment of frame size, and setup of link monitoring.
� Network Control Protocol (NCP): once the link is established, this provides a framework to
enable the network layer protocols to establish a connection.
� A link quality monitoring protocol provides for checking link quality during the session.
PPP has two methods of authentication:
� Password Authentication Protocol (PAP): in this option the originator sends a password; the
receiver either accepts the password or closes down the link.
� Challenge Handshake Authentication Protocol (CHAP): The host transmits a challenge that
contains a random character string to the distant station; the distant station responds with a
calculated value using a private algorithm and the receiving station�s identifier.
� The most common use is over networks that include dial-up connections as opposed to
leased-line, which is what the PPP operates with.

FTP
As we have seen, the need for network came up primarily to facilitate sharing of files between
researchers. And to this day, file transfer remains one of the most used facilities.The protocol that
handles these requests is File Transfer Protocol or FTP.

Using FTP to transfer files is helpful in these ways −

 Easily transfers files between two different networks
 Can resume file transfer sessions even if connection is dropped, if protocol is configure
appropriately
 Enables collaboration between geographically separated teams

Important Network Concepts

Networking communication is full of some very technical concepts based

on some simple principles.
 Packet: The fundamental unit of data transmitted over the Internet. When a device intends to
send a message to another device (for example, your PC sends a request to YouTube to open
a video), it breaks the message down into smaller pieces, called packets. Each packet has the
sender’s address, the destination address, a sequence number, and a piece of the overall
message to be sent.
 Hub:A simple network device that connects other devices to the network and sends packets
to all the devices connected to it.
 Bridge: A network device that connects two networks together and only allows packets
through that are needed.
 Switch:A network device that connects multiple devices together and filters packets based on
their destination within the connected devices.
 Router: A device that receives and analyzes packets and then routes them towards their
destination. In some cases, a router will send a packet to another router; in other cases, it will
send it directly to its destination.
 IP Address: Every device that communicates on the Internet, whether it be a personal
computer, a tablet, a smartphone, or anything else, is assigned a unique identifying number
called an IP (Internet Protocol) address. Historically, the IP-address standard used has been
IPv4 (version 4), which has the format of four numbers between 0 and 255 separated by a
period. For example, the domain Saylor.org has the IP address of 107.23.196.166. The IPv4
standard has a limit of 4,294,967,296 possible addresses. As the use of the Internet has
proliferated, the number of IP addresses needed has grown to the point where the use of
IPv4 addresses will be exhausted. This has led to the new IPv6 standard, which is currently
being phased in. The IPv6 standard is formatted as eight groups of four hexadecimal digits,
such as 2001:0db8:85a3:0042:1000:8a2e:0370:7334. The IPv6 standard has a limit of
3.4×1038 possible addresses.
 Domain name: If you had to try to remember the IP address of every web server you wanted
to access, the Internet would not be nearly as easy to use. A domain name is a human-
friendly name for a device on the Internet. These names generally consist of a descriptive
text followed by the top-level domain (TLD). For example, Wikepedia’s domain name is
wikipedia.org; wikipedia describes the organization and .org is the top-level domain. In this
case, the .org TLD is designed for nonprofit organizations. Other well-known TLDs include
.com, .net, and .gov.
 DNS: DNS stands for “domain name system,” which acts as the directory on the Internet.
When a request to access a device with a domain name is given, a DNS server is queried. It
returns the IP address of the device requested, allowing for proper routing.
 Packet-switching: When a packet is sent from one device out over the Internet, it does not
follow a straight path to its destination. Instead, it is passed from one router to another across
the Internet until it is reaches its destination. In fact, sometimes two packets from the same
message will take different routes! Sometimes, packets will arrive at their destination out of
order. When this happens, the receiving device restores them to their proper order.
 Protocol: In computer networking, a protocol is the set of rules that allow two (or more)
devices to exchange information back and forth across the network.