These search terms have been highlighted: convert 1056000000 bits bytes 

Page 1

Fast, Cost-Effective xDSLTesting

Gigabit Ethernet: The Challenges of Network Analysis

ATM and DECT Background

Page 2



WG SLK: A New Range of Test Equipment

for Economical Testing of xDSL Links



LinkView ± A New Software Analysis

Product Family


WG Domino Goes Gigabit


Handheld Analyzer for Radio Fields

up to 2 GHz



1.5 Mbit/s Monitor


An Essential New Function for IBT-10 and IBT-20:

Detailed Real-Time Analysis


Safety in Constant Magnetic Fields:

ETM-1 Tesla Meter for Measurements down to 0 Hz


Measurements on Optical Networks: Two Paths

Leading to Reliable Results


WG RTU-500 Remote Test Unit:

A Valuable Business Investment


PA/PFA Family ±

A Worldwide Success!




Technology On The Road To Success



Testing Radio-Link Modems at Bosch Telecom.

ANT-20 with CATS in VEE Test Environment



CVI Applications Test Sequencer


Football World Cup 1998:

Victory for France Telecom


ATM Channel Explorer

Makes Monitoring Easy


Monitoring MPEG-2/DVB

Digital Broadcasting


WG Extends its Digital Broadcast Product Range



ISDN Testers and Analyzers Compatible with QSIG

and CorNet




WG IBT-5 Purchased by Telecom Organization

of Thailand (TOT)


TELSTRA (Australia) Visits Wandel & Goltermann

in Rennes (France)



The Challenges of Network Analysis

at Gigabit Speeds


ATM Performance Measurements:

The Information is in the Cell


Digital Signal Processing

in Analog Products


The Disturbing Effect

of Echo


Publishing information

ªbitsº is published at irregular intervals.

Issue 82: October 1998

Published by Wandel & Goltermann GmbH & Co,

Elektronische Meûtechnik,

PO Box 1262, 72795 Eningen u.A., Germany

Editor-in-chief: Burkhard Braach,

Tel. +49 7121 86 1793, Fax +49 7121 86 1333

Coordination: Bernd Schepper,

Tel. +49 7121 86 1287, Fax +49 7121 86 1333

The Editor will be pleased to grant permission

to reproduce parts of this publication on request.

Source acknowledgment is required.

Printed in Germany

Subscribe now at!


W a n d e l & G o l t e r m a n n b i t s 8 2

On the Back Page: WG Literature

Advanced Network Testing ±

A Newsletter for ANT-20 Users

InterNetworking ± A Magazine

for Data and Protocol Testing

SONET Pocket Guide

SDH Pocket Guide

GSM Pocket Guide

LAN Troubleshooting and Baselining,

Pocket Guide

ISDN Supplementary Services,

Pocket Guide

ISDN PPP Troubleshooting,

Pocket Guide

These publications can be ordered or downloaded

bits 8221

bits 8222

bits 8231

bits 8232

bits 8233

bits 8234

bits 8235

bits 8236

Page 3

Wandel & Goltermann: Member of

the Network Management Forum

With effect from July 1998, WG is an official

(NMF). Information about this body and an index of

products complying with NMF specifications is

available from their web page under

The NMF homepage is often used by professionals

in the network management field as a source of the

latest information in the NM scene.

New TMN Analysis Software

Wandel & Goltermann now offers a new class of

software for TMN/Q3 analysis: QMonitor TAP

decodes the complete protocol stack (layers 1 to 7)

and provides a comprehensive analysis of the

management information. It is based around the

WG Domino Analyzers. More information under

MA-10A Mobile Radio Analyzer:

Get The Feel Of It On Your PC

The new demonstration software CD is now

available, free of charge. Minimum PC

specification required: 486 DX, 33 MHz

Memory: 8 MB RAM

Video card: 256 colors, at least 6406480 resolution

CD ROM drive: At least double speed

Operating system: Windows 3.x or Windows 95

Application Notes

These are listed on page 33. You can order them

using the enclosed fax form or from

You can also download them



W a n d e l & G o l t e r m a n n b i t s 8 2


Wandel & Goltermann and W


Together meeting the challenges of

tomorrow's communications

Wandel & Goltermann and Wavetek are both renowned throughout

the world for high-quality test solutions for the communications

industry. The link-up between these companies combines Wandel &

Goltermann's leading position in the field of datacoms and telecoms

test equipment with Wavetek's strength in test solutions for cable TV

and mobile communications systems. Together, we can now offer

our customers around the globe a comprehensive range of test and

measurement solutions, products and services covering installation,

diagnostics, maintenance and servicing of voice, video and data

networks from a single source.

