3rd sem atm basics for wcdma networks M.TECH ( PDF FILE )

1
ATM BASICS FOR WCDMA NETWORKS
A
Dissertation
submitted
in partial fulfillment
for the award of the Degree of
Master of Technology
in Department of Electronics & Communication Engineering
(with specialization in Digital Communication)
Supervisor: Submitted By:
Mrs. Archana Mewara Manmohan Singh Chandoliya
Asst. Professor Enrolment No.: 12E2YTDCM3XP603
( Department of Electronics & Comm. Engg.)
YIT, Jaipur
Department of Electronics & Communication Engineering
Yagyavalkya Institute of Technology
YIT Lane, RIICO Industrial Area
Sitapura, JAIPUR.
Rajasthan Technical University
February, 2014

2
CERTIFICATE
This is to certify that Mr. MANMOHAN SINGH CHANDOLIYA a student of M .Tech. in
DIGITAL COMMUNICATION ( Electronics & communication Engineering / Electronic
Instrumentation & Control Engineering ) 3rd
semester has submitted her Dissertation synopsis
entitled “ ATM BASICS FOR WCDMA NETWORKS ” under my guidance.
Mrs. Archana Mewara
Asst. Professor
Yagyavalkya Institute of Technology, Jaipur

3
CANDIDATE’S DECLARATION
I hereby declare that the work, which is being presented in the Dissertation, entitled “ATM
BASICS FOR WCDMA NETWORKS” in partial fulfilment for the award of Degree of
“Master of Technology” in Department. of Electronics & Communication Engineering with
Specialization in Digital Communication, and submitted to the Department of Electronics &
Communication Engineering, Yagyavalkya Institute of Technology, Jaipur, Rajasthan
Technical University is a record of my own investigations carried under the Guidance of Mrs.
Archana Mewara, Asst. Professor, Department of Electronics & Communication
Engineering, Yagyavalkya Institute of Technology.
I have not submitted the matter presented in this Dissertation anywhere for the award of any
other Degree/Diploma.
(Name and Signature of Candidate)
( Manmohan Singh Chandoliya )
Digital Communication
Enrolment No. – 12E2YTDCM3XP603
Counter Signed by-
Mrs. Archana Mewara
Asst. Professor

4
ACKNOWLEDGEMENT
I would like to thank my supervisor Mrs. Archana Mewara, Asst. Professor, Department of
Electronics and Communication Engineering for providing me opportunity to work under him
and his consistent direction, he has fed in my work. It‟s my privilege to acknowledge my
profound sense of gratitude to my supervisor for his comments, suggestions, encouragement and
inspiring guidance throughout the course of the dissertation work.
I also wish to extend my thanks to Prof. Dr. Vishnu Sharma, Principal, YIT to give me an
opportunity to carry out my Master of Technology program.
I also wish to extend my sincere thanks to Mr. L. N. Balai, H.O.D., Department of Electronics
and Communication Engineering for providing software and laboratories as additional facilities
to do Master of Technology.
At last but not the least I would like to place on record my sincere gratitude to faculties & staff of
Department of Electronics and Communication Engineering for providing fruitful environment
and continuous encouragement throughout the course of completion of my dissertation.
With sincere thanks from
Manmohan Singh Chandoliya
Enrolment No. 12E2YTDCM3XP603

5
TABLE OF CONTENTS
Front Page 1
Certificate 2
Candidate‟s Declaration 3
Acknowledgement 4
Contents 05-09
List of Abbreviations 10-12
List of Symbols 13-15
List of Figures 16
List of Tables 17-18
1. Abstract 19
2. Introduction 19
2.1 Modern Requirements 19
2.1.1 Today‟s Perception 19
2.2 ATM & ISDN 20
2.2.1 Broadband ISDN 20
2.2.2 B-ISDN and N-ISDN 21
2.3 ATM Technology 21
2.3.1 Legacy Networking 21
2.3.2 Combining Technology 21
2.4 The ATM Cell 22
2.4.1 Fixed Size 22
2.5 ATM Multiplexing 23
3.Standardisation 24
3.1 ATM Standards Bodies 24
3.1.1 The Beginning 24
3.1.2 ITU-T 24
4. The Physical Layer 25
4.1 ATM's Physical Layer 25
4.2 ATM Interface References 25
4.2.1 Private UNI 25

6
4.2.2 Public UNI 26
4.2.3 Public NNI 26
4.2.4 Public NNI 26
4.2.5 B-ICI 26
4.2.6 AINI 26
4.3 SDH/SONET 27
4.3.1 SDH and SONET 27
4.4 SDH Topology 28
4.5 SDH Frames 28
4.5.1 Repeater Section 29
4.5.2 Multiplexer Section 29
4.5.3 Path 29
4.5.4 Payload 29
4.5.5 Pointer 29
4.6 Mapping Cells to Frames 30
4.7 Cell Delineation 30
4.8 The Physical Implementation of SDH 31
4.8.1 Unable Media 31
5. The ATM Layer 32
5.1 ATM Layer 32
5.1.1 ATM Layer Services 32
5.2 UNI Cell Header 33
5.3 NNI Cell Header 34
5.4 Generic Flow Control (GFC) 34
5.5 Virtual Path Identifier (VPI) 35
5.5.1 Virtual Paths 35
5.6 Virtual Channel Identifier (VCI) 36
5.6.1 Virtual Channels 36
5.7 Virtual Paths 37
5.8 Reserved Virtual Connections 37
5.9 Payload Type Identifier (PTI) 38

7
5.9.1 PTI MSB (bit 3) 38
5.9.2 PTI MSB (bit 2) 39
5.9.3 PTI MSB (bit 1) 39
5.10 Congestion Control 39
5.10.1 Round Trip Time 39
5.10.2 Early Packet Discard (EPD) 40
5.10.3 Buffers 40
5.11 Cell Loss Priority (CLP) 40
5.11.1 CLP Operation 41
5.12 Header Error Check (HEC) 41
5.13 Virtual Paths and Channels 42
6. The ATM Adaptation Layer 42
6.1 The Adaptation Layer 42
6.1.1 Adaptation Function 43
6.2 Quality of Service (QoS) 43
6.2.1 QoS Service Categories 43
6.2.2 CBR 44
6.2.3 VBR-RT 44
6.2.4 VBR-NRT 44
6.2.5 ABR 44
6.2.6 UBR 44
6.2.7 GFR 44
6.3 ATM Traffic Classes 45
6.4 General Principles of Adaptation 46
6.5 Usage of the Adaptation Layer 46
6.6 AAL 1 46
6.6.1 AAL 1 Operation 47
6.7 AAL 2 47
6.7.1 AAL 2 (Contd.) 48
6.7.2 Initial AAL 2 Header 48
6.7.3 Secondary AAL 2 Header 48

8
6.8 AAL 3/4 48
6.8.1 AAL ¾ Frames 49
6.9 AAL 3/4 CS & SAR fields 49
6.9.1 AAL ¾ CS 49
6.9.2 AAL ¾ SAR 50
6.10 AAL5 (SEAL) 50
6.10.1 AAL5 Frame Format 50
6.10.2 AAL5 Trailer 50
6.10.3 AAL5 Transmission 51
7.Signaling 51
7.1 Address Formats 51
7.1.1 DCC ATM Address Format 51
7.1.2 ICD ATM Address Format 51
7.1.3 General 51
7.2 Point-to-Multipoint Connections 51
7.3 The Traffic Contract 52
8. UNI Signaling 53
8.1 UNI Signaling 53
8.2 Signaling Format 53
8.2.1 Call Establishment 53
8.2.2 Call Clearing 54
8.2.3 Miscellaneous 54
8.2.4 Point-to-Multipoint 54
8.2.5 Information Elements 54
8.3 Called Party Number (IE Example) 55
9. B-ICI, AINI, IISP 56
9.1 Broadband – Inter Carrier Interface (B-ICI) 56
9.2 ATM Internetworking Interface (AINI) 57
9.3 Interim Inter-switch Signaling Protocol (IISP) 57
10. ATM Traffic Descriptors 58
10.1 Traffic Management 58

9
10.2 Traffic Descriptor Parameters 59
10.3 Required Parameters 60
10.3.1 CBR 60
10.3.2 VBR 60
10.3.3 ABR 60
10.3.4 UBR 61
10.3.5 GFR 61
10.4 Peak Cell Rate (PCR) 61
10.5 Sustainable Cell Rate (SCR) 62
10.6 Maximum Burst Size (MBS) 62
10.7 Minimum Cell Rate (MCR) 63
10.8 Cell Delay Variation and Tolerance 64
11. Quality of Service Parameters 65
11.1 Quality of Service (QoS) Parameters 65
11.2 Cell Loss Ratio 66
11.3 Maximum Cell Transfer Delay 66
11.4 Peak-to-Peak Cell Delay Variation 67
11.5 Relating Peak-to-Peak CDV to max CTD and CLR 67
11.6 Accumulation of QoS Parameters 68
11.7 Measuring Delay Parameters (MDP) 68
11.8 Non-Negotiable QoS Parameters 68
11.8.1 CER 68
11.8.2 SECBR 68
11.8.3 CMR 69
12 . References 70-71

10
LIST OF ABBREVIATIONS
ACF Auto-Correlation Function
ACI Adjacent Channel Interference
A/D Analogue/Digital ( Converter )
AGC Automatic Gain Control
AK Authorization Key
ARIB Association of Radio Industries and Businesses
ASIC Application Specific Integrated Circuit
ASN Access Service Network
A-TDMA Advanced TDMA
AWGN Additive White Gaussian Noise
B3G Beyond 3G
BER Bit Error Rate
BPSK Binary Phase Shift Keying
BWA Broadband Wireless Access
CA Certification Authority
CCI Co-Channel Interference
CDM Code Division Multiplexing
CDMA Code Division Multiple Access
CDMA-S2000 Code Division Multiple Access Standard 2000.
C/N Carrier-to-Noise Power Ratio
C/(N+I) Carrier-to-Noise and -Interference Power Ratio
CSI Channel State Information
D/A Digital/Analogue (Converter)
DECT Digital Enhanced Cordless Telecommunications
DFT Discrete Fourier Transform
DFTS-OFDM DFT-Spread OFDM
DL Downlink
D-QPSK Differential-QPSK
DS Direct Sequence (DS-CDMA)
DSP Digital Signal Processor

11
DVB Digital Video Broadcasting
EDGE Enhanced Data for Global Evolution
EGC Equal Gain Combining
EGT Equal Gain Transmission
ETSI European Telecommunication Standard Institute
FDMA Frequency Division Multiple Access
FEC Forward Error Correction
FFT Fast Fourier Transform
FH-CDMA Frequency Hopping (FH-CDMA)
GMSK Gaussian Minimum Shift Keying
GSM Global System for Mobile Communications
HHT Hilbert Huang Transform
HT Hilbert Transform
ICI Inter-Carrier Interference
IDFT Inverse Discrete Fourier Transform
IEEE Institute of Electrical and Electronics Engineers
IFFT Inverse Fast Fourier Transform
IHT Inverse Hada-Mard Transform
ISI Inter-Symbol Interference
MA Multiple Access
MAI Multiple Access Interference
MC Multi-Carrier
MC-CDMA Multi-Carrier CDMA
MCM Multi-Carrier Modulation
MC-SS Multi-Carrier Spread Spectrum
MC-TDMA Multi-Carrier TDMA (OFDM and TDMA)
M-QAM QAM Constellation With M Points, e.g. 16-QAM
MRC Maximum Ratio Combining
OFCDM Orthogonal Frequency and Code Division Multiplexing
OFDM Orthogonal Frequency Division Multiplexing
OFDMA Orthogonal Frequency Division Multiple Access

