Chapter 14
ATM
(ASYNCHRONOUS TRANSFER MODE)
Contents
• Background information on ATM technology
• Difference between STM & ATM
• ATM protocol
• Different switching
• ATM interfaces and connections
• ATM network architecture
• ATM type of switches
• ATM cell format, UNI/NNI format
• ATM RM & layer functions
• ATM benefits
• ATM switch architecture
• ATM services
• Underlying transmission system for ATM
Objectives
After completion of this module you will be able to:
14.1 Introduction
1st generation switches are dedicated to specific purposes such as telephony, facsimile and
low speed data transfer used circuit switched telephone network. So high-speed data
transfer over this network is not possible due to lack of bandwidth, flexibility, quality of
transmission media and equipment. Then for the purpose of high-speed data transfer,
another network called packet switched network came into existence.
ITU-T (the new avatar of CCITT) set new standards for public telecom network. In 1984,
ITU-T defined a new method called 2nd generation switch known as ISDN " a network
that provides end to end digital connectivity to support a wide range of services including
voice and non-voice service, to which users have access by a limited set of standard
multipurpose UNI". For this, 2 interfaces called BRI or BRA (192Kbps) and PRI or PRA
(2.048Mbps) are defined at the basic rate of 64Kbps. By this, maximum transmission is
restricted to 2Mbps only.
With the basic bit rate of 64Kbps, the network can offer a maximum of 1.544Mbps
(called T1 link) or 2.048Mbps (called E1 link). So, such a type of working is called N-
ISDN. However with the concept of LAN, transmission of images with good resolution
may require higher bit rates. This leads the new conception and realization of 3 rd
generation switch, based on B-ISDN. ITU-T in 1993, defines B-ISDN as "a service or
system requires transmission channels capable of supporting rates greater than PRA or
PRI".
The underlying technology that makes B-ISDN possible is ATM (Asynchronous Transfer
Mode).
Mode means specific method or way.
Transfer means transmission and switching aspects. Switching by means of Cell
Switching. Transmission by means of Primary rate of 155.52Mbps or above.
Asynchronous means information packets will be transferred based an irregular
or random occurrence pattern as they are filled according to the demand.
Hence "ATM is a method of transmission & switching of information in the form
of packets which may occur an irregular occurrence pattern as they are filled according to
the demand of the user".
14.3 STM
Y X
B Y A X
M
B A U
X
Fig-1
In the above Fig-1, even though the Cell X and B are empty, they will also be
Multiplexed and sent on the output side. By this, the bandwidth is not used effectively.
14.4 ATM
Y X
Y A
M
B A U
X
Fig-2
In the above Fig-2, the empty Cells X and B are not at all transferred towards
output side. By this, the output bandwidth is effectively used. This technique is used in
ATM switching
Packet switching technology is used.
Statistical multiplexing (another name of Asynchronous Time Division
Multiplexing) is used.
Cell Relay method is used.
ATM is the protocol designed by ATM Forum and adopted by the ITU-T. ATM
can be thought of as the “Highway” of the information Super highway.
So ATM can do every thing that N-ISDN can do but with better quality.
In ATM System, the packet size is fixed to 53 octets known as a CELL. Any type
of traffic viz Voice, Data, Video, Synchronous or Asynchronous, Short or Long packets
can be converted into ATM Cells by a process known as emulation.
So ATM can also be called as Cell relaying technology or Cell switching
technology.
Can be called as B-ISDN services switch.
Primary rate of transmission in ATM is 155.52Mbps.
Switching means creating a temporary connection between two or more devices linked to
the switch, a Hardware and/or Software devices. Traditionally, 3 methods of switching
have been important called Circuit Switching, Packet Switching and Message Switching.
Circuit switching create a direct physical connection between two devices such as phones
or computers. As in Fig-3, devices A & G are connected by the switches 1,2 and 4 via
path I and III. Circuit switching is mostly used at the physical layer of OSI Model
D
II 3
A
E
I
B 1 2
F
C III
4
G
Fig-3
For Data communication Packet switching technology was designed. User data are
packetized and sent packet by packet using the path in shared manner. Two different
approaches are available under packet switching. One is called Datagram approach and
second is called Virtual circuit approach. The latter is used in ATM.