As one of the largest

companies in communi-

cations test technology,

we intend in future to

concentrate fully on being

the leading supplier of

communications test

solutions for the global

telecommunications industry. By supplying market-oriented,

competitively superior products and services with lifetime support,

we want to become the kind of competent, reliable partner that is

able to provide lasting solutions to problems and thus able to make

a valuable con-

tribution to the

success of our


At present, our


test solutions

cover all areas of voice, video and data communications. The

merger of Wandel & Goltermann with Wavetek means that the

combined company now has one of the most complete ranges of

test and measuring equipment and services on offer for the fields

of servicing, calibration and application-oriented training in the

member of the Network Management Forum

communications industry.

Our consistent approach towards close contact with our customers

means that we are continuing to expand our global market access.

The combination of a worldwide presence in the form of regionally-

managed sales companies and our multinational network of

production divisions remains the key to best serving the various

needs of our international clientele under consideration of

geographical and cultural aspects.

We will in future, too, improve our focus on our core business

within our selected markets by working together with other

companies in this sector through strategic alliances or financial


Linking our two traditional, forward thinking companies means

that we are ready right now to meet the future challenges of the

continuing communications age.

bits 8240

Page 4


xDSL is the modern, highly-promising strategy

for drastically increasing the performance of

existing copper access networks. It is expected

that within the next four years alone, some

20 million xDSL communications units will be

installed. Wandel & Goltermann supplies the

right test equipment for this technology.


A New Range of Test Equipment for Economical

Testing of xDSL Links

Comprehensive pilot projects have been

conducted worldwide that were mainly aimed at

investigating the use of ADSL in the private sector.

Now that the experimental stage has been largely

completed and mass production of modems has

started, wide coverage using this technique is

now imminent.

One important result of these pilot projects was

the fact that qualification of the lines for the high

bit rate services cannot as a rule be avoided,

despite the sophisticated technology used in the


The SLK Line Testers are the Wandel &

Goltermann solution to the problem

of qualifying copper wire pairs in the

minimum of time and with maximum

cost-effectiveness. In addition to

xDSL applications, the instru-

ments can also be used for

ISDN, PCM and traditional

telephone services.

Possible applications for SLK Test Sets

A handy carry bag contains the complete line

test set comprising two test instruments and

appropriate accessories. The instruments are

rugged enough to withstand the rigors of field

use for pre-qualification, installation and mainten-

ance of balanced copper pairs. Example

applications are:

. xDSL, ISDN, PCM and POTS line qualification

. Line troubleshooting, e.g. short circuits, poor

contacts and interference

. Location of hits in characteristic impedance

caused e.g. by different cable types, loading

coils or stubs

The advantages of the new Line Testers

To qualify a line, end-to-end measurements

with one test set at each end of the line

have to be performed. Such measure-

ments can be performed by just one

person, thanks to the communication

between the two instruments. Operation

is made extremely simple by means of

pre-defined test sequences, and the time

required for a complete line test is reduced

to just a few seconds, during which all

relevant parameters are measured,

compared with the stored limit values and

the result (good or bad) displayed using

colored light-emitting diodes. As well as signal

generator and receiver functions, the test sets

also perform as digital multimeters and reflect-

ometers. This means that users have all the major

functions essential for troubleshooting right at hand

in a single instrument.

. Only one operator required

. Automatic testing (Auto Test)

. Short test time

. Test result immediately visible

. Complete solution with built-in TDR


W a n d e l & G o l t e r m a n n b i t s 8 2

Page 5

All important functions in just one instrument

xDSL is the generic term for various DSL techno-

logies (e.g. ADSL and HDSL) which differ from each

other in terms of transmission frequency, system

impedance and signal levels. The SLK Line Test

Sets take these system differences into account

and also provide facilities for testing lines for other

services such as ISDN, PCM or the plain old

telephone service (POTS). The operating modes

and measurement functions available are shown

for the SLK-22 Test Set in the following table:

Defining instrument parameters and test


Operation of the instruments is considerably

simplified by the fact that users need not be

concerned on-site with equipment settings, test

sequences or measurement limit values. All that is

needed is to select the desired communications

service and start the measurement process.