12
P/S Parallel-to-Serial ( Converter )
PSD Power Spectral Density
QAM Quadrature Amplitude Modulation
QPSK Quaternary Phase Shift Keying
RF Radio Frequency
Rx Receiver
SC Single Carrier
SNR Signal-to-Noise Ratio
S/P Serial-to-Parallel ( Converter )
SS Spread Spectrum
SS-MC-MA Spread Spectrum Multi-Carrier Multiple Access
TDD Time Division Duplex
TDM Time Division Multiplex
TDMA Time Division Multiple Access
Tx Transmitter
UHF Ultra High Frequency
UL Uplink
WP Wavelet Packet
WP MC Wavelet Packet Based Multi Carrier

13
LIST OF SYMBOLS
A (k) Source Bit of User k
A (k) Source Bit Vector of User k
AP Amplitude of Path p
B (k) Code Bit of User k
B (k) Code Bit Vector of User k
B Bandwidth
Bs Signal Bandwidth
C Speed of Light
C (k) The Spreading Code Vector c(k)
C (k) Spreading Code Vector of User k
Cn Spatial Pre-Coding Vector
C Capacity
C Spreading Code Matrix
D (k) Data Symbol of User k
D (k) Data Symbol Vector of User k
DO Diversity
DF Frequency Diversity
DT Time Diversity
DB Decibel
DBM Decibel Relative to 1 mw
E{.} Expectation
Ebb Energy Per Bit
Ec Energy Per Chip
Es‟ Energy Per Symbol
F Frequency
F Carrier Frequency
H Channel Matrix
FD Doppler Frequency
FDf Filter Maximum Doppler Frequency Permitted in the Filter Design
FD max Maximum Doppler Frequency

14
FD p Doppler Frequency of Path p
FN Nth Sub-Carrier Frequency
F Noise Figure in dB/Feedback Information
Fs Sub-Carrier Spacing
Gll lth
Diagonal Element of the Equalizer Matrix G
G Equalizer Matrix
G[j] Equalizer Matrix Used for IC in the jth
Iteration
H (t) Impulse Response of the Receive Filter or Channel Impulse Response
H (τ, t) Time-Variant Channel Impulse Response
H (f, t) Time-Variant Channel Transfer Function
IC Size of the Bit Interleaver
K Number of Active Users
L Spreading Code Length
La Length of the Source Bit Vector a (k)
Lb. Length of the Code Bit Vector b (k)
Ld. Length of the Data Symbol Vector d (k)
M Number of Bits Transmitted Per Modulated Symbol
M Number of Data Symbols Transmitted Per User and OFDM Symbol
n (t) Additive Noise Signal
N Noise Vector
NC Number of Sub-Carriers
Nl lth
Element of the Noise Vector N
NF Pilot Symbol Distance In Frequency Direction
N Grid Number Of Pilot Symbols Per OFDM Frame
N ISI Number of Interfering Symbols
Ns Number of OFDM Symbols Per OFDM Frame
NT Pilot Symbol Distance in Time Direction
N Tap Number of Filter Taps
P(.) Probability Density Function
P{.} Probability
Pb BER

15
PG Processing Gain
Q Number of User Groups v Received Vector After Inverse OFDM
R (k) Received Vector of the k the User After Inverse OFDM
R Code Rate
Rb Bit Rate
Rect (x) Rectangular Function
R1 Element of the Received Vector
Rs Symbol Rate
S Symbol Vector Before OFDM
S (k) Symbol Vector of User k Before OFDM
SL lst Element of the Vector s
Sinc(x) Sin(x)/x Function
T Time/Number of Error Correction Capability of an RS Code
T Source Symbol Duration
Tc Chip Duration
Td Data Symbol Duration
Tfr OFDM Frame Duration
Tg Duration of Guard Interval
Ts OFDM Symbol Duration without Guard Interval
Ts OFDM Total Symbol Duration with Guard Interval
Tsamp Sampling Rate
U Data Symbol Vector at the Output of the Equalizer
Ul lth
Element of the Equalized Vector u
V Velocity
V Guard Loss in SNR Due to the Guard Interval
V Pilot Loss in SNR Due to the Pilot Symbols
W (k) Soft Decided Value of the Code bit b (k)
W (k) Soft Decided Value of the Code bit Vector b (k)
Wn Power Normalization Factor on Subcarrier n
X (t) Transmitted Signal

16
LIST OF FIGURES
Figure Number Title Page No.
2.4.1 Diagram of ATM Cell Fixed Size 23
8.3 Diagram of ATM Cell 55

17
LIST OF TABLES
Table Number Title Page No.
2.1.1 Modern Requirement 20
2.3.2 ATM Combining Technologies 22
3.1.2 ATM Standards Bodies 24
4.3.1 SDH and SONET 27
4.8.1 Usable Media physical Aspects of SDH/SONET 32
5.2 UNI ATM Cell Header 33
5.4 Generic Flow Control (GFC) 34
5.5.1 Virtual Paths Virtual Path Identifier ( VPI ) 36
5.6.1 Virtual Channels Identifier ( VCI ) 37
5.8 Reserved Virtual Connections 38
5.10.3 Buffers Congestion Control 40
5.11.1 Cell Loss Priority (CLP) 41
5.12 Header Error Check (HEC) 42
6.2.1 QoS Service Categories 45
7.3 The Traffic Contract 53
8.1 UNI Signaling 53
8.2.5 Signaling Format Information Elements 54
9.1 Broadband – Inter Carrier Interface (B-ICI) 56
9.2 ATM Internetworking Interface (AINI) 57
9.3 Interim Inter-switch Signaling Protocol (IISP) 58
10.1 Traffic Management 59
10.2 Traffic Descriptor Parameters 60
10.4 Peak Cell Rate (PCR) 61
10.5 Sustainable Cell Rate (SCR) 62
10.6 Maximum Burst Size (MBS) 63
10.7 Minimum Cell Rate (MCR) 64
10.8 Cell Delay Variation and Tolerance 65
11.1 Quality of Service (QoS) Parameters 65

18
11.2 Cell Loss Ratio 66
11.3 Maximum Cell Transfer Delay 67
11.4 Peak-to-Peak Cell Delay Variation 67
11.8 Non-negotiable QoS Parameters 69

19
1. Abstract
In contrast to TCP/IP, the Asynchronous Transfer Mode (ATM) network architecture
incorporated features for supporting real-time traffic such as voice and video in the initial
implementation. The principal aspects of ATM directed to real-time support are short, fixed-
sized packets (cells), short headers, and no link-by-link error control.* ATM is a standardized
architecture of packet-oriented transmission and switching originally proposed for a Broadband
Integrated Services Digital Network (BISDN). ATM has since been expanded in scope to
support a wide variety of service types: wideband, narrowband, busty, non-real time, and real
time. The synchronous TDM (circuit-switched) network that evolved primarily for voice services
supports the same services with external adaptations, but the adaptations come from a variety of
suppliers necessitating the need for multiple, nonintegrated, none standardized equipment and
support. ATM standardizes the wide range of services by defining quality-of-service
requirements for various traffic types. The quality-of-service parameters specifically intended for
voice services are maximum delay, delay variation, and cell loss probability.
2 Introductions
2.1 Modern Requirements
2.1.1 Today's Perception
Today's networking requirements are extensive. All the traditional services are still required plus
many more, which are all bandwidth-hungry. The Internet brings with it the promise of many
new services, and figures show that the demand for Internet bandwidth is growing at about 200%
compound per year. The demand for services such as video transmission, where a typical
uncompressed TV picture requires around 166 MHz of bandwidth, is booming. Video can now
be compressed to lower bandwidths without significant loss of quality. However, this has led to
the introduction of a number of incompatible standards. Video and voice over the Internet can be
of very poor quality. However in the commercial environment, the provision of adequate
bandwidth is all that is necessary - and ATM has been the lever to provide it. Video applications
tend to be of two separate types - real time video conferencing, and video retrieval and playback
using a video server. If a 'wish list' for the network of the future were to be drawn up by an
operator, then on it would surely appear the items above. In addition, it is clear that any new
network design must be able to cope with the large bandwidth demands of today and tomorrow,

20
and be infinitely scalable. Operators have not been traditionally good at estimating future
bandwidth demands. An easily managed network, which offers infinite scalability, is the ultimate
goal.
Also, any new network should be able to carry synchronous voice traffic - not at 64 kbit/s
level, but at first order multiplex level (E1 - 2.048 Mbit/s or DS1 - 1.544 Mbit/s) - and mix it
with variable bit rate traffic from a variety of sources, plus unspecified bit rate traffic. It should
also be able to allocate priorities, transfer data with a guaranteed quality of service (QoS) and
provide the highest quality of error-free bandwidth, whilst reducing the demands on network
management..
Modern Requirement
1 Traditional voice, data, fax etc.
2 Plus
– High-quality audio
– High-quality video
– Voice over the Internet
– HTML
– Multimedia etc.
3 Only networking standards supported
– Scalability from office through corporate to operator
– Same standards from operator down to the desktop
– Guaranteed quality of service where appropriate
Table 2.1.1 Modern Requirement
2.2 ATM & ISDN
2.2.1 Broadband ISDN
ATM technology has been adopted by the International Telecommunication Union
Telecommunication Standardization Sector (ITU-T) to support its Broadband ISDN service (B-
ISDN). B-ISDN was originally defined as a carrier-based technology with fixed (high)
bandwidth channels. Considering the lack of availability of H channels today in ISDN, imagine a
further 10 varieties of H channel and the prospect of trying to get one across national boundaries.

21
2.2.2 B-ISDN and N-ISDN
B-ISDN supports multiple bearer channels on a single physical access link like Narrowband
ISDN (N-ISDN). Unlike N-ISDN the bandwidth of these channels can be selected on an as
required basis. B-ISDN supports far more channels than N-ISDN, potentially 224 at the user-to
network link (16777216 simultaneous channels). This number of channels is not currently
supported, but provides an indication of ATM's commitment to future proofing. Signaling
Signaling in a B-ISDN/ATM service is carried in-band (on a dedicated channel) as opposed to
standard ISDN, which carries signaling on a separate data channel.
2.3 ATM Technology
2.3.1 Legacy Networking
Existing data networks are not suited to the sheer volume of data generated by video and
graphics applications, nor can they easily carry video or audio data streams. On the other hand,
networks provisioned for video and audio services are not suited to the transmission of data and
computer images. Also, the cost of leasing WAN bandwidth from operators has always
constrained network designers to ensure that the maximum possible use is made of the
bandwidth, whereas in the LAN, bandwidth can be seen as essentially 'free'. Therefore, the
designer can feel free to run as few or as many services over the circuits as suits the individual
circumstances.
2.3.2 Combining Technologies
Circuit and packet-switching technologies have been available for many years. Each has its
strengths and weaknesses. What ATM seeks to do is to exploit those strengths whilst eliminating
the weaknesses. The circuit switch's connection-orientation is used in combination with the
routing capabilities of packet switching. This is done by breaking the data into small fixed-length
packets known as 'cells', and statistically multiplexing these cells onto high-speed bearer circuits.
Connections reserve bandwidth in forward or reverse directions, or both, in order to guarantee
capacity. Therefore, there should be sufficient bandwidth in an ATM system to cater for the sum
of the demands. Transmission control is by access limitation, imposed through a connection
algorithm.