The identifier that is actually used for data transfer in Virtual circuit approach is
called the Virtual circuit identifier. A VCI is a smaller number that only has switch scope.
It is used by a frame.
When a frame arrives at a switch, it has one VCI. When it leaves, it has another
VCI. Fig-4 shows how the VCI in a data frame changes from one switch to another
VCI VCI
Switch
Data 21 Data 88
X
Fig-4
ATM or B-ISDN offers 2 types of connections called PVC & SVC and ATM services are
connection oriented.
A source and a destination may choose to have a dedicated virtual circuit. In this case, the
corresponding table entry is recorded for all switches by the system administrator. An
outgoing VCI is given to the source and an incoming VCI is given to the destination. The
source always uses this VCI to send frames to that particular destination. The source
always uses this VCI to send frames to that particular destination. The destination knows
that the frame is coming from that particular source if the frame carries the corresponding
incoming VCI. In a simple word, PVC is like a Hotline/P Wire/ Point to Point/ Leased line
and the nature is static. Fig-5 shows the PVC setup.
31 41 51
A x x B
11
Data 14 21 Data 54
22
x
Data 34 Data 44
Incoming Outgoing
Fig-5 Port VCI Port VCI
21 34 22 44
If a source needs connection with several destinations or any other destination, it needs a
PVC for each destination which is costly. An alternative approach is the SVC. So SVC
creates a temporary, short duration connection which exists only whenever data are being
transferred by the end users. In other words, this is dynamic in nature. This approach
requires a series of action called connection setup, setup acknowledgement, data transfer
and tear down phases. ATM supports both types of connections
ATM network consists of access devices called the end points, available at user end, are
connected through a interface called UNI to the ATM switch. Another ATM switch of the
network is connected through an interface called NNI. The architecture is shown in the
Fig-6.
UNI UNI
D
NNI NNI
A
Switch Switch Switch
E
B 1 2 3
F
C ATM
Network
End End
Points Points
Fig-6
Connection between two end points is accomplished through transmission path (TP),
virtual path (VP) and virtual circuit (VC).
A transmission path (TP) is the physical connection (wire/wireless) between an end
point and a switch or between two switches.
A TP is divided into several virtual paths (VPs). A virtual path provides a
connection or set of connections between two switches.
Within a VP, many circuits called virtual circuits (VCs) will be available which is
used for connection.
Cell networks are based on virtual circuits. All cells belonging to single message
follow the same VC and remain in their original order until they reach their destination.
TP, VP and VC are shown in Fig-7.
14.15 VPI/VCI
In a virtual circuit network, to route data from one end point to another, the virtual
connection need to be identified. For this purpose, the designer of ATM, created a
hierarchical identifier with 2 levels called virtual path identifier (VPI) and virtual circuit or
channel identifier (VCI). The VPI defines the specific VP and the VCI defines a particular
VC inside the VP. Both the connection identifier are shown in Fig-8.
VC1
VC2
VP1 VC3
VP2
VP3
Most of the switches (Core switch) within typical ATM network are routed using VPI (VP
switch). (i.e) The switching can be taken place by changing the VPI but keeping VCI
within VPI intact. Such switches are called VP switch. If switching can be taken place by
changing both the VPI and VCI, then such switches are called VC switch. The switches at
end points (Edge switch) of the ATM network use both VPIs and VCIs (VC switch). Both
switches are shown in Fig-9.
VP and VC Switching
VC Switch
VCI 1 VCI 2 VCI 3 VCI 4
VCI 1
VPI 1 VPI 3 VCI 3
VCI 2
VCI 1 VCI 1
VPI 4 VPI 5
VCI 2 VCI 2
VP Switch Port 3
Fig-9
At present, rate of transmission is 155Mbps called primary rate. Higher order is also
possible in multiple of 4 times.