The selected service is defined by a ªtemplateº

stored in the instrument. Factory pre-programmed

templates or user-defined templates matching

specific requirements can be transferred from a PC

to the test set via the built-in RS 232 interface.

This practically eliminates measurement errors due

to incorrectly set parameters.

Software support for result data transfer

The configuration software is included with the Line

Testers. It is also used to transfer the measurement

results to a PC, thus greatly simplifying the produc-

tion of test records. For example, with the SLK-22,

the results of up to 200 line tests can be stored and

transferred to a PC. Detailed evaluation of each

individual result is, of course, also possible using

the test set itself.

The new SLK range of instruments yield enormous

savings in the costs of line qualification and

troubleshooting compared with previous test

methods. Users can also match the type and

scope of measurement to suit their exact require-


We will be pleased to send you detailed product

information on request.

Peter Ziemann

Wandel & Goltermann,

Division 1, Germany





Auto Test or

Single Test

± Attenuation

± Near-end crosstalk

± Far-end crosstalk

± Signal to noise ratio

± Loop resistance

± Impulse noise

± Broadband noise

± Reflection / Impedance

Pair Detect

Tone Tx

Tone Rx

± Generates or receives signal

tones for quick location of the

wire pair to be tested




± Indicates impedance mismatches

± Shows coupling between wire pairs


W a n d e l & G o l t e r m a n n b i t s 8 2


Operating modes and

test functions of the


The templates can be

selected from an ex-

works menu or you

can create your own.

The configuration

and transfer software

lets you load the test

equipment with

templates of all the

test parameters and

transfer the results

back to the PC.

bits 8201

Page 6

W a n d e l & G o l t e r m a n n b i t s 8 2



As reported in the last issue of ªbitsº, the

acquisition of Tinwald Networking

Technologies gives WG customers direct

access to leading software analysis tools that

will be integrated into our Domino and

NetForce product families. Here is a brief

product summary.

LinkView ± A New Software Analysis Product Family

LinkView Internet Monitor is designed to provide

network administrators and managers with the

ability to assess the impact of Internet use on both

network capacity as well as the organization's

productivity, user services, and security aspects.

LinkView Internet Monitor can be run for a period

of time ranging from minutes to hours to days

depending upon the objectives of the user of the

software. Available for purchase through the

Company's Website at

LinkView LAN Monitor, formerly known as

LinkView 1030, is a self-contained, software-only

network analyzer that works with most third-party

network interface cards. LinkView LAN Monitor is

available for purchase through the Company's

Website at

LinkView PRO is a popular cost-effective network

analyzer which includes a broad array of network

analysis reports, data capture and protocol

decoding capabilities in an

easy-to-use Windows 95/NT

interface. LinkView PRO

is available for purchase

through the Company's

Website at

LinkView PRO with

WG Examine builds upon the

feature-rich LinkView PRO and

integrates WG's powerful

decode engine which features

over 300 protocol decodes.

The WG Examine decode

engine uses quick filters which

provide quick application of

filters to like frame types, and

advanced filters to allow

increased focus on specific

problems to help resolve

protocol interaction problems.

The decode engine offers

customizable display options

and symbolic name support.

LinkView PRO Distributed, formerly marketed as

NIMS, is the most affordable distributed analysis

solution for small and mid-sized companies.

Drawing on the key strengths of LinkView PRO,

LinkView PRO Distributed provides cost-effective,

easy-to-use remote network analysis. This

software-only management system allows you to

efficiently manage distributed multi-segmented

LANs over Internet/Intranet or TCP/IP WANs ±

installing in minutes without modifying any existing

network devices.

LinkView PRO Collision Expert combines

LinkView PRO with a specialized collision expert

that will analyze any collision, at any speed, from

anywhere on the Ethernet segment.