22
ATM Technology
1 Traditional Communications:
– Circuit switching
_ Inefficient for bursty data
– Packet switching
_ Unpredictable delays
2 ATM blends the best features of each system
– Establishes end-to-end connections
– Traffic is sent in packets and multiplexed on connections
– Standards-based handling allows predictable delays
Table 2.3.2 ATM Combining Technologies
2.4 The ATM Cell
2.4.1 Fixed Size
The concept of the fixed-size cell comes from work done in Australia during the development of
Distributed Queue Dual Bus (DQDB) (IEEE 802.6) which is a Metropolitan Area Network
(MAN) standard. In DQDB the unit of traffic is the slot, which is 53 octets in length and a very
close approximation to AAL3/4 (this term will be formally introduced later). The size is usually
regarded as a compromise between transmission efficiency for data (efficiency equals the
ratio of total bits to overhead bits) and delay requirements for voice and video traffic.
In the USA, echo cancellation is deployed as default (unlike Europe) and slightly larger
delays due to a larger cell size were not considered as important as transmission efficiency. Echo
is the degrading effect that occurs where there is 2 to 4-wire conversion. The effect is to reflect
part of the signal back to the sender. When this reflection occurs approximately 30 ms (or later)
after sending, it becomes noticeable and intrusive. Echo cancellation is a process that removes
the reflection, often by mixing a small inverted signal into the signal returning to the sender to
cancel the reflection. The size of the ATM cell was chosen to allow operators to transmit over
relatively long distances - round trips of 1000 km - whilst avoiding the need for expensive echo
cancellers.

23
Figure 2.4.1 Diagram of ATM Cell Fixed Size
2.5 ATM Multiplexing
ATM was designed to be able to handle voice, data and video information. The Nature of each of
these is different. The digitization of voice is based on sampling the analogue signal at 8000 Hz
and converting each sample into an 8-bit word, thus producing a data speed of 64000 bit/s (or 64
kbit/s). This data is time-sensitive and synchronous. As the bit rate does not change - either on a
single channel or on a number of multiplexed channels - it is referred to as Constant Bit Rate
(CBR). Other traffic varies in speed; compressed video or data, for example, is bursty by nature
and requires a high bandwidth, but only for a short period of time. This type of traffic is Variable
Bit Rate (VBR). ATM has devised a method of mixing both types and, at the same time,
allocating a certain amount of traffic priority through the use of an ATM adaptation layer (AAL).
This method is called ATM multiplexing. AAL is responsible for segmenting and reassembling
the information at the sending and receiving ends. Traffic whose bit rate is not known or

24
specified may also be transmitted. In this case, transmission is on a 'best effort' basis, that is, the
network tries to provide an end-to-end service but with no guarantees. This is Unspecified Bit
Rate (UBR).
3 Standardization
3.1 ATM Standards Bodies
3.1.1 The Beginnings
The roots of ATM can be traced back to the early 1980s when the PPTs were investigating the
use of Broadband ISDN for the transport of integrated data.
3.1.2 ITU-T
The Comity Consultative International Télégraphique et Téléphonique (CCITT), as it was known
then, received a proposal from British Telecom, France Telecom, Deutsche Telecom and AT&T
for a new system, which was studied over the next four years. The CCITT eventually became the
ITU-T (International Telecommunications Union), and it undertook the study of the ATM
standards, which were eventually issued in 1989.
The ITU-T is the dejure standards body, and its standards are very largely based on the
application of ATM to the public telecommunications infrastructure. LAN issues - LAN
emulation, for instance - are of no interest.
ATM Standards Bodies
Main international standards organization for ATM is the ITU-T
– ITU-T decision-making is relatively slow
– Based on absolute majority voting
Tables 3.1.2 ATM Standards Bodies

25
4 The Physical Layer
4.1 ATM's Physical Layer
The physical layer is at the lowest level of the ATM stack. It takes the full cells from the mid-
layer and transmits them over the physical medium. The ITU-T originally defined only two
speeds which should be supported by ATM: 155.52 Mbit/s and 622.08 Mbit/s. However, over
time a number of additional speeds and interfaces have evolved, going as low as E1 and as high
as the G bit/s range. The physical layer itself is subdivided into two sub-layers: the Transmission
Convergence (TC) sub layer and the Physical Medium Dependent (PMD) sub layer. These two
sub layers work together to ensure that the optical or copper interfaces receive and transmit the
cells efficiently, with the appropriate timing structure in place. ATM, being an international
transmission technology, has to be able to work with a variety of formats, speeds, and
transmission media and distances that may vary from country to country. The standardization of
the physical layer interfaces enabled just such connectivity. Single-mode fiber, multi-mode fiber,
coaxial pairs, and shielded and unshielded twisted pairs are all standardized for use in the ATM
environment. The Transmission Convergence sub layer takes care of Header Error Check (HEC)
generation and verification, cell scrambling and descrambling, cell delineation and decoupling.
The HEC is a one-byte field in the ATM cell header, which protects the header from errors. The
Physical Medium Dependent sub layer covers bit timing, line coding, the physical connectors
and signal characteristics
4.2 ATM Interface References
The diagram opposite is crucially important, it defines the location of all reference interfaces in
ATM networks.
4.2.1 Private UNI
The private user-to-network interface is the interface between an end-user and a private ATM
switch.

26
4.2.2 Public UNI
This is the interface between an end-user and a public ATM switch. Note that in this definition
the public network regards a private network as an 'end-user' and the private network is the 'user
end' of the UNI.
4.2.3 Public NNI
The public network-to-network interface is the province of the ITU. The most noticeable
problem here is the lack of clear definition of a signaling standard between public carrier
switches. The UNI defines all the interface characteristics (Physical, electrical, optical, and
management and data structures) between a host and the first ATM switch in a system.
4.2.4 Public-NNI
This describes the procedures and protocols to be used between private switches. The signaling
mechanism here is a derivative of UNI signaling and therefore P-NNI does not suffer the
standards shortfall experienced in the public NNI. The NNI specifies how switches are to be
connected together. Included here are some elements of inter-switch signaling, which are
necessary for ATM switches to be able to find routes through large complex systems.
4.2.5 B-ICI
The Broadband Inter-Carrier Interface (B-ICI), as its name suggests, is an interface between two
carrier networks. The recommendations provide a framework for the definition of service hand-
over from carrier to carrier. An example here may be how SMDS traffic carried over an ATM
network from one carrier to another is ultimately delivered to a distant SMDS network. Similar
consideration can be given to the hand-over of Frame Relay traffic.
4.2.6 AINI
The ATM Inter-Network Interface (AINI) is a work-in-progress document developed with the
aim of defining the interface between autonomous systems. The word „autonomous‟ here refers
to networks that utilize other networks but do not wish to know the details of the other networks'
implementation. The BICI could be considered an example of an AINI. There is a great deal of
work still to be done on the AINI definition.

27
4.3 SDH/SONET
Synchronous Digital Hierarchy (SDH) and its American counterpart, Synchronous Optical
Network (SONET) form the basic byte-delivery systems of B-ISDN, and hence for ATM. These
transmission systems were designed as add-drop multiplexer (MUX) systems for operators, and
the line format contains significant management bytes to monitor line quality and usage. SDH
and SONET are gradually replacing the older Plesiochronous Digital Hierarchy (PDH) systems.
PDH systems have been the mainstay of telephony switching and require significant space to
accommodate them. An advantage of SDH is the small footprint of the equipment cabinets,
leading to lower building costs, less heat generation and lower overall maintenance. The
advantages of higher bandwidth, greater flexibility and scalability make these systems ideal for
ATM.
4.3.1 SDH and SONET
As the name suggests, SDH is a byte-synchronous multiplexing system, but also has to support
the transport of plesiochronous data streams. This feature has been included in SDH primarily so
that as providers install an SDH backbone system, they can continue to support their legacy
circuits. At lower speeds (up to 155.52Mbit/s) SDH and SONET are different with SONET
providing the greatest granularity (speeds down to 51.84Mbit/s). At and above 155.52Mbit/s the
rates are the same. SONET and SDH are not interoperable however, as they use control and
alarm indication bits in different ways. These differences are not severe however, and are easily
reconciled (unlike the differences between E3 and T3 for example).
SDH/SONET
1 The basic standard defined to support ATM is:
– European/world standard
_ Synchronous Digital Hierarchy (SDH)
– American standard
_ Synchronous Optical Network (SONET)
2 The two systems are identical at transmission rates of 155 Mbit/s and above
Tables 4.3.1 SDH and SONET

28
4.4 SDH Topology
An SDH network makes use of dual contra-rotating fiber rings. Multiplexers are placed around
the ring to add and drop information. These are known as Add Drop Multiplexers (ADMs) and
they replace whole racks of the PDH multiplexers referred to earlier. The ADM can accept a
wide range of input data rates and services, including ISDN, ATM, FDDI and LAN. All PDH
rates can be interleaved. On the high-speed aggregate side, SDH transmits at the three rates of
155 Mbit/s, 622 Mbit/s and 2.4 Gbit/s. A single synchronous ADM with such a wide range of
interfaces can provide access to cater for a variety of needs - 1.544 Mbit/s, 2.048 Mbit/s, 6
Mbit/s, 34 Mbit/s, and so on. Higher rates of transmission can be accommodated by higher-rate
multiplexers. The optical interface of the ADM can be duplicated for protection; this is normally
done by providing dual contra-rotating rings operating in east-west mode, with automatic
detection of loss of signal and ring self-healing. This increases resilience, helping to keep
network costs down. With SDH and SONET it is usual to deploy two separate rings of fibre in
what is known as a dual-protection ring topology. In this configuration the secondary ring is used
to provide redundancy. In the event of a cabling failure, the traffic can be diverted to the
protection ring. The time to achieve this changeover is in the order of 20-40 ms. often, a single
spare ring is used as a protection facility for a larger number of working rings. Overhead bits in
the frame structure are used to activate the switchover process. Using these methods allows a
SONET/SDH transmission system to achieve very high levels of reliability. When the system
deployed in a local environment (even across an equipment rack), point-to-point cabling is the
normal.
4.5 SDH Frames
SDH and SONET are synchronous, octet-aligned, frame-based transmission systems. All
equipment in a SONET/SDH network share a common (atomic) clock referred to as the Primary
Reference Source (PRS). In the event of a clock distribution failure, an on-board local clock will
generate timing. When several streams of traffic are multiplexed together, octets (not bits) from
each of the sources are interleaved. All transmission on SONET/SDH systems occurs in discrete
frames. The frames are located by a recognizable framing pattern and all frames (irrespective of
size) are transmitted in 125 micro seconds. This 125 s figure betrays the telephony origin of
SONET and SDH. The frame itself has a relatively high overhead. The overhead area is in three