ATM Cell consists of 2 fields called Header Field and Information Field as in Fig-10.
Fig-10
Header field is different for UNI and NNI in the ATM network
14.18.1.1 GFC
FC ( Generic Flow Control - 4 bits)
It is used to assist the customer network in the cell flow control, but not carried
through the network.
This label identifies a particular virtual path and virtual channel or circuit on a
transmission link. The switching nodes use this information and along with the routing
information established at connecting setup, routes the cells to the appropriate output
ports. The switching nodes changes the input value of VPI/VCI fields to new output
values. Since VPI field is 8 bits (at UNI) and VCI has 16 bits field, a host can have
theoretically 256 bundles, each containing up to 65,536 circuits.
8 VPI bits provide 28 = 256 bundles
Having one of the two values ‘0’ or ‘1’, the CLP indicates priority of a cell when
the network element has to make the decision to drop the cell when its throughput
bandwidth exceeds its transfer rate.
In congestion situations, cells with CLP =1 may be dropped and not transferred at
all.
14.18.1.4
It identifies the payload type i.e. whether the cell payload contains user data or
network information and also provides congestion identification.
HEC code detects and corrects a single bit error or detects multi bit errors in the
header field. It is based on CRC-8 with the devisor polynomial as X8+X2+X+1.
The Information Field does not contain all the 48 octets of user data. One or two
octets are dedicated for administration and call sequence purpose.
The first octet (after the overhead bits or Header octets) consists of three sub
fields.
The first bit is known as the convergence sub layer indicator (CSI). It is used to
indicate whether the pointer is used or not.
The next three bits are sequential number (SN) from 000 to 111 used to detect the
type of cells.
The next three bits are the Sequence Number Protection (SNP). It performs error
detection on the CSI and SN sub fields.
One bit is not used at present.
The second octet is optional and is used as a pointer to mark the start of long
encapsulated messages.
48-octet information field is only scrambled.
14.19 Format
Management Plane:
All the management functions that relate to whole system are located in the
management plane, which is responsible for providing coordination between all planes.
Two types of functions i) Layer Management ii) Plane Management.
Layer Management:
1.Management functions relating to resources and parameters residing in its
protocol entities.
2.Handles specific OAM information flow for each layer.
Plane Management
Management of all the planes for its proper functions.
Control Plane
Responsible for the call control and connection control functions.
These are all signaling functions for setup, supervise and release a call or
connection.
User Plane
Deals with transport of user information, flow control and recovery from errors.
ATM standard defined 3 layers. They are from top to bottom, the AAL (ATM Adaptation
or Application Layer), the ATM Layer and the Physical Layer as in Fig-11
AAL Layer-3
AT M Layer-2
P HY SI CAL Layer-1
Fig-11
Normally the end switches use all the 3 layers while the intermediate switches use
only the bottom 2 layers as in Fig-12
ATM ATM
AAL
AAL
PH Y SI CAL PHY SI CAL
ATM
ATM
PHY SI CAL
PHY SI CAL
X X
End Switch Switch End
Point Point
ATM N/W
Fig-12
This Layer deals with issues related to physical connectivity of the transmission
medium and transmission of ATM Cells.
This layer is divided into 2 sub layers called
1.Physical Medium Dependent (PMD)
2.Transmission Convergence (TC)
Functions Of Physical Medium
It is the lowest sub layer and includes 2 functions namely
1.The PMD functions.
2.Bit timing functions.
PMD functions provide the bit transmission capability, including bit alignment.
Line coding and if necessary, electrical/optical conversions is performed by this layer. In
many cases PM will be an OFC. Other media such as coaxial and twisted pair cables are
also possible. The transmission functions are medium specific.
Bit timing functions are the generation and reception of waveforms suitable for
the medium, insertion and extraction of timing information, and line coding if required.
The TC sub layer performs 5 functions namely
1.Transmission frame generation & recovery.
2.Transmission frame adaptation is responsible for all actions to adapt the Cell
flow according to the payload structure of the transmission system (interface). Two
interfaces are defined namely (1) SDH based interface or Byte structured interface and
(2) Cell based interface.