LinkView PRO

Collision Expert

LinkView PRO

with WG Examine

LinkView Token Ring Expert is a truly unique

product allowing network professionals to quickly

identify and pinpoint the source of a number of

low-level Token Ring problems, including hard to

diagnose jitter-related problems. LinkView Token

Ring Expert is a must have product for people

charged with the responsibility of managing and

troubleshooting Token Ring networks ± of which

there are still some 30,000 world wide, including

many that support the largest and most successful


For in-depth product information and free

downloadable demos, visit

Malcolm Aldridge

Wandel & Goltermann,

Division 3, USA

WG Domino Goes Gigabit

With an ever increasing number of network users

and data-intensive office, intranet, and multimedia

applications, who doesn't need more bandwidth

these days? Gigabit Ethernet is the latest

technology poised to bring relief to network

managers. Delivering raw bandwidth of 1000 Mbps

and compatible with existing 10/100 Ethernet

standards, this evolution of current Ethernet

technology promises to provide ten times the

performance of Fast Ethernet at only a fraction of

the cost. According to In-Stat Inc., Gigabit

Ethernet equipment sales are expected to jump

from $58.9 million last year to more than $266

million this year. By 2000, the market is expected

to exceed $1 billion. How is Gigabit affecting the

IT plans of your organization? If you're planning on

implementing Gigabit, how do you plan to test and

troubleshoot this high-speed network?

At a time when the market for Gigabit Ethernet is

so promising, Wandel and Goltermann introduces

the newest member of the Domino Family of

Internetwork Analyzers ± DominoGigabit. As a

member of the Domino Family, it provides the

flexibility and portability benefits offered by

Domino: Multisegment analysis with up to eight

multiple Domino analyzers for testing network

devices, autoconfiguration for fast, easy setup,

easy-to-use Microsoft Windows-based user

interface, lightweight and highly portable, and

compatibility with most notebook computers.

Key features include:

± Full line rate TX and RX including capture

± 128 Mbytes capture RAM

± Extensive real-time line rate filters and triggers

including IP addresses

± Power TX Wizard helps users define packets to

TX by simply choosing the desired protocol at

each level required, including 802.3 zp and

802.3 zq

± Automatically discovers and graphs top users

± Gives utilization graphs and protocol


± 32 nsec timestamp resolution on captured data

± Supports full duplex operation and monitoring

± Supports single mode and multi mode


The first portable Gigabit analyzer suited for field

service operations, this newest addition to the

Domino family can monitor activity, decode all

major protocols and generate network traffic,

making it an excellent tool for troubleshooting

Gigabit backbones. Due to its ability to link to any

other Domino analyzer, DominoGigabit is more

than a Gigabit Ethernet instrument. It also performs

as an internetwork analyzer, allowing the user to

track a problem across multiple segments of

variable types.

This instrument generates and receives traffic

simultaneously, making it an ideal tool for

benchmarking routers, bridges, gateways, and

other equipment prior to installation or after an

upgrade. Don't plan your entry into Gigabit without

considering the testing challenges.

Curtis Johnson

Wandel & Goltermann,

Division 3, USA

W a n d e l & G o l t e r m a n n b i t s 8 2



bits 8202

bits 8203

Page 8



W a n d e l & G o l t e r m a n n b i t s 8 2

The rate of change in the internetworking

market is extremely fast, causing network

vendors to invest in new technologies that raise

both performance and functionality. This trend

also impacts test equipment vendors who need

to do the same. Nowhere is the challenge

greater than in the area of portable field service

solutions which must balance cost,

performance, functionality, and size.

The Challenges of Network Analysis

at Gigabit Speeds

Before we go into a detailed discussion about

Gigabit Ethernet, here are some basic observations

about Ethernet at these speeds:

Different frame sizes are presented above to give

an indication of lower (64) and upper limits (1500),

as well as the most common frame sizes (250 to

500 bytes). When analyzing a solution for field

service applications it is important to distinguish

between frame sizes used in benchmarking

analysis (64 bytes) and real-world traffic

(250 ± 500 bytes).

Basic Network Analysis

Basic network analysis consists of:

. Frame capture and decode

. Basic statistical information

Capture and decode

At first glance it may appear a daunting task to

capture data at a rate of 2 Gbps (full-duplex 1 Gbps

Ethernet). In fact this is true for some technologies,

but not for others.