29
parts as shown in the diagram opposite: repeater section, line section and path. These terms relate
directly to the previous diagram.
4.5.1) Repeater Section
Information exchanged between repeaters. Information includes a parity value for the previous
frame and for an engineering voice channel. Often referred to as the F1 flow.
4.5.2) Multiplexer Section
Information exchanged between multiplexers, again engineering and parity checks and other
performance monitoring and alarm indications. Often referred to as the F2 flow.
4.5.3) Path
Further monitoring and control information between the ingress/egress points of the
SONET/SDH network. This may be the physical interface on the customer's equipment. Often
referred to as the F3 flow.
4.5.4) Payload
As the name suggests, this is where we find the user traffic. In the case of ATM cell transfer,
much of the detailed structure of the payload area can be ignored. One thing is worth noting
however: in ATM we define two more 'F' flows (4 and 5) to provide monitoring between end
points. These end-points are ends of paths and channels (more detail will be provided
later).These flows are in the form of cells injected into the user cell flows (that is in-band as
opposed to F1- F3 which, having reserved capacity for their exclusive use, are out-of-band). Also
one particular ATM adaptation layer, AAL2, defines two more „F‟ flows, F6 and F7. It is an
important characteristic of ATM that it can use one consistent set of operation and maintenance
features (the „F‟ flows) throughout the network, from one set of customer premises equipment to
another, monitoring each level.
4.5.5) Pointer
The pointer is used to adjust the emission time of the SDH payload in order to cater for the
differences in clocks between the SDH network and the source. This procedure allows extra

30
'filler' octets to be inserted into the bit stream in the event of a slow tributary signal. It also
allows extra capacity to be utilized in the event that the tributary signal is running slightly fast.
4.6 Mapping Cells to Frames
SDH and SONET present complex mechanisms for mapping many types of traffic to a frame for
high-speed long-haul transport. This traffic can include structures from the plesiochronous
hierarchy: FDDI, DQDB and ATM. How ATM cells are mapped is important for our discussion
here. The payload area of a SONET/SDH frame is not exactly divisible by 53 (the cell size) and
therefore in most cases a cell bridges two frames. This is of no concern, as the transmission
convergence sub layer at the receiver will rejoin the two parts. When cells are mapped to the
payload area they are mapped back-to-back, that is there is no delimiter octet to identify the start
of a cell.
In early definitions of the interrelationship between SONET and ATM, a mechanism was
defined whereby an octet of the path overhead (called H4) could be used to identify the first cell
boundary after the occurrence of the H4 octet. This mechanism fell into disuse when it was
pointed out that corresponding structures do not exist beyond 155.52 Mbit/s. The mechanism
was never fully defined for implementation in SDH systems. This presents a problem. Given that
SONET and SDH are octet-aligned (that is the PMD sub layer identifies octet boundaries) and
the ATM layer expects to receive cells from the physical layer, how are the cell boundaries
discovered.
4.7 Cell Delineation
Many transmission systems that are qualified for use with ATM have in-built mechanisms to
delineate cell boundaries. FDDI and Fiber Channel use a special symbol that cannot occur in the
user traffic flow. DS-1, DS-3, E-3 etc. can use an intermediate framing structure that is cell-
aligned, referred to as PLCP (Physical Layer Convergence Procedure). PLCP was developed
from SMDS/DQDB technologies.
The delineation process has two forms that function in the same way. In non framed
Transmission, the process is bit-aligned, while in a frame-based system such as SDH or SONET,
an octet-aligned process is more appropriate. With reference to the diagram opposite, we start in
the 'hunt' state. In the hunt state we use brute force to locate any cell. The method is to compute

31
the HEC value for any randomly chosen sequence of 5 bytes (we guess that this is a header). If
this test fails, then we try the next five. Given that headers are five bytes long and that cells are
53 bytes long, we are bound to find a header (and the HEC calculation will indicate
success).Keep in mind that there may coincidentally be a correct calculation on what is not a
header. In this case we wait 48 bytes and try again. After the first correct calculation we move to
pre-sync state where we again wait for 48 bytes. When this wait-and-try-again process produces
the correct answer (HEC checks good) six times in succession, then we regard ourselves to be in
sync and make the transition to sync state where we pass the now identified cells to the ATM
layer.
4.8 The Physical Implementation of SDH
Although optical interface and carrier systems are preferred for the transport of ATM cells due to
their inherently very low error rates, other interfaces are specified. Fiber is the preferred
connection for Postal, telegraph and telephone company (PTT) interfaces over WAN and MAN
connections using high performance single-mode fiber and laser devices, and for distribution
around campus sites and buildings using industry-standard 62.5/125 m multi-mode fiber.
Category 5 Unshielded Twisted Pair (UTP) is also widely accepted for ATM tail circuits,
delivering ATM to the end-user workstation and to devices such as videoconferencing codes.
LAN implementation of ATM is widely carried out using a combination of speeds and media. A
typical backbone, for example, could be constructed in mono mode fiber at 622 Mbit/s to link a
number of campus buildings together. Within each building, a tributary is extracted at 155 Mbit/s
in multi mode fiber to feed the different floors within each separate building, and then Category
5 UTP delivers ATM to the desktop at 25 Mbit/s.
4.8.1 Usable Media
Although Category 5 UTP is preferred for local ATM distribution, Category 3 UTP has been
tried and tested. Also, Category 3 has been found to be satisfactory for the lower-end speeds and
short-haul distances. However, if UTP is used, then all the cable plant, including punch-downs
and wall-face plates, must be of the same grade throughout.

32
The Electronics Industries association (EIA) provides a standard known as 568A. This
standard requires that end-to-end cabling in UTP Category 5 should be no more than 100 m, with
90 m reserved for the horizontal and 10 m combined for the patch cords.
Physical aspects of SDH/SONET
1 Fiber
_ Single mode
– Preferred connection to operator connection
_ Multimode
– Used for private ATM networks, for example, a university
2 Campus
– UTP
_ Category 5
– Used among workgroups
– To replace traditional LANs with ATM
Table 4.8.1 Usable Media physical Aspects of SDH/SONET
5 The ATM Layer
5.1 ATM Layer
The ATM layer describes ATM fundamentals such as cell structure and the use of virtual paths
and channels. Cells filled at this layer are passed to the lower physical layer for transmission.
5.1.1 ATM Layer Services
The ATM layer provides for the transfer of the Service Data Units (SDUs) over the UNI or NNI.
An ATM Service Data Unit (SDU) is 53 octets long. The structure of this cell depends on
whether it has to pass between an end system and an ATM switch, or between switches.
Connections are set up through ATM using predefined virtual circuits to establish the end-to-end
connections. The quality of the service provided to the end-user is specified in a traffic contract
between the network and the user.

33
This traffic contract itself will specify a number of items including the bandwidth to be
occupied in the forward and reverse directions, the specific ATM adaptation layer to be used for
data, the quality of service (QoS) required and other traffic parameters.
5.2 UNI Cell Header
The payload of an ATM cell is always 48 bytes, whether UNI or NNI is being used. There are
small differences between the UNI and NNI cell headers. The UNI contains a Generic Flow
Control field (GFC) which is as yet undefined, and is therefore not used.
User-Network Interface Cell Format:
1 Generic Flow Control (GFC)
2 Virtual Path Identifier (VPI)
3 Virtual Channel Identifier (VCI)
4 Payload Type Identifier (PTI)
5 Cell Loss Priority (CLP)
6 Header Error Check (HEC)
Table 5.2 UNI ATM Cell Header

34
5.3 NNI Cell Header
The ATM Forum NNI cell header drops the four-bit GFC field, allowing a larger range of
Virtual Paths to be supported.
1 Virtual Path Identifier (VPI)
2 Virtual Channel Identifier (VCI)
3 Payload Type Identifier (PTI)
4 Cell Loss Priority (CLP)
5 Header Error Check (HEC)
An end-to-end connection in ATM is sometimes known as a Virtual Channel Connection (VCC),
and consists of the Virtual Path and Virtual Channel numbers:
VCC = VPI + VCI
5.4 Generic Flow Control (GFC)
The GFC field was added at the insistence of Bellicose in 1988 when it was originally thought
that multiple ATM devices could be dropped on a single UNI. However, it is unlikely that this
function will ever be implemented. Two modes of operation have been discussed, uncontrolled
and controlled access. For equipment using uncontrolled access, the GFC field is not used and
the bits are always set to 0000 binary for transmitted cells. For equipment using controlled
access, the actual definition of the field is not yet agreed. There is, however, a definition for the
use of the GFC field. If a device receives 10 non-zero GFC fields within 30000, plus or minus
10000, cells received, it considers the other ATM device to be using controlled access and
notifies its layer management function.
Generic Flow Control
1 Locally significant only (at UNI)
– Any value will be overwritten by the switch
2 Two modes of operation:
– Controlled mode
– Uncontrolled mode
3 Currently only uncontrolled mode is defined
– Uncontrolled GFC = 0000
Table 5.4 Generic Flow Control (GFC)

35
5.5 Virtual Path Identifier (VPI)
ATM is a connection-orientated protocol which needs to use the end-station addresses only once,
at the time of call set-up. Thereafter, the ATM address is not required, as the system only needs
to attach to the cell a reminder of which conversation - voice, data or video - a cell belongs to.
This reminder, in the form of the VPI/VCI combination, is used by the look-up table in each
ATM switch, through which the cells travel, to ensure accurate cell-switching between input port
and output port. The end-station ATM addresses mentioned are covered in a later chapter.
The VPI has 8 bits available at the UNI and 12 at the NNI, giving either 256 or 4096
simultaneous VPs, over which data will travel in virtual channels (VCs). It is normal in private
ATM implementations to use a single virtual path (VPI=0), whereas operators make extensive
use of VPIs to simplify the switching of large numbers on VCs which travel within a path.
5.5.1 Virtual Paths
The values for the VPI and the VCI are unique to a specific link between two ATM sites. The
values may change when routing through a switch, but this presents no problems since the switch
will handle the cross-connection of a VPI/VCI combination on one link to a VPI/VCI
combination on the next link. But why both VPI and VCI, and not just a larger range of VCIs?
The virtual path concept originated with concerns over the cost of controlling B-ISDN networks.
The idea was to group together a number of connections sharing a common path through the
network into identifiable units (paths). Network management actions could then be applied to the
smaller groups of connections instead of the larger number of individual connections. For
example, a customer may take an ATM service from a carrier to connect many users' computer
systems to a central site host computer. In the event of a disaster the recovery would consist of
redirecting the virtual path to the disaster recovery site, rather than redirecting every individual
channel.

36
Virtual Path Identifier ( VPI )
1 Identifies this cell‟s path
2 8 bits available at the UNI
3 12 bits available at the NNI
– 256/4096 possible simultaneous paths
– Maximum number of usable bits is negotiable between user and Network
Table 5.5.1 Virtual Paths Virtual Path Identifier ( VPI )
5.6 Virtual Channel Identifier (VCI)
Each cell which originates from an ATM host and which is destined for a remote ATM host must
be identified throughout the system. The cell therefore carries a unique VCI number in the
header. This number is either assigned to it by the network manager in the case of a permanent
end-to-end connection (a PVC), or is chosen by the ATM switch at the time of call set-up by
selecting from an unassigned pool of VCIs, in the case of a switched connection (an SVC). The
VCI is unique to both port and path - a VCI of 100 on the output of a switch does not have to
support the same parameters as a VCI of 100 on the input of the same port. Virtual channels may
be unidirectional, unlike telephony circuits, where calls are always set up as bi-directional. To
establish a bi-directional call the call set-up will need to specify whether bandwidth is required in
the reverse direction. Virtual circuit identifiers may be remapped when transmitted through a
switch; this is a standard function. If this is the case, the HEC is recalculated each time.
The size of the VCI field - two bytes - allows for a total of 65535 VCIs.
5.6.1 Virtual Channels
The physical link carrying ATM cells can be visualized as being subdivided into smaller
pathways, each containing multiple smaller pathways. This way envisaging virtual circuits may
be helpful in understanding how operators apply the concept of the VPI to enable bundles of
circuits to be easily managed. The VPI can be thought of as a small multi-core cable within a
larger cable, each of the cores being a VCI. A pipe can carry many such VPIs and, if a bearer
failure requires the re-routing of VPIs, this can be carried out much more easily at the VPI level
than the VCI level. At the VPI level it is only necessary to issue a command to re-route one VPI

37
from source to destination, and all the VCIs travelling within that VPI are automatically switched
as well. This minimizes the load on control mechanisms.
Virtual Channel Identifier ( VCI )
1 Identifies this cell‟s channel
216 bits available at the UNI & NNI
– 65,536 possible simultaneous channels per path
– Maximum number of useable bits is negotiable on a per-path basis
Table 5.6.1 Virtual Channels Identifier ( VCI )
5.7 Virtual Paths
By creating a permanent virtual path through a network, multiple channels may be associated
with that path at one end. These channels will be switched through the network over the path
without having to be individually switched at each intermediate ATM switch. The virtual path
technique saves both time and resources within those switches.
5.8 Reserved Virtual Connections
The first 32 VCIs (decimal 0 to 31) are reserved for specific functions and, therefore, the first
VCI which is user-assignable is VCI 32. The ITU-T uses the first 16, and the ATM Forum has
been allocated the second group of 16. Technically speaking, only certain values of VCI are
reserved on particular VPIs. However, by popular usage carriers have adopted the convention
that VCI values 0 through 31 are reserved for all values of VPI. Given the large range of values
available, this is not seen as a problem and it simplifies implementation. Recently the
combination VP 0 VC 32 was reserved for Multi Protocol Label Switching (MPLS) control
traffic. This reserved channel will only be used by switches that have MPLS software installed.
This differs from the reserved channels (VCIs 0 to 31) in that 0 to 31 are permanently reserved,
regardless of what software is installed on the switch.