Under SDH based interface, 155.520Mbps (STM-1) & 622.080Mbps (STM-4)
rates are recommended for UNI.
3. Cell delineation is the process, which allows identification of the Cell
boundaries.
4. HEC sequence generation/verification. This is the value for the 1st 4 octets of
the Cell header and inserts the result in the 5th octet HEC field. This is capable of
detecting and correcting single bit error & detecting certain multiple-bit errors.
5.Cell rate decoupling. The insertion & discarding of idle Cells is called Cell rate
decoupling.
It deals with flow issue of ATM Cells, Cell header related and path related issues.
Functions of ATM Layer
This layer is above the Physical Layer
This layer has got four functions:
1.Cell multiplex/demultiplex. VC and VP are multiplexed and demultiplexed.
2.VPI and VCI translation.
3.Cell header generation/extraction.
4.Generic Flow Control.
This layer lies between ATM Layer and Higher Layer. It has two functions.
1. Segmentation And Reassembly (SAR)
2. Convergence Sub layer (CS)
1) Service Specific Convergence Sub layer (SSCS)
2) Common Part Convergence Sub layer (CPCS)
AAL can be classified by four methods namely
1. Based on Timing. Timing between source and destination required or not
required. Real time services like voice & video required timing syn where as non-real
time services like data transfer not required syn.
2. Based on Bit rate. Bit rate constant or variable. Switched speech has CBR
where as packet transfer has VBR.
3. Based on Connection. Connection oriented or not.
4. Based on Services offered. 5 layers called AAL1 to AAL5.
Knockout switch, cross bar switch or single stage switch, shared memory switch, shared
medium switch, fully interconnected switch, space division switch, banyan switch or
multi stage switch, batcher-banyan switch and sunshine switch are the different type of
ATM switches. Batcher-Banyan switch is widely used
The basic function of the ATM switching system is to route the cells from the input port
to the appropriate output port of the switch. The ATM switching system must contain the
function defined by the U-Plane, C-Plane and M-Plane of the B-ISDN PRM in addition
to relaying of cells. Also the ATM system should support & implement the traffic control
function based on ITU-T & ATM-F recommendations.
All these functions are distributed within the ATM system switch architecture as
in the Fig-13 with the following functional parts:
1) Input Modules (IMs)
2) Cell Switch Fabric (CSF)
3) Output Modules (OMs)
4) Connection Admission Control (CAC)
5) System Management (SM)
6) Muliplexer/Demultiplexer (Optional)
Non IM M
U OM
Native
ATM IM X
Sw itching
N etw o rk
(CSF )
Native IM OM
ATM
IM OM
Control
Fig-13 CAC SM
CAC
The signalling/control information is routed to CAC through CSF or from IM
directly. It performs the connection admission discussion and resource allocation for all
connections in the switch
SM
It is responsible for managing the entire switching system. It includes fault
management, performance management, configuration management, security
management, accounting management and traffic management by means of congestion
control. Also responsible to support Interim Local Management Interface for each UNI.
Mux/Demux
It is an optional item. It will be available only if non-native ATM devices are to be
interconnected with ATM switch.
Numbering Convention
It is defined as per ITU-T recommendation I-361 which says that:
Octets are sent in increasing order starting with octet 1. Therefore the header field
will be sent 1st followed by the information filed.
Bits within an octet are sent in decreasing order starting with bit 8.
So, for all fields, the 1st bit is the MSB.
CELL
A Cell is a block of fixed length. It is identified by a label at the ATM layer of the
B-ISDN PRM.
A TM
N ode
A TM
ATM SDH N od e
No de R I NG
ATM
N od e Fig-14
14.31 Conclusion
The key to efficient utilization of the ATM networks is the integration of multiple
services over a common infrastructure. Traffic management with QoS plays a significant
role. ATM is going in a big way to play in different flavor like BB etc. To support this,
various native ATM & as well as non native ATM protocols are defined.