For example, the average PCI bus-based network

interface card (NIC) operates at a theoretical

maximum of about 1,056,000,000 Bps

(33 MHz632 bits) or 1 Gbps, about half of a

full-duplex Gigabit Ethernet link ± and that

theoretical capability does not include the bus

arbitration, generalized I/O interrupt or Windows

operating system overhead applied to each I/O

from the NIC to the hard drive where the data is

saved (and these drives usually top out in write

performance in the sub 100 Mbps range). So even

when 66 MHz, 64 bit PCI-based NICs become

available (which have a theoretical bandwidth of

4.2 Gbps) there is some doubt as to what the real

performance will be, given bus, operating system,

and hard drive limitations.


On the other hand, specialized customer hardware

solutions can save directly to RAM over proven

custom bus technology that already runs at

50 MHz (1.6 Gbps) or above per receiver (for each

side of a duplex link).


Some NIC-based solutions will have a limited amount of capture

RAM on the card.

More information



Inter frame gap

0.0000000096 seconds (12 byte times)


8 bytes

Half-duplex theoretical bit rate

1,000,000,000 bits per second

Full-duplex theoretical bit rate

2,000,000,000 bits per second

Effective bit rate (64 byte frames) ± half-duplex


Effective bit rate (250 byte frames) ± half-duplex


Effective bit rate (500 byte frames) ± half-duplex


Effective bit rate (1500 byte frames) - half-duplex


Frames per second (64 byte frames) ± half-duplex

1,488,095 pps

Frames per second (250 byte frames) ± half-duplex

462,963 pps

Frames per second (500 byte frames) ± half-duplex

240,385 pps

Frames per second (1500 byte frames) ± half-duplex

82,237 pps

Page 9

So the claim of line rate capture by recognized

providers of high performance test equipment is a

credible one with proven technology. More

complex is the issue of how much capture RAM is

needed to save to full line rates.

High-end solutions for Gigabit Ethernet will be

introduced to the marketplace with a base capture

RAM of 128 Megabytes (64 MB per receiver).

If the average frame size was 64 bytes long, and

the link was operating at full line rate, then this

would be enough capture RAM per receiver for

about 0.67 seconds (the effective bit rate for

64 byte frames is 761,904,762 bps).

0.67 seconds is not a lot in most people's minds

(though long enough to capture about 1 million

frames per receiver which someone will need to

sort through). In reality, most networks have

average frame sizes between 250 to 500 bytes

(some also have 128 byte frames). With 250 byte

average frame sizes, a 64 MB capture RAM will fill

in 0.55 seconds, or in 0.53 seconds for 500 byte


So a capture capability of between 0.67 and

0.53 seconds sounds fairly dismal on the surface,

and to some extent it is, but there are some

techniques to dramatically improve utilization of

capture RAM.

First of all, when initially troubleshooting problems,

troubleshooters claim that they need to capture a

great deal of data because they do not know what

they are looking for. This claim has merit, but the

depth to which they have to drill down in the first

pass is not great. For example, if you only capture

64 bytes of data from each frame (more than

enough to capture layer 2, layer 3, and often layer 4

information) you can start to get a high-level view

of what is going on. For a capture RAM of 64 MB

per receiver, a Gigabit Ethernet link with an average

frame size of 250 bytes would yield a capture

capability of 2.16 seconds (up from 0.53 seconds)

and 4.19 seconds for an average frame size of

500 bytes. That is for a link running continuously at

line rate. For a link that is 30 percent loaded, the

capture capability would be 7.2 seconds for an

average frame size of 250 bytes and 13.3 seconds

for 500 byte average frames.

So, 7 to 13 seconds is much better than the 0.67 to

0.53 seconds we started out with as the theoretical

limit, but is understandably still not satisfactory

for some. Once the initial high-level scan of the

problem is complete, you can really begin to

narrow the scope of what is captured through a

combination of filters and triggers.

In the long term, the raw capture RAM capability of

high-performance solutions is sure to increase as

the cost of memory continues to decrease and as

physical memory density increases, but in the

short-term, a combination of pre-capture filters,

triggers, and frame slicing will provide great value

to troubleshooters.