38
Reserved Virtual Connections
1 The following VPI/VCI combinations have been reserved:
– VPI = 0 VCI = 0 to 15 ITU-T
– VPI = 0 VCI = 16 to 31 ATM Forum
– VPI = ALL VCI = 1 to 5
2 In practice, Carriers regard VCIs 0 to 31 as reserved for all VPIs
Table 5.8 Reserved Virtual Connections
5.9 Payload Type Identifier (PTI)
Cell control uses the three-bit PTI and the one-bit cell loss priority. The PTI defines the
information which is carried in the cell. User data and network management cells are
differentiated. Bit 2 of the three-bit field can be used to indicate congestion in data traffic,
although this is not yet fully defined. Bit 1 is used to indicate the final cell in a stream of cells
which have been filled with higher-level packet traffic. This is used in assisting the switch to
manage output cell queue congestion. Remember that there is no flow control in ATM, so cells
will continue to arrive at a switch even if the switch cannot handle the consequent congestion. If
a cell queue is congested to the point where it becomes necessary to begin to discard cells, it is
better to discard cells from specific channels rather than randomly. To this end, if a group of
cells from one specific VCI is marked in such a way that the switch knows where the start and
end of that cell group is, and then the switch can discard all the cells from the group in order to
reduce congestion. The alternative is to discard random cells from multiple cell streams. This
will inevitably lead to increased congestion when the receiving end reassembles the data cells
into the higher level packets, checks them and discovers that data is missing. Retransmission will
then be requested from all the channels which detect missing data, which will lead to further
congestion when the repeats arrive at the switch.
5.9.1) PTI MSB (bit 3)
The most significant bit of the PTI field indicates whether the cell is a management cell (bit = 1),
or a user/signaling cell. If the cell is a management cell then the coding of the next two bits is as
follows. If bits 2 and 3 are 1, then this is a reserved code. If those same bits are 10, this is a

39
resource management cell. The two remaining codes indicate that the cell is part of an F5 OAM
cell flow. If the coding is 00, this is F5 OAM on a segment-by-segment basis. If the coding is 01
the flow is F5 OAM end-to-end.
This bit is a congestion identifier and we will address this on a separate slide.
This bit has many titles: Service Data Unit (SDU) bit, SDU-type bit, The Bit (older
documentation). Its use is to identify the end of last cell in a higher-layer (Protocol Data Unit)
PDU when the ATM adaptation layer 5 is used. Remember this bit for later, as the course
contains a section on AALs. If this bit is 1 then this cell is the last cell.
5.10 Congestion Control
Congestion occurs in an ATM network when a resource has been overbooked, when a network
failure has been encountered, or when cells fail to conform to the traffic contract. Under normal
circumstances, congestion should not occur because a Connection Admission Control (CAC)
algorithm will not allow it. Explicit Forward Congestion Indicator (EFCI) is a congestion
notification mechanism which the end-user uses to improve the utility of the connection. A cell
transiting an area of congestion may set the EFCI bit of the PTI to indicate to the destination that
there is a problem somewhere between source and destination. It is then up to the end device to
implement a protocol in order to lower the cell rate of the connection during congestion or
pending congestion. Although EFCI is widely implemented, the time between emitting a cell and
the time at which the source is notified (round trip time) of this congestion may be quite long.
5.10.1 Round Trip Time
Consider an extreme case: if the round trip time (RTT) is 0.5 seconds and the interface rate at the
source is 40Mbit/s, then 20 million bits will be sent before congestion is notified to the source.

40
5.10.2 Early Packet Discard (EPD)
EPD is a procedure that can help in cases of congestion. If the connection is established for an
end station that uses the type 5 adaptation (AAL5), then there is a bit in the header that indicates
the last cell in a 'set' of cells belonging to a higher layer PDU. There are two possible techniques:
in the event that a cell is dropped, drop the rest up to the cell marked as the last (let that pass), or
drop an entire packet from the cell immediately after the last marked (end) cell, up to and
including the next marked cell. The first technique is sometimes referred to as Intelligent Tail
Packet Discard and the second is called Early Packet Discard.
5.10.3 Buffers
The techniques described above both operate at the buffer level in the switch. If a cell is lost in a
congested condition (say output buffer has reached a high water mark) we can switch into the
Intelligent Tail Packet Discard or Early Packet Discard mechanisms and discard the cells from
the buffers. Usually the switch contains a 'back pressure' mechanism so that the cells are
discarded at the egress interface to the switch. The processes indicated here operate at the
connection level.
Congestion Control
1 Bit 2 of the PTI may be used to indicate to the destination that congestion has taken place in
the network
2 The bit is called Explicit Forward Congestion Indicator (EFCI)
3 This will occur when switches are discarding cells with CLP =1
Table 5.10.3 Buffers Congestion Control
5.11 Cell Loss Priority (CLP)
This single-bit field can be regarded as the key to internal ATM traffic management. The bit can
be set by the end-station ATM equipment but reset by ATM switches. This may occur when the
bearer circuits fail, or if the end station ATM equipment transmits at a higher rate than that
specified in the traffic contract. If the Cell Loss Priority (CLP) bit is set, the network is allowed
to discard that cell in the event of there being insufficient bandwidth for the cell to transmit it and
all other cells which are requesting bandwidth. The purpose of the bit is to identify cells that

41
should be discarded before cells that do not have the bit set. Under normal conditions the
reservation of bandwidth should obviate the need to discard cells, but it may be that a user will
choose CLP = 1 to take advantage of lower tariffs for traffic. Traffic with CLP = 0 will normally
be terrified at a higher rate.
5.11.1) CLP Operation
If a cell arrives at a switch at a higher data rate than that agreed in the traffic contract, the switch
has the right to change the CLP bit setting to indicate that the cell may be discarded. It is not
appropriate for traffic to pass at data rates higher than those contracted, as the end-station ATM
equipment is then enjoying the use of bandwidth which has not been paid for. The traffic
policing policy which decides whether cells are forwarded at a lower Priority.
Cell Loss Priority (CLP)
1 CLP operates independently on each active VPI/VCI
2 A switch may flip CLP from 0 to 1, for example, if traffic on a VPI/VCI exceeds the maximum
agreed sustainable cell rate
Table 5.11.1 Cell Loss Priority (CLP)
5.12 Header Error Check (HEC)
The payload in ATM carries no error detection or correction mechanism, relying instead on low
error-rate bearer circuits, receiving end software, or the receiving end not noticing that cells have
been lost. This latter condition only applies in real-time voice or video traffic connections, where
a small proportion of cells may go astray without being noticed. The Header Error Check (HEC)
is there to protect the entire header, including the HEC itself. It is capable of single bit detection
and correction, as well as multiple bit detection but not multiple bit correction. If a cell is
forwarded without the correction taking place, not only would the correct destination not receive
its cell but another destination may receive it instead (a misinserted cell), with unpredictable
results. The ATM source calculates the HEC based on all five bytes including itself. The receiver
at the other end of the link has two modes of operation, detection mode and correction mode.
Correction mode support is optional. When errors occur, they are typically characterized by
either single-bit errors or relatively large bursts of errors:

42
1 In the event of a single-bit error, it will be corrected. The mode of operation will then
switch to detection. If the next HEC is correct, the mode of operation goes back to correction.
2 In the event of a burst of errors, the first cell affected will be discarded. The next and
subsequent cells may or may not be discarded, depending on the length of the burst. Whether
errors occur singly or in bursts will be dependent on the physical media used. Recall from the
chapter on ATM‟s physical layer that the HEC is used for cell delineation.
Header Error Check (HEC)
1 The HEC is performed on the header only
– Supports forward correction of single-bit errors
– Supports detection of multiple-bit errors
2 Faulty cells are discarded
– At the UNI:
3 Error detection is mandatory
4 Error correction is optional
5 The HEC is generated/verified at the TC part of the physical layer
Table 5.12 Header Error Check (HEC)
5.13 Virtual Paths and Channels
Inside a node we can choose to base switching decisions on the VPI/VCI values as a pair (VC
Switching) or use the VPI value alone (VP Cross connect). In the case of VP Cross connect, the
switch can switch a group of channels (all those with the same VPI) to the same destination. This
can be a simplification in terms of resource management and can be faster. For the greatest
functionality, the switch uses the combined VPI/VCI pair. In this case, individual channels can
be 'routed'.
6 The ATM Adaptation Layer
6.1 The Adaptation Layer
The ATM adaptation layer is the highest layer in ATM. It is used to adapt traffic into an ATM
format. In other words, the AAL maps application data into the ATM 48-byte cell payloads. The

43
AAL function is performed at the edges of an ATM connection, and not within the network.
There are several different AALs which have been defined by the ITU-T and the ATM Forum.
The layer itself is further divided into two sub layers:
1 The Convergence sub layer (CS)
2 The Segmentation and Reassembly sub layer (SAR)
6.1.1 Adaptation Function
It would not be possible to use only one standard adaptation layer in ATM for all types of data
because voice, video and data have entirely different characteristics. These characteristics vary
between synchronous and asynchronous, constant and variable bit rate, real time and non-real
time. In addition there is the need to be able to adapt ATM to carry traffic which is not
connection-oriented in its original form, for example, LAN traffic emanating from a
connectionless source. ATM is also designed to be able to handle Frame Relay data, which is
packet-switched and connection-orientated. ATM must also be able to accommodate the
differing requirements of timing associations and data priorities, and allocate different qualities
of service to the traffic.
6.2 Quality of Service (QoS)
Quality of service is a major issue in information transfer. Dividing data into four separate
classes and allocating each its own QoS parameters simplifies the setting up of calls, as each type
of data can then be automatically allocated its appropriate QoS.
6.2.1 QoS Service Categories
QoS relates to the needs of the user for the particular application. A number of categories have
been defined. Voice services, for example, require a minimal end-to-end delay to minimize the
need for echo cancellers. Video applications cannot tolerate any data loss that will result in
image degradation. Data services are more tolerant of delays and variations in delays. See the
glossary at the end of this text for complete definitions of the various categories.

