The World Wide Web’s
(WWW) primary means of communicating with every Internet-connected device at
the same time is through the protocol TCP/IP.
It is commonly referred to as the Internet
Protocol Suite. Basically, it is a
computer language that exists on every computer on every type of network. Its early development as part of the United
States Defense Advanced Research Projects Agency (DARPA) and its evolution into
its current state has enabled the Internet as we know it today. This paper will discuss the TCP/IP framework,
how it has enabled the Internet’s growth and some issues this type of network
faces. It will also touch on what is
being done as a way forward for these roadblocks.
The name itself
is a culmination of a dozen different protocols. The main protocols that are at the core of
the suite itself are Transmission Control Protocol (TCP) and Internet Protocol
(IP). TCP is the transport layer
protocol that is responsible for establishing connections and reliable data transport
between devices while IP provides addressing, datagram routing and other
internetwork functions (TCP/IP Guide, 2005).
These two protocols perform the heavy workload at layers three and four
but the entire suite requires the work of many different protocols to provide
users the functions they need.
TCP/IP was initially created as part of DARPA’s attempts
to create a research network for military and universities. This network, coined ARPAnet, was designed to
use existing technologies but had its limitations. The creators understood there would be
problems with capacity if the network scaled to a larger size. The first attempts at TCP, in which the “P”
stood for Program instead of Protocol, ran into issues by trying to do too
much. One of its pioneers, Jon Postel,
understood and provided:
are screwing up in our design of internet protocols by violating the principle
of layering. Specifically we are trying
to use TCP to do two things: server as a host level end to end protocol, and to
server as an internet packaging and routing protocol. These two things should be provided in a
layered and modular way. I suggest that
a new distinct internetwork protocol is needed, and that TCP be used strictly
as a host level end to end protocol (TCP/IP Guide, 2005).”
This observation split
the early TCP into TCP and IP. The split
enabled different protocol sets to be used to provide network-layer and
transport-layer capabilities. There are
currently other internetworking protocols but TCP/IP is the universal accepted
The Core of the Internet
provided by TCP/IP helped shape the Internet.
During the late 1980s, the Department of Defense (DOD) and many of the
U.S. Government chose to adopt Open Systems Interconnect (OSI) protocols. TCP/IP was viewed as a proprietary solution (BSD
Unix bundled it with their operating systems) that was at best temporary so the
DOD mandated all computer communications products would use OSI protocols and
TCP/IP would be phased out. Despite this
order, as the Internet grew, so did the development of TCP/IP. It would become the real open systems
interconnecting protocol suite since the OSI protocols were years away from
completion. Efforts were made to combine
OSI and TCP/IP so both suites and their functionalities could be taken
advantage. OSI applications would run
over TCP/IP (Kessler 2017).
At the elementary level, this protocol suite sets the
rules for “talking” amongst networks. It
governs Internet communications through its four-layered protocol stack. Each layer performs a specific function:
Interface Layer: This layer is designed
to operate over any underlying technology.
Two of the most notable network interface protocols maybe used where no
other protocol may be used; Serial Line Internet Protocol (SLIP) and
Point-to-Point Protocol (PPP). A remote
computer can connect to a server, then to the Internet using IP versus an
asynchronous connection like a leased line or dial-up.
Layer: Internet Protocol version 4
(IPv4) provides unreliable services because it does not guarantee delivery or
conduct packet checks or flow control.
IP addresses are 32 bits in length and are divided into network and host
subfields. To make room for different
size networks, IPv4 has address classes that are used for host addressing.
Layer Protocols: TCP/IP actually uses
TCP and User Datagram Protocol (UDP).
The higher-layer applications establish a connection using a three-way
handshake to determine and acknowledge each other’s initial sequence number
(ISN). Once established, data exchange
can occur. Then the connection is
Layer: This layer supports applications
that are the Internet (Kessler,
2017). TCP and UDP applications use
common protocols that are supported by the TCP/IP stack.
As shown, TCP/IP is not
only a stack of communication protocols but it is a combination of protocols,
applications and utilities that provides data connection over any internetwork
and underlying subnetwork using common network technologies.
Due to being the
most used protocol on the Internet, the framework has its drawbacks and
limitations. Because all of the major
networking devices that communicate over the Internet use this protocol, it is
considered the default communication standard.
A notable flaw within the framework is that it combines identity and
location in a single address using Tempered Networks. This creates a lack of security due to its
innate openness. The security
vulnerability is caused by TCP/IP’s connected devices. Since its identity and location is visible,
it can be spoofed by hackers anywhere in the world (Kaplan, 2017).
used together, TCP and IP are separate.
IP is “connectionless” and permits information to be broken into
segments called data packets. The source
of IP creates a listing of the route the packets have to take to reach their
destination. At this stage, attackers
can gain access to the source path and modify the route. Known as a source route attack, the attack
can also read the data in the packet. TCP
is a connection-based protocol and requires a formal connection between source
and destination. The source assigns
sequence numbers to the packets for reassembly at the destination. With the right kind of tools, it is possible
for an attacker to guess the numbers and hijack the transmission (Finjan,
The expansion of
internet usage has caused a shortage in address space availability especially
in densely populated countries like India and China. The limits of IP version 4 using a 32-bit
system was identified as early as 1992.
The Internet Engineering Task Force (IETF) designed a suite of protocols
known as IP version 6 as a way to transition into a larger pool of addresses. It uses a 128-bit system and will allow as
many addresses to cover every inhabitant on the earth many times over. It was designed from the ground up with many
levels of hierarchy and flexibility in addressing and routing (IPv6.com, 2006). It boasts features such as:
header format: designed to keep header
overhead to a minimum
address space: reduces the need for
address conservation (network address translation)
security: IPSEC compliance used along
the entire route
The last feature is
important in many ways. Not only does it
provide end-to-end encryption but the integrity-checking is standard to IPv6
which is available for all connections.
It makes man-in-the-middle attacks more difficult to execute.
TCP/IP is nonproprietary and can be modified to be compatible with all
operating systems to communicate with any other system. It is highly scalable and routable. It is one of the factors the Internet is a
global communication tool today. The
whole internet uses it; without this framework data communication and
internetworking of devices is not possible.
This paper discussed how important the framework is for moving data from
one place to another. It also touched on
the flaws of the architecture and explained its IP addressing successor. TCP/IP works in layers; each layer performs a
specific function; resulting in a reliable form of internetwork communication.