Basic Statistical Information

Line rates, frame rates, error rates, and network

statistics are the hallmark of high performance

solutions. The ability to provide this information is,

to some extent, getting easier as MAC chip set

vendors integrate some of this into their chips.

However, early versions of these chips often lack

this statistical information, so it requires the

specialized skill and capability to develop Field

Programmable Gate Arrays (often performing at

speeds approaching their design limits) to fill in the


Advanced Frame Processing

Beyond basic statistical information, capture, filter,

trigger, and transmit capabilities, functions such as

baselining and expert systems are also provided.

The problem with baselining and expert systems is

the sheer number of frames per second that is

available for processing.

With an average frame size of 250 bytes, there

would be 1,000,000 frames per second

(500,00062) and 500,000 frames per second

(250,00062) at an average frame size of 500 bytes

for a full-duplex link running at full line rate.

Given how far down into a frame these types of

applications go, this is a phenomenal amount of

processing. Once again, a network that was only

30 percent loaded would bring the processing load

down to between 300,000 and 150,000 frames

per second, which is within the realm of possibility

for modern micro-processors for some types of


High-performance hardware-assisted solutions

have an advantage over software-based solutions

in that they can have a dedicated processor for this

task in addition to running the processing on a

light-weight real-time operating system.

It is possible to implement baselining systems that

can get fairly accurate statistical analysis (over

time) by doing sampling. Nevertheless, the more

you can process, the more accurate the analysis.

Expert systems are a little more tricky. It is possible

to do iterative processing by applying specific

filters to look for specific problems.

Clearly, for the near future, high-performance

solutions will have an advantage in this area of

protocol analysis because of the dedicated

processing they can apply.


The raw performance capabilities of Gigabit

Ethernet look daunting at first glance, and indeed

provide many real challenges to providing good

analysis solutions at these speeds. Very few

vendors have the capability to build the high-

performance hardware needed to provide these


But sit back, consider real-world traffic levels,

selectively and appropriately apply filters, triggers,

and frame slicing, and the results will be


Curtis Johnson

Wandel & Goltermann,

Division 3, USA

W a n d e l & G o l t e r m a n n b i t s 8 2



Page 10



W a n d e l & G o l t e r m a n n b i t s 8 2

Bosch Telecom manufactures digital radio-link

modems in Offenburg, Germany. New test

benches for final production series testing are

being set up within the framework of a new

test strategy, with the aim of reducing test

expenditure and simplifying evaluation of

quality. The following is a report by GuÈnter

Arnswald and Gebhard Rall, responsible for

test engineering development.

Testing Radio-Link Modems at Bosch Telecom

ANT-20 with CATS integrated into VEE test environment

The Offenburg factory of Bosch Telecom is

equipped with an extended test infrastructure.

UNIX workstations and Windows NT PCs are

used as test computers. These control the test

equipment via LAN (TCP/IP) or IEC bus links. The

existing test software is based on Hewlett-Packard

VEE, which is a special graphic programming

language for test bench control. It forms the

foundation of a modular test software concept:

New test software is assembled from existing

and completely tested software measurement

modules where possible.

This environment had to be taken into account

when setting up new test benches for the radio-link


The aim: Low cost of production,

rapid implementation

Digital radio-link modems convert SDH data

streams from the baseband level (STM-1,

155 Mbit/s) to the IF level (70 or 140 MHz) and

vice versa. As well as the typical measurements

of system characteristic, signature, tolerance to

jitter, jitter transfer function and intrinsic jitter, the

collection of so-called internal quality data is also

important for Bosch Telecom. The new test bench

must therefore meet several criteria at the same


± Testing the modem in a real system

configuration with a short-circuit in the IF level.

± Sequential, fully-automatic testing of several

modems, with data storage and test reports.

± Reduction of test time requirements.

± Expansion of tests to include evaluation of

internal quality data.

The aim of test planning was to realize the test

software as quickly as possible so as to shorten

the implementation phase and keep production

costs low.

The solution: Existing WG test software

integrated into VEE

The test hardware consists of the following

components: A RFS-1 Fading Simulator from

Wandel & Goltermann (WG) links the devices

under test on the IF level. The STM-1 signals are

transmitted and received on the baseband level by

two ANT-20 Advanced Network Testers. WG-CATS

(CVI Application Test Sequences) software which

runs pre-defined test sequences automatically is

installed on the testers themselves, being triggered

by the UNIX workstation as the test computer.

