The data link layer is
also responsible for:
Encapsulation of network layer data packet into
In order for the
data to be transmitted to other devices, the protocol data units (PDUs) from
the upper layers or application layers of an OSI model must be passed down to
the next lower level for transmitting and eventually passed to the physical
layer as it is the only layer with hardware connection. The data is passed down
from the 7th layer where it has a header and the data to the 6th
layer where the 6th layer’s header is added to the 7th
layer’s header and data and then passed down to the 5th layer where
5th layer’s header is added to the 6th layer’s data. It
is then passed down to the 4th layer (Transport layer) which is one
of the most important places where encapsulation occurs where the virtual
circuit is set up. The two main transfer layer protocols, TCP or UDP allow data
passage to the next lower network layer, 3rd layer by attaching
TCP/UDP headers. The data stream from the above layers is broken into headers
at the transport layer which is also called ‘segment’. Each segment is sequenced
so that receiver gets the data stream in the same way as it is transmitted from
the sender. The data sometimes can be too large to be transmitted, in that case
the data is ‘fragmented’ by which an IP datagram may contain only a portion of
data received from the upper layer. And many such datagrams put together will
from the data originally sent from the above layer. The receiving device’s
transport layer reassembles the data by combining all these datagrams.
In the network layer,
the data is again encapsulated into the body of an IP message which is usually
called as IP Datagram or IP packet. This
process of encapsulating at the network layer is also called ‘Packaging’. Here a header is added to the segment
received from the above layer and then pass it down to the data link layer. The
3rd (Network) layer and the 4th (Transport) layer work
together to build the data steam to be sent to the destination, however they
cannot directly place their protocol data units on the destination and it has
to pass through or be transmitted through the lowest layer.
The data layer then
encapsulates the IP datagram or IP packet from the network layer into a frame
by adding a header which has the source MAC (media access control) address and
destination MAC address. If the destination is not on the same local network,
then the frame is forwarded to the router to be transmitted through an inter
connected network. The information in the IP datagram’s header allows
transmission across the internetwork and ensure the datagram delivery. It has
information such as the destination’s address, identification of the type of
the frame and control bits.
frames are handed to the lowest level of the OSI model, the physical layer. The
frames which are the logical representation of digits 1s and 0s, are then
encapsulated to digital symbols at this physical layer and punt on the network
medium to be transmitted to the destination.
Once the sender is done
with the transmission of the data, the receiving device at the destination
establishes synchronization using the digital signals and then extracts 1s and
0s. the 1s and 0s are built into data frames at the receiving end’s data link
layer and a Cyclic Redundancy Check (CRC) is performed to check the output
generated against the output in the received data frame from the sender. If the
information matches, then the data packet information is processed, if there is
no match then it is rejected. This process is called decapsulation or
de-encapsulation. The receiving end processes the packet and sends it to the
network layer where IP address is validated and checked against what is sent
from the sender. If the IP address matches, the segment is extracted from the
packet. It is then sent to the layer above it which is the transport layer,
where the data stream is rebuilt, and an acknowledgment of delivery is sent to
source device that each segment is received without any errors.
The data link layer also ensures error detection
and correction. In order to detect and identify the error and then to correct
it, the data link layer engages in the process where the sender device appends
an error detection code to the frame which is usually in the form of bits and then
transfers the data frame to the receiver. Once the receiver starts decoding the
data and rebuilds the frame, it recomputes the error detection code and if
their code matches the sender’s code, that means that frame is received
correctly, and it is error free.
Flow Control: The data link layer ensures flow
control between sender and receiver so that if the receiver is slower at
processing than at receiving and sender continuously sends frames, there is a
possibility of frames to get lost in that process. Data link layer makes sure both
the sender and receive are on the same pace to transmit and receive.
IEEE 802 standards are used in many
areas of computer networking like token ring, VLANs, LAN, bridging etc. Each of
the standards focus on one area of networking. Some examples are:
802.1(LAN/WAN) – It is responsible for managing the local and other networks
and ensuring the connection between them is established. It has many sun
standards in it. It is also responsible for creation of the networks and
enables configurations settings to enable transfer through multiple bridges or
802.3(Ethernet) – It provides its support to the 802.1 standard and enables
transfer with access control and collision detection by which it senses if
there are other frames being transmitted, so it stops and sends signal before
retransmitting again or resuming the transmission again.
802.4(Token Bus) – this is the concept of virtual ring created for successful
transmission of data. It is defined as the nodes carrying the token ring can only
transmit data; hence the token ring is passed across all nodes that arranged in
the form of a virtual ring. Protocols were added to this to ensure success.
802.11(Wireless LAN) – It can popularly be called as Wi-Fi creating the
possibility of many people accessing the local network if they are in the range
of the network.
As multiple devices attempt to use a medium
simultaneously, there is a high possibility of frame collisions to occur. The
protocols which are a set of rules to govern the communication are located at
each layer are responsible for passing down the data from the upper layer to
the layer below it. The data link layer houses the layer 2 protocols. The
Data-link protocols have many specific mechanisms in place to define how the
connected devices detect and identify the errors and what errors schemes they
use to recover from such errors or data collisions during transmission. These
data link protocols are responsible for communication between two nodes in the
form of bits and bytes on the same local area network or neighboring/adjacent
nodes in the wide area network. There are multiple ways to connect the nodes or
devices to transfer the data as there are the data link protocols. Based on the
technology and other factors, the data link layer uses diverse set of protocols
when encapsulating the data into frames. Data link layer cannot encapsulate
using a universal protocol for various IP datagrams or packets coming from
various WAN technologies. The factor that classifies them is the way the
devices are connected.