44
6.2.2 CBR
CBR has been defined to support constant bit rate connection-oriented traffic where end-to-end
synchronization is required. This is otherwise known as ITUT Class A performance
requirements. This service should meet the current requirements for performance comparable to
today‟s digital private line services such as E1.
6.2.3 VBR-RT
VBR-RT has been defined to support variable bit rate connection-oriented traffic where end-to-
end synchronization is required. This is otherwise known as Class B performance requirements.
This service is intended for packetized video and voice applications, such as video conferencing
systems.
6.2.4 VBR-NRT
VBR-NRT is for types of traffic which are predictable; yet do not require a timing relationship to
be maintained end-to-end. This service can be used for interconnecting LANs.
6.2.5 ABR
The ABR service is designed for economical support of applications with vague requirements for
throughputs and delays. Although ABR is mentioned here, it is not specified in either UNI 3.0 or
3.1. ABR is covered under the latest UNI, which is V4.0. This has been ratified and many ATM
switch manufacturers are already offering ABR support.
6.2.6 UBR
UBR operates on a 'best effort' basis, with no reservation of bandwidth. Signaling used to set up
and clear down calls is normally transmitted as UBR, as is Local Area Network Emulation
(LANE) traffic.
6.2.7 GFR
GFR is a new service category which is still being defined. It is intended to provide a mechanism
that will deliver frames (as cells). If one cell is lost they are all lost. What is guaranteed is a
frame rate rather than a cell rate. It is included here in order to provide a complete overview.

45
QoS Service Categories
1 CBR Constant Bit Rate
2 VBR-RT Variable Bit Rate - Real Time
3 VBR-NRT Variable Bit Rate - Non-Real Time
4 ABR Available Bit Rate
5 UBR Unspecified Bit Rate
5 GFR Guaranteed Frame Rate (later)
Table 6.2.1 QoS Service Categories
6.3 ATM Traffic Classes
The ITU-T I.362 standard provides the functional descriptions for the AALs. Traffic classes are
based on the following parameters:
 Whether a timing relationship is required between source and destination
 Whether the traffic is CBR or VBR
 Whether the traffic is connection-oriented or connectionless
Class A: defines traditional synchronous data, such as that containing E1 voice circuits or
uncompressed broadcast video.
Class B: covers compressed video which requires a timing relationship.
Class C: defines busty data such as Frame Relay, X.25 or large file transfer.
Class D: includes broadcast data such as SAP messages in NetWare or an ARP packet in
TCP/IP.
There is also a Class X which covers UBR where traffic type and timing requirements are
defined by the user. Reference may be found, in ANSI documentation, to Class Y. In the ITU
and ATM Forum documentation this is known as ABR. The four main classes map onto the
AALs directly as shown here. AAL5 is the ATM Forum's response to the ITU-T's excessively
complex AAL 3/4. AAL5 is sometimes known as SEAL (Simple and Efficient Adaptation
Layer).

46
6.4 General Principles of Adaptation
Although the use of a Convergence Sub layer (CS) is specified for the AAL, not every type of
data requires it. Class A data (that is synchronous voice or uncompressed video) does not, for
example, need the CS. Higher-layer data traffic must be manipulated through the CS in order to
ensure that end-to-end transmission can be undertaken on a packet basis without loss of session.
In addition, each ATM cell must be filled with data. If the source produces less than 48
bytes, or if the volume of data is not divisible by an integer number of 48 bytes, then padding
must be added. The Convergence Sub layer adds its own header and trailer before segmentation
and reassembly.
6.5 Usage of the Adaptation Layer
As the AAL is used to adapt traffic to an ATM format, it is needed at the entry point to the ATM
network. Once the traffic is adapted to the ATM cell format it travels across the network in ATM
cells, which are switched in the ATM layer of switches along the path of the ATM connection.
Once the cells reach their destination there is a need to reassemble the traffic back into the
format of the original application. The AAL is thus also used at the exit point of the ATM
network. From the diagram opposite and the above description it would seem that there is no
need for an AAL function in core ATM switches. All switches, however, require an AAL
function. For example, signaling traffic needs to be interpreted by all switches. To interpret
incoming signaling traffic, the ATM switches needs firstly to reassemble the signaling traffic.
This requires an AAL function. The switch knows that traffic coming in from a given connection
is signaling traffic, as a special reserved VPI/VCI value is used. Similarly, all ATM switches
need to interpret operation and maintenance traffic and management traffic. Both signaling and
management are covered in a .
6.6 AAL 1
Class A data (E1, DS1, and data from other voice circuits) is by its very nature synchronous. A
function of the AAL associated with Class A data, AAL1, is to ensure that there is timing
integrity between the sending and the receiving end. Another function is to carry out clock
recovery at the destination. The AAL also provides a mechanism to detect lost cells, and inserts a
dummy into the cell stream to ensure that the timing information is not lost. This section

47
provides an overview of the operation of AAL1. A specific application of AAL1 that of Circuit
Emulation, is examined in the chapter on Voice over ATM.
AAL1 Operation
To format Class A data into cells, the data stream at the defined operating speed is simply
chopped up into 47-byte chunks. These chunks are known as Service Data Units (SDUs). Each
47-byte SDU is preceded by a one-byte header, resulting in a 48-byte payload.
6.6.1 AAL 1 Operation
The one-byte header added to each 47-byte SDU comprises two parts, the Sequence Number
(SN) field and the Sequence Number Protection (SNP) field. The SN field is then split into two
parts: the Convergence Sub layer Indication bit (CSI) which is normally set to 0, and three bits
for the Sequence Number. This cycles through from 0 to 7 and back to 0 again, and is suitable
for identifying missing or misinserted cells. To ensure the integrity of the SN field, it is protected
by the SNP field, which is a three-bit CRC check with an additional even-parity bit.
6.7 AAL 2
AAL2 defines the transport of VBR traffic that is timing-sensitive, such as VBR audio and
video. AAL2 is new and the data which would be supported by it was traditionally transmitted in
AAL5 instead. A problem with the utilization of AAL 5 however is the lack of delay parameters.
By contrast AAL2 is inherently designed for the support of VBR traffic, for which timely
delivery is an issue. A feature of AAL2 is the ability to accept several streams of traffic and
multiplex them together. The manner of multiplexing is to accept samples and to append a small
header to each sample. The primary function here is to add a channel number to identify the
higher-layer stream. Once so labeled, blocks are then transferred to the ATM cell payloads. Part
of this transfer is to add yet another header. The principal use of this header is to identify the start
of a flow after a short period of inactivity (one of the streams may show a blank screen for a
couple of seconds and the stream may produce no output).

48
6.7.1 AAL 2 (Continued)
AAL2 is important as it will be used in the third generation mobile telephony standard, Universal
Mobile Telecommunications System (UMTS). The formats of the AAL 2 headers are shown in
more detail opposite.
6.7.2 Initial AAL 2 header
Key fields of the initial packet header are the Channel Identifier (CID), the Length Indicator (LI),
and the User-to-User Indication (UUI) fields. These are defined below. CID Field The channel
identifier field identifies the individual user channels within the AAL2, and allows up to 248
individual users within each AAL2 structure. The CID field is actually 8 bits, thus allowing a
theoretical maximum of 255 individual users. However, several CID values are reserved for
management functions and future functions, hence the figure of 248 users. LI Field The length
identifier identifies the length of the packet payload associated with each individual user, and
assures conveyance of the variable payload. The value of the LI is one less than the packet
payload and has a default value of 45 octets, or may be set to 64 octets. UUI Field One current
use for the User-to-user field is to negotiate a larger Maximum Transfer Unit (MTU) size for IP.
This function was originated from third generation mobile telephony standard, Universal Mobile
Telephony Standard (UMTS) development. AAL2 is important as it will be fundamental to this
UMTS standard. Carrying IP traffic is perceived as a major application in UMTS.
6.7.3 Secondary AAL 2 header
The Offset Field identifies the location of the start of the next packet within the flow. For
robustness the Start Field is protected from errors by the Parity bit (P) and data integrity is
protected by the Sequence Number (SN).
6.8 AAL 3/4
When work began on the definition of adaptation processes it was felt that different adaptations
would be required for connectionless and connection oriented data. However after six months
work the two workgroups discovered that they had produced near-identical processes. The two
work groups subsequently joined forces to produce the single adaptation known as AAL 3/4.

49
You will be shown in the next few pages the differences between AAL 3/4 and AAL 5 which is
another, simpler data adaptation that has found greater favors due to its lower overheads.
A complete packet of data, for example an IP or IPX packet up to 64 kbit, is taken and
encapsulated within a convergence sub layer. Padding of 0 to 3 octets is added to ensure a 32-bit
alignment to simplify processing. The SAR sub layer then adds a header and trailer to make up
the 48-byte PDU, which then has a 5-byte ATM header added by the ATM layer to make the 53-
byte cell. By comparison with AAL5 this adaptation has a further 4 octets of overhead per cell. It
effectively reduces the efficiency of AAL3/4 (in comparison with AAL5) by 4/48, that is to
8.3%.
6.8.1 AAL ¾ Frames
The general structure of the convergence sub layer and SAR layers is as shown in the diagram.
The sub-structures are examined in the following diagrams. As can be seen, AAL 3/4 has a
relatively high overhead. In this case, 4 octets are consumed by the header and trailer fields.
After subtracting this overhead, the payload has been reduced to 44 octets. Although originally
designed to carry all manner of traditional data traffic, AAL 3/4 was seen as overly complex to
implement and also as inefficient due to its high overheads. Consequently, most data traffic is
carried in AAL 5. The primary role of AAL 3/4 today is in transporting an SMDS service.
SMDS is essentially a niche application, and it is believed that SMDS will migrate to the more
efficient AAL 5. It is likely that AAL 3/4 will die out in time.
6.9 AAL 3/4 CS & SAR fields
6.9.1 AAL3/4 CS
Type Indicates the units used by the BA and Length fields. BTag/Etag These two 'tags' are a
numerical value (the same value), which help to ensure that it is a single CS unit that has been
received and not a damaged CS unit created by joining together parts of two CS units. BA Size
Length of the user information subfield of the CS payload. Pad Padding added to ensure that the
total length of the CS is divisible by 4 (32 bits).This is an engineering consideration to simplify
processing by 32-bit processors. Length of the user information subfield. Other fields and
subfields are reserved for future definition.