This solution has advantages because the existing,

comprehensive library of prepared sequences,

so-called test cases, provided by CATS could be

used. This makes use of the built-in ªintelligenceº

of the test set. Basically, the only test software that

needed to be developed was for calling up the

WG-CATS test cases and for processing the test

data with the Bosch system software. Only a few

Page 11

measurements not directly available as WG test

cases needed to be created in VEE, a relatively

small outlay in comparison to that needed to

completely redevelop the fully-automatic measure-

ments running on the ANT-20.

The result: A test bench that fits into the Bosch


After a trying-out period, during which problems

due among other things to the timing behavior of

the TCP/IP network occurred, the test software is

now stable. Naturally, the advantages of rapid,

cost-effective implementation are offset by some

disadvantages: Invoking the WG test cases is

somewhat slower than direct instrument setting via

IEC bus, which might prove troublesome during

alignment measurements, for example. The current

version also does not provide intermediate results

during long-term measurements, making this

solution less suitable for long-term thermal stress

measurements. On the other hand, the test

computer and measuring equipment can be

widely separated, an advantage that makes it

possible to initiate measurements from practically

any location.

Gebhard Rall and GuÈnter Arnswald

Bosch Telecom, Offenburg, Germany

WG CATS ± CVI Applications Test Sequencer

WG CATS is a comprehensive library of pre-defined

ªtest casesº (test algorithms providing good / bad

evaluation) for telecommunications equipment

(SDH/SONET, PDH, ATM and optical communica-


The test cases have been created using

LabWindows CVI and normally run under the corre-

sponding CATS Test Executive. This allows the

creation of complete test sequences made up from

several test cases and provides records of the

results. In this case, however, an existing test

executive (VEE) was used by the customer to

implement the pre-defined WG test cases and so

reduce the large amount of programming required

to create the test algorithms.

The next stage of development will be to provide

the WG CATS test cases as DLLs (dynamic link

library) so that they can be directly integrated into

applications already being used by the customer.

This will eliminate the speed problem hinted at in

the above article.

Walter Besse

Wandel & Goltermann,

Division 1, Germany



W a n d e l & G o l t e r m a n n b i t s 8 2






















with hp-VEE

IEC bus


Start test







with WG CATS

The test bench. The

two devices under test

are linked by a Fading

Simulator on the

IF level and two

ANT-20 Advanced

Network Testers

transmit and receive

the STM-1 signals on

the baseband level.

Test bench control:

A UNIX workstation as

test computer controls

the ANT-20 via LAN.

The CATS test

sequences on the

ANT-20 (Domino

version) are simply

started via the LAN

and the results are

transmitted to the test


IF level (70/140 MHz)


under test

Baseband level (STM-1)

bits 8204

Page 12


W a n d e l & G o l t e r m a n n b i t s 8 2


32 teams, 64 matches, 280 radio and TV stations

in the link-up and a cumulative TV audience

of 40 billion ± just some of the obvious

superlatives. The hidden talents were in the

technology that made the whole thing possible.

One such superlative was the largest ATM

network to date which was set up by France

Telecom for TVRS 98, using measurement

technology from Wandel & Goltermann.

Football World Cup 1998:

Victory for France Telecom

The headquarters of the CIRTV transmission center

(Centre international de radio teÂleÂvision) was huge:

Two entire exhibition halls totaling some 20,000 m


at the Porte de Versailles in Paris were filled with

the equipment for distributing the pictures and

sound from all ten stadiums to all points of the


CIRTV was the technical center for TVRS 98, a

consortium set up by the French members of the

EBU (European Broadcasting Union): TDF

(TeÂleÂdiffusion de France, part of the France-

Telecom group), Canal+, France Television, TF1,

Radio France and Radio France International.

TVRS 98 held the rights to the programs which

were sold to foreign radio and TV broadcasters

such as ZDF, RAI, RTL and the BBC.

Gilbert Brachet and Hughes Launay of France

Telecom (National networks division) were chosen

to provide support for the ATM project for the 1998

Football World Cup. They were already at work at

Porte de Versailles by the beginning of April.