Networks: The devices are connected to by point-to-point links.
Broadcast Networks: The
connected devices share a common link or cable.
Transmission – Each bit or character is transmitted independent of other
bits/characters. The characters are separated by a start bit in front of each
character and a stop bit at the back of each character as there is no clock
information. The start bit signals the receiver that bits of a character are coming,
and the stop bit signals the receiver that the character has been received and
resets the receiver for the next start bit to be recognized.
of Asynchronous protocol:
is communication protocol which is half duplex and mostly used for file
transfer. It is mainly used to transmit files between two computers when they
use modem. Xmodem manipulates the data into packets that are sent by receiver.
Files are transferred one packet at a time by using ‘check sum’ and waits for
file confirmation receipt. All that has communication package has XModem
protocol. It can be implemented in both software and hardware, and used to
detect errors. The main disadvantage with Xmodem is data transmission speed.
Ymodem is successor to Xmodem and was designed to give better
performance. It is mainly used to transmit files between two
microcomputers when they use modem, and it has bigger packet size when compared
to Xmodem. Also, the wait time is less between file transmissions to confirm
the file receipt when compared to Xmodem. It process the data transfer in
batches. Data transmission speeds are faster in Ymodem than Xmodem,and it can
transmit exact file size. It also sends file name as part of file transmission
hence we are not required to file name every computers time.
Zmodem: Zmodem was designed for more
reliability. It has sliding window support for high performance. It is also
used for faster file transmission rates and error findings when compared to
Xmodem. It is similar to Ymodem where it won’t wait for file confirmation
receipt. We can able to alter or cancel the file transmission anytime using
Zmodem. It is mostly used in bulletin board systems. It provides crash recovery
when something happens.
Blast: It was one of
the most important asynchronous protocols during the 80s with features like,
error detection, retransmission and data delivery assurance etc. It allows a
two-way file or data transfer allowing for a faster communicaton.
Kermit: It is very similar
to XMODEM, but better in cases of poor connections. It allows data transfer in
a two-way direction simultaneously or half duplex by which sending and
receiving is done one after another. It uses a sliding window protocol by which
delivery acknowledgement is necessary. The packets of data are assigned a
sequential number which the receiver uses to form the data stream in the
correct order and identify the missing ones or duplicate ones, hence better
Protocols: These are categorized into two types.
Character-oriented protocols – It is also called as byte-oriented protocols.
The data frames or data packets are represented in the form of bytes or
Binary Synchronous Communication (BSC) – It
allows a half-duplex communication by which there is two-way communication
between the devices but not at the same time. It can be used with
ASCII, EBCDIC, and Six Bit Transcode. It includes control characters
which have information in the form of code words or data words. It
regulates link control by which it tries to minimize the errors and also
acknowledge each and every data transfer, it also
regulates flow control by ensuring that both the sender and
receiver are in sync and the receiver is not overwhelmed with messages and try
to detect error and correct if possible using the
error control mechanisms.
Digital Data Communications Message Protocol
(DDCMP) – It works both in asynchronous and synchronous communications. This
protocol is situated above the physical layer and is responsible for defining
on the structure, the data, the sequencing of data into blocks etc. It works on
both point to point link or multipoint links by establishing data paths for
transmission of data. It is mainly focused on integrity of the data and the
Bit-oriented protocols – The data frames or data
packets are represented in the form of bits.
High-level Data Link Control (HDLC) – The
bit-oriented protocols are in someway related to HDLC. It works both on point
to point and multi point networks in both half duplex and full duplex (two-way
communication simultaneously) way of communication. It is mainly concentrated
on establishing the reliable communication between the sender and the receiver
to transmit the data and acknowledgement of the success of the data transfer.
The data is encapsulated here into a frame where a header is attached which has
address information at the beginning of the frame and a trailer at the end of
the frame which contains cycle redundancy code for possible error detection
during the transmission. The frames at this data link layer are separated by
sequence of bits which are also known as ‘flags’.
Synchronous Data Link Control (SDLC) – It was
basically formed to replace binary synchronous protocol. It is also concerned
with successful transfer of data between the sender and receiver. It uses the
primary and secondary station model for communication where it has full control
of the link and it sends commands to other secondary stations and sometimes
respond to the commands. It works well with multi point networks but can also
be used on point to point networks.
Link Access Procedure (LAP): It is developed
originally from HDLC. It mainly focusses on transmitting error free data and
ensuring that the data is received and arranged in the correct sequence. It was
renamed later to Link Access procedure balanced (LAPD). Types of LAPs are:
Link Access Procedure
for Modems (LAPM) – It is an
error control protocol which is concerned with error free transmission of data
and retransmitting it, if there are any errors. It sends data frames with a
header added in the beginning and a trailer in the end which has the redundant
code (Cycle code redundancy) to detect errors and wait for the acknowledgment
and retransmit the data through ‘Automatic Repeat Request’ (ARQ) to ensure successful
Link Access Procedures,
D channel (LAPD) – It is concerned with
error detection and sequencing of the frames. It forms the second layer of
protocol in the D channel of ISDN (Integrated Services Digital Network) which
houses the control information and various signaling related data.
Link Access Procedure
for Frame Relay (LAPF) –
It is heavily based on SDLC. It focuses on communication standards in terms of
signaling between physical and data link layers of source computer to the
destination through routers and frame relay switches.