50
6.9.2 AAL 3/4 SAR
Segment Type Indicates whether a cell is the first (at the beginning) of a message (BOM),a
continuation of a message (COM), or the last (at the end) in a message (EOM). There is also the
possibility of a message that fits in one cell, called a single segment message. Sequence indicates
the position in a convergence PDU of a SAR PDU. MID this is multiplexing ID field. This can
be used to allow the multiplexing of several traffic streams into a single connection. All the cells
on a connection will of course have the same VPI and VCI values. This extra field is required to
Identify a particular end point. This feature is of greatest utility if the connection is point-to-
multipoint. Len this is the length of the actual data in the last cell of a message. The last cell of a
message may not be completely full; we do add padding but only to align on a 32-bit boundary at
the CS. CRC A 10-bit Cyclic Redundancy Check computed over the SAR PDU.
6.10 AAL5 (SEAL)
AAL5 has significantly lower overheads than AAL 3/4 and is, therefore, very widely adopted. In
practice, since AAL 2 is not yet widely used and AAL 3/4 is seen as overly complex and
cumbersome, only AAL1 and AAL5 are widely used. AAL1 is used for CBR traffic and AAL5
for all others: VBR, UBR and ABR.
6.10.1 AAL5 Frame Format
Rather than the multiple convergence sub layers of AAL3/4, AAL5 simply takes the network
layer packet and adds a single trailer. The PAD field is there to pad out the complete PDU so that
it can be divided into an integer number of 48-byte segments for loading into the cells.
6.10.2 AAL5 Trailer
The AAL 5 8-byte trailer consists of:
1 Two 1-byte fields which are unused
2 A 2-byte length field which indicates the length of the data, not including the trailer
and pad
3 A 4-byte CRC

51
6.10.3 AAL5 Transmission
With AAL5 there is no obvious method of working out which cell represents the end of one PDU
and which the start of the next PDU. To achieve this, the PTI field in the header is used. Bit 1 is
set to 1 when the last cell representing the PDU is assembled, and all other cells have the bit set
to 0. When the receiver sees the PTI field with bit 1 set to 1, it assumes that the next cell with the
same VPI/VCI number will be the first cell of a new PDU.
7 Signaling
7.1 Address Formats
Work is still proceeding on defining the most effective addressing structures for use in ATM.
Below are listed three formats that are used in private networks. The carriers have already
declared their intent to use E.164 addresses. AESA The preferred term for an ATM address is
ATM End System Address (AESA). It was decided that existing address formats would be used,
in particular Network System Access Point (NSAP) addresses. It is important to realize that
when used in this manner the address does not identify a Network SAP. Only the NSAP address
syntax is employed. In spite of this they are commonly referred to as NSAP addresses.
7.1.1 DCC ATM Address Format
Authority Format Indicator (AFI) = 39 Data County Code (DCC)
7.1.2 ICD ATM Address Format
Authority Format Indicator (AFI) = 47 International Code Designator (ICD)
NSAP Encapsulated E. 164 Address Format
Authority Format Indicator (AFI) = 45 E.164 - An E.164 format (telephone) number
7.1.3 General
Domain Specific Part (DSP) End System (or Station) Identifier (ESI) Sel Selector
7.2 Point-to-Multipoint Connections
Multipoint connections are a feature of ATM networks. They are used in all LAN techniques.
They will be a most important feature of broadcast networks such as those providing video on

52
demand. The process of setting up a point-to-multipoint connection involves first of all setting up
a point-to-point connection. It must be specified that this connection is to be multipoint (This
must be done as multipoint are uni-directional.) Once the initial point-to-point is set up
additional destinations (leaves) can be added. There are two alternative mechanisms that can be
used here:
1 Send a request to the root (the originator of the original point-to-point)
2 With signaling version 4.0 issue a Leaf Initiated Join (LIJ) request to the network.
In the case of LIJ, the root does not necessarily know of the existence of the new leaf. This is
problematic: how does the potential leaf identify the connection to which it wants to be added?
The answer is by the use of a Globally Unique Connection Identifier (GUI) and a server which
can allocate GUIs to the required cell stream.
7.3 The Traffic Contract
The traffic contract is the sum total of all the parameters required to define the characteristics of
a connection. The contract includes an indication of how the network is to verify that the user
does not use more resources than were requested at set-up time. The contract consists of a series
of requirements that are encoded for transmission to the network at the ingress switch to the
network (this includes a value of required bandwidth and delay). The call set-up is the longest
message in ATM signaling, for it is at the set-up stage that the network will pass across all the
necessary details of the call, based on the agreed traffic contract. The set-up message carries the
Destination 20-byte ATM addresses, plus the basic bandwidth parameters forward and reverse,
and the QoS class. The set-up message may also carry the source ATM address.
The Connection Admission Control (CAC) algorithm of the switch will then assess the
network in the light of the request, before allowing the connection to proceed to set-up. Passing
the local CAC check is no guarantee that the connection will be successfully set up. The ingress
switch will retain a copy of the pertinent parameters (such as PCR, SCR and MBS - covered in
detail in a later chapter) and will use this information to check that the connection stays within its
contracted bounds (a policing function).

53
The Traffic Contract
1 Destination address and the traffic contract from the essential parts of call set-up
2 The traffic contract between user and network establishes:
–Virtual bandwidth reserved in each of the forward and backward directions
– QoS class for cells in each of the forward and reverse directions
Table 7.3 The Traffic Contract
8 UNI Signaling
8.1 UNI Signaling
Recall that the user-network Interface (UNI) is that point between the end-point ATM equipment
and the first ATM switch. There have been several versions of the UNI specification, defined by
the ATM forum: UNI 2.0, UNI 3.0, UNI 3.1 and UNI 4.0 (also known as Sig 4.0). Of these
specifications UNI 2.0 supports only PVCs, while the latter three versions also support SVCs.
An important point to note is that UNI version 3.1 and later versions are not backwards
compatible with UNI version 3.0. As ATMF signaling (from UNI 3.1 onwards) was aligned
with the ITU-T Q.2931 signaling standard we will now examine the Q.2931 standard.
UNI Signaling
1 The set-up message will carry the source and destination ATM addresses, plus the bandwidth
and the QoS parameters
– The call set-up message is chopped up using AAL5 and sent on reserved channel (VPI= 0,
VCI=5).
Table 8.1 UNI Signaling
8.2 Signaling Format
Signaling under ATM consists of joining together a variety of basic building blocks containing
the necessary information. These building blocks are known as information elements (IEs) and
each element has a standard 4-byte header followed by the IE content. Message Types:
8.2.1 Call Establishment

54
CALL PROCEEDING
CONNECT
CONNECT ACKNOWLEDGE
SETUP
8.2.2 Call Clearing
RELEASE
RELEASE COMPLETE
RESTART
RESTART ACKNOWLEDGE
8.2.3 Miscellaneous
STATUS
STATUS ENQUIRY
8.2.4 Point-to-Multipoint
ADD PARTY
ADD PARTY ACKNOWLEDGE
ADD PARTY REJECT
DROP PARTY
DROP PARTY ACKNOWLEDGE
8.2.5 Information Elements
A call set-up message has up to 19 separate IEs, although not all are used. Information elements
such as AAL parameters and calling party number are optional. While some, such as traffic
descriptor, called party number and broadband bearer capability, are mandatory. The IEs used in
setting up point-to-point or point-to-multipoint calls are listed here.
Signaling Format
1 The layer three signaling data unit:
– Standard header
_ AAL type (5)
_ Message type
2 Body consists of building blocks called Information Elements (IE)
3 IEs are built as required by the message type and service type
Table 8.2.5 Signaling Format Information Elements

55
8.3 Called Party Number (IE Example)
An IE is a clearly defined set of parameters. An IE may contain a single parameter or several
related parameters. For example, one IE contains all of the PCR, SCR and MBS values for each
direction of the connection. Each IE has a code, of 8 bits, that identifies which IE it is. There is
also a length field that specifies how long the IE is. Given the IE length the next IE can be
located from the length field. There is an IE that specifies 'sending complete' to indicate the end
of message. The fact that the identifier field is only 8 bits would seem to restrict the protocol to
256 different IEs. Although this may seem large, provision has been made to extend it
indefinitely. This is done by an IE that specifies a 'shift' rather like the case shift on a keyboard.
After shift, the next IE will be interpreted from a different list.
Each IE is decoded in the message and Connection Admission Control will examine it
before attempting a connection set-up.
Figure 8.3 Diagram of ATM Cell

56
9 B-ICI, AINI, IISP
9.1 Broadband – Inter Carrier Interface (B-ICI)
The B-ICI protocol specifies signaling and routing protocols that manage switched connections
between public networks. The most recent ATM Forum specification for B-ICI signaling is BICI
version 2.0. B-ICI signaling is based on signaling system number 7, which is divided into
Different user parts and a common transport system. The common transport system is called the
Message Transfer Part (MTP) and is responsible for the reliable transfer of signaling messages
between nodes. The specific user part pertinent to B-ICI is the Broadband Integrated Services
User Part (BISUP). Thus the B-ICI protocol consists of:
1 The Message Transfer Part 3 (MTP-3) signaling message handling protocol
2 The Broadband Integrated Services Digital Network User Part (BISUP) BISUP is a
node-to-node signaling protocol that provides signaling capabilities for interconnected
ATM networks. BISUP allows for the carriage of signaling messages across B-ICI. These
messages provide for the establishment, supervision and release of virtual connections
between inter-connected nodes. B-ICI routing tables are static tables configured by the
operator. This means that the operator has full control of the network routing.
Broadband – Inter Carrier Interface (B-ICI)
1 B-ICI = Broadband Inter-Carrier Interface
2 Used for creating SVCs between different public networks
– Static routing - routing tables defined manually
3 Based on SS7 family of signaling protocols
4 Consists of Message Transfer Part 3 (MTP-3) and the Broadband Integrated Services User Part
(BISUP)
5 Current ATM Forum specification is B-ICI version 2.0
Table 9.1 Broadband – Inter Carrier Interface (B-ICI)

57
9.2 ATM Internetworking Interface (AINI)
The ATM Inter-Network Interface (AINI) service provides functions for connecting PNNI
networks, or inter-connecting PNNI and B-ISUP networks. AINI is based on the ATM Forum
PNNI specification version 1.0, but uses static routing. It is likely that AINI will be used to link
PNNI domains. Used in this way, the network operator gains the dynamic benefits of PNNI
while retaining control of inter-domain routing since AINI uses manually defined routes.
ATM Internetworking Interface (AINI)
1 AINI = ATM Internetworking Interface
2 Based on PNNI protocol
3 Used for connecting PNNI networks or interconnecting PNNI and BISUP networks
4 Static routing - routing tables defined manually
Table 9.2 ATM Internetworking Interface (AINI)
9.3 Interim Inter-switch Signaling Protocol (IISP)
IISP was an interim solution to dynamic set-up for private networks. The greatest effort in PNNI
development was the creation of the routing algorithms. While the routing algorithms were being
perfected, IISP defined how to create static routing tables. IISP is not a scalable solution, as the
amount of configuration required grows exponentially with the number of switches and
connections. The signaling system in IISP is based on Q.2931, with the definition of some new
IEs specific to the private network. With IISP (Q.2931) the signaling process is asymmetric. This
means there is a clearly defined user side and network side. The administrator has to pick which
is which. Being based on static routing tables, the switches cannot adapt to changing topology. If
a link is down, alternatives will not be considered unless explicitly entered in the tables. Once
PNNI version 1 was issued, the IISP solution to call set-up in the private network was considered
to be far inferior. IISP should be considered irrelevant once a vendor implements PNNI.

58
Interim Inter-switch Signaling Protocol (IISP)
1 IISP = Interim Inter Switch Signaling Protocol
2 IISP can be thought of as „PNNI without the routing‟
3 Similar to UNI, IISP is based on UNI version 3.1 specification
4 Static routing - routing tables defined manually
Table 9.3 Interim Inter-switch Signaling Protocol (IISP)
10 ATM Traffic Descriptors
10.1 Traffic Management
Traffic management is necessary to ensure that both the priorities of the user and the
performance criteria of the network are met. The main threat to orderly traffic transmission is
congestion and the possibility of having to start discarding cells, due to too many cells arriving at
a point in the network. The network must cater for all types of traffic and deal with the different
QoS requirements for the various traffic types. CBR connections, for example, are very delay
and delay-variation sensitive. They require different parameters to VBR connections, such as
LAN traffic. LAN traffic, due to its busty nature, is well able to tolerate variations in end-to-end
delay. It also has upper-layer software to protect it against lost or misinserted cells. Traffic
Management is divided into three essential subjects:
1 Connection Admission Control, which decides whether the requested connection can be
allowed
2 The Traffic Contract, which defines the parameters of traffic at the time of connection
3 Traffic Shaping and Policing, which polices the connection once it is Allowed Traffic
Management is the function of the ATM Layer, and is subject to the QoS requested at the
call set-up time. Other features of Traffic Management are:
4 Resource Management, which is the allocation of network resources to separate traffic
flows according to their service characteristics
5 Traffic Policing, which is used to monitor traffic volumes
6 Priority Control, where cells may be selectively discarded to meet network Objectives

59
Traffic Management
1 Traffic Management is essential for the proper operation of ATM
2 The aim is to ensure that all the different classes of traffic receive the appropriate handling
3 Main features of traffic management:
– Traffic Contract
– Connection Admission Control
– Traffic Shaping
– Traffic Policing
Table 10.1 Traffic Management
10.2 Traffic Descriptor Parameters
To set up an SVC connection on an ATM network, a call set-up request is transmitted over VPI
= 0, VCI=5. This sequence of data consists of a number of information elements, each of which
represents an aspect of the traffic. If a CBR connection is being requested, only the Peak Cell
Rate (PCR) needs to be specified in each direction. Recall that it is possible to set up asymmetric
connections. If a VBR connection is being requested then the Peak Cell Rate, the Sustainable
Cell Rate and the Maximum Burst Size must be specified. ABR traffic is specified by the Peak
Cell Rate and the Minimum Cell Rate, while UBR traffic is specified by the PCR and a 'best
effort indicator‟. The traffic descriptor is an attempt to capture the characteristics of the source
traffic in terms of rate and volume. The mechanisms used by the network to ensure a connection
is established that can accommodate these requirements are examined later.
In the slide opposite there is a distinction made between source traffic descriptors and
connection traffic descriptors. The distinction is that the source traffic descriptors originate with
the end-station‟s call request, while the Cell Delay Variation Tolerance (CDVT) is set by the
operator within the switch.

60
Traffic Descriptor Parameters
1 These are the parameters requested at connection setup time:
– Peak Cell Rate (PCR)
– Sustainable Cell Rate (SCR)
– Maximum Burst Size (MBS)
– Minimum Cell Rate (MCR)
– Cell Delay Variation Tolerance (CDVT)
Table 10.2 Traffic Descriptor Parameters
10.3 Required Parameters
The table opposite summarizes the connection parameters required to support the listed service
category.
10.3.1 CBR
Characterized by the Peak Cell Rate, by definition the output is expected to be constant. There
are no defined mappings between Traffic Classes and Service Categories (some of them are quite
obvious). Class A traffic uses AAL1, which in turn uses a CBR service category connection.
Bursty data can be sent over a CBR connection if timely delivery is an issue. The tolerance of
delay variation is also important in order to prevent audio/video buffer overflow or starvation.
10.3.2 VBR (Real time and Non Real-time)
For VBR the PCR is still important, but so is the sustainable rate. SCR is a parameter in
provisioning buffers at the ingress point to the network. It is important in configuring the
conformance definition (more on this later). Again CDVT is considered.
10.3.3 ABR
PCR is required, as always. Minimum Cell Rate here defines the minimum amount of traffic that
is required to sustain the application (in order, for example, to prevent a database update timing
out). A CDVT is specified again. ABR is unique because it relies upon a feedback mechanism

61
that provides the end system with an indication of impending network congestion (The ABR
endpoint is expected to throttle as necessary.)
10.3.4 UBR
Only needs to call out a PCR. The connection is then more or less free to send at any rate
between zero and PCR. The traffic sent has, however, no guarantees. It is 'best effort'.
10.3.5 GFR
Guaranteed frame rate (GFR) is a service category being defined by the ATM Forum and the
ITU. Its aim is to carry entire frames, such as IP or Frame Relay frames, across the ATM
network. This is in contrast to other service categories which deliver only a specified cell rate,
GFR delivers a specified frame rate.
10.4 Peak Cell Rate (PCR)
VBR traffic is transmitted at the PCR specified. However the CAC algorithm endeavors to
allocate bandwidth based on the SCR, by equating the VBR traffic with an equivalent channel of
CBR.
PCR is always specified. It is tempting to regard PCR in relation to CBR as 'guaranteed'
traffic, and some service providers elect to reserve bandwidth for CBR connections at the PCR.
It will be seen, however, that when buffering is used it is possible to reserve less than the PCR
for a CBR connection, provided that the aggregate CBR traffic on a connection does not
approach the link rate. PCR places an absolute maximum limit on the traffic that can be sent over
a connection. Any traffic sent at a rate beyond the PCR is unconditionally discarded.
Peak Cell Rate (PCR)
1 Peak Cell Rate (PCR) is the absolute maximum rate at which the network guarantees cell
delivery
– A user may send cells at this rate for a short period of time
– Rate is reduced to maintain an average (SCR)
2 PCR is used by CBR, VBR and ABR service categories
Table 10.4 Peak Cell Rate (PCR)

62
10.5 Sustainable Cell Rate (SCR)
ATM networks calculate the SCR of a traffic stream over a period of time (typically a short
interval, for example 50 ms), in order to compare the actual SCR with the requested SCR. As
ATM circuits are unidirectional, the SCR and PCR must be specified for both forwards and
backwards transmission.
Generally speaking, an application can send at the SCR without hindrance. As with the
PCR, SCR bandwidth is not actually reserved across the network but a value derived from SCR.
If an application sends beyond SCR, but not greater than PCR, then those 'extra' cells will be
accepted up to the limit specified by BT but they will be marked for possible discard. A good
analogy here would be Frame Relay, where a user sends data at a rate greater than the committed
information rate (CIR) but less than CIR + EIR, in which case the frames are accepted with the
discard eligibility bit set.
SCR places an upper limit on the average cell rate. Note that average is not a good
engineering parameter as its value is only known when the connection is terminated (at which
time the volume of traffic sent and the duration of the connection are known).
Sustainable Cell Rate (SCR)
1 Sustainable Cell Rate (SCR) is the average rate that a network guarantees cell delivery
2 Users may burst above the SCR to the PCR (up to a maximum of BT) as long as they reduce
their rate of flow to maintain this rate
3 SCR is only used by the VBR QoS category
Table 10.5 Sustainable Cell Rate (SCR)
10.6 Maximum Burst Size (MBS)
Maximum Burst Size (MBS) is the size of the largest burst that can be accommodated by the
input buffer at the switch. If the transmission exceeds the MBS, then cells may be discarded.
MBS is measured in cells. Burst Tolerance (BT), which is related to MBS, is defined as the
maximum period during which the network will accept cells at the PCR.

63
he two terms are not the same, apart from different units, as cells can leave the input
buffer during the arrival of the burst. The burst tolerance is therefore greater than the M
BS.When a call is set up, the traffic parameter is MBS. The following computation
can be made to derive the BT:
BT = (MBS-1)(1/SCR - 1/PCR)
BT is a parameter that is used as a value for the size of a buffer, used to monitor the size of a
burst. The principle, which is explained later, is often referred to as a 'leaky bucket'. In operation
if the burst is too large, then it overflows the bucket and is lost (or at least it is indicated as a cell
that is outside defined limits as specified in the contract).
Maximum Burst Size (MBS)
1 Burst Tolerance (BT) is the maximum time that the network will accept cell rates of PCR
2 BT is only used by the VBR QoS category
Table 10.6 Maximum Burst Size (MBS)
10.7 Minimum Cell Rate (MCR)
Available Bit Rate (ABR) is specified in UNI 4.0. This traffic makes use of gaps in previously
reserved but unused bandwidth in order to provide a low-cost service unsuited to real-time
services (such as voice, video or multimedia traffic). It is, however, ideal for services such as
overnight file back-up.
The MCR and the PCR are specified, and the system uses a relatively high overhead of
resource management (RM) cells to try to pass traffic at a rate within the range specified
(approximately 1 in 30 cells are RM cells). MCR is a parameter that needs to be specified in the
GFR service category, although the MCR is directly related to the guaranteed frame size.
Note Although GRF is mentioned several times in this document, very little reference
is made to it in either the ITU or ATM Forum documentation. GRF is designed to characterize
frame-based traffic more closely, since some guarantee needs to be provided in relation to the
carrying of complete frames (for example, Frame Relay and mapping CIR to a series of ATM
cells). Much work remains to be done here, but references will be made where appropriate.

64
Minimum Cell Rate (MCR)
1 Minimum Cell Rate (MCR) is the highest rate at which the network guarantees delivery of
cells
2 A user may attempt to send at higher rates at the risk of losing cells
3 This parameter is used to support an ABR service
Table 10.7 Minimum Cell Rate (MCR)
10.8 Cell Delay Variation and Tolerance
The absolute cell delay in a network is only critical to CBR traffic, as VBR traffic is synchronous
only within the burst of data. However, while cell delay may not be critical it is possible that
variations in delay could be a problem. Consider television for instance. Delay in itself may not
be a problem, provided the delay is not several minutes, but if the delay and rate at which the
frames are displayed is varies then this would definitely be a problem.
The causes of delay variation or 'jitter' can be divided into four specific areas:
1 The statistical multiplexing on bearer circuits
2 ATM switch queues
3 Network Management cell insertions
4 the physical overhead in the lowest level of ATM
When cells from two or more ATM connections are multiplexed onto a common link, the cells of
one connection may be delayed while cells from another are being inserted. In a similar way
some cells may be delayed while OAM cells are inserted, or while a physical layer overhead is
added to the data stream. As a result, a random element is introduced into the time interval
between the end point of the ATM connection receiving the cells and the time at which the cells
are put into the network.

65
Cell Delay Variation and Tolerance
1 Variation in Cell Delay is a fact of life
2 Delays are caused by :
– Multiplexing
– Queuing
– OAM cell insertion
– Physical Layer overhead
3 An application may need a guaranteed limit on the degree of variation, a specified tolerance
Table 10.8 Cell Delay Variation and Tolerance
11 Qualities of Service Parameters
11.1 Quality of Service (QoS) Parameters
QoS in an ATM network is defined by six parameters which characterize the performance of a
given connection. The parameters quantify the performance of the connection all the way across
the ATM network, but excluding the end stations. Three of the six parameters are negotiable
between the end-station.
Quality of Service (QoS) Parameters
1 Negotiable QoS Parameters
– Cell Loss Ratio (CLR)
– Maximum Cell Transfer Delay (Max CTD)
– Peak to peak Cell Delay Variation (peak-to-peak CDV)
2 Non-negotiable QoS Parameters
– Cell Error Ratio (CER)
– Severely Errored Cell Block Ratio (SECBR)
– Cell Misinsertion Rate (CMR)
Table 11.1 Quality of Service (QoS) Parameters

3rd sem atm basics for wcdma networks M.TECH ( PDF FILE )

3rd sem atm basics for wcdma networks M.TECH ( PDF FILE )

Recommended

Recommended

More Related Content

What's hot

What's hot (19)

Similar to 3rd sem atm basics for wcdma networks M.TECH ( PDF FILE )

Similar to 3rd sem atm basics for wcdma networks M.TECH ( PDF FILE ) (20)

More from rajasthan technical university kota

More from rajasthan technical university kota (20)

Recently uploaded

Recently uploaded (20)

3rd sem atm basics for wcdma networks M.TECH ( PDF FILE )