The pattern of work at CIRTV matched the project.

Mr. Brachet had to admit that he seldom left the

building before 10 p.m. ± not to mention all the time

spent working at weekends. But the goal was fixed:

Everything had to be ready by June 10th, and no

sacrifices in quality were allowed.

As far as test equipment for setting up and running

the ATM network was concerned, France Telecom

relied on Wandel & Goltermann's ANT-20 Advanced

Network Tester. The global success of this instru-

ment has made it the reference for testing PDH,

SDH, SONETand ATM networks.

France Telecom used eight ANT-20s altogether. Six

of these were rented from Livingston. They were

initially used to qualify the ATM paths from each of

the stadiums. Later on, they were used to analyze

transmission quality for network maintenance and


The France Telecom engineers and technicians

quickly recognized the advantages of ANT-20 in

use: Easy handling, high performance, small size

and wide range of interfaces.

This was not the first international project for

Gilbert Brachet and Hughes Launay. They had also

been involved in the 1992 Winter Olympics in


Gilbert Brachet showed us around CIRTV. The

actual node was surrounded by the studios for

each of the TV companies. Mr. Brachet explained

that each match day was simulated before the

games even began, especially the days when four

matches were to be played.

Mr. Brachet and Mr. Launay received support in

their job from the Centre de Supervision du

ReÂseau, France Telecom's network supervision

center located in Paris Saint-Amand, which kept

an eye on all the stadiums, and from the CIRTV

headquarters in Blagnac, which monitored all the

alarms in the national network.

In setting up the largest video network based on

ATM to be realized to date, France Telecom

succeeded in providing the necessary technical

and human resources and so can rightly be said to

have had a major share in the success of the

Mr. Brachet and one of

his ANT-20s at CIRTV

(Centre International

de Radio TeÂleÂvision)

Page 13

World Cup, particularly satisfying in view of the

resounding victory of the French team!

Jean-Raoul Boyer

Wandel & Goltermann


ATM Channel Explorer

makes monitoring easy

The ATM Channel Explorer is a virtual instrument

(VI) for the ANT-20 and ABT-20 Analyzers. It

monitors and analyzes live traffic on an ATM link.

Various scan and analysis criteria can be set for

this purpose.

Activity Scan registers up to 1000 active virtual

channels (VC) and determines the important

attributes such as VPI/VCI values and various

bandwidth parameters. The recorded data can

be further investigated, e.g. for AAL type.

Trouble Scan registers all virtual paths (VP) and

virtual channels (VC, up to 1000) exhibiting an

alarm state (AIS or RDI). The results for the

VP and VC levels (F4 and F5 flow) are listed


AALType Analysis determines and displays the

AAL types for the active ATM channels.

AALType Distribution Analysis determines

and displays the distribution of active ATM

channels according to their AAL types. The

distribution is shown as a pie chart and as a

table of percentages.

ATM Channel Explorer is part

of the Broadband Analyzer/

Generator option for ANT-20

and ANT-20E. It is included in

the base software for the

ABT-20 Advanced Broadband


Jochen Hirschinger

Wandel & Goltermann,

Division 1, Germany


W a n d e l & G o l t e r m a n n b i t s 8 2

The national ATM

network linked all the

stadiums to CIRTV.

The network backbone

was a STM-16 ring

(2.5 Gbit/s) to which

all stadiums were

linked via add/drop

multiplexers. ATM

communications has

the advantage, among

other things, that

bandwidth can be

adjusted freely. The

video signals required

9 Mbit/s to 34 Mbit/s,

depending on the

degree of compression,

with the HDTV trans-

missions for France

Supervision and the

Japanese broadcaster

NHK also using

45 Mbit/s. In addition,

the ATM Adaptation

Layer 1 protocol also

allowed automatic

correction of trans-

mission errors.

ATM Channel Explorer,

main window showing

list of results

AAL type distribution

results as pie chart

and numerical values


Parc des Princes

Stade de France











bits 8205

Page 14


W a n d e l & G o l t e r m a n n b i t s 8 2


Customers rightly expect the agreed quality

of service. Network providers need the

required performance. Technicians know

what's needed for both. This article gives an

insight into the background for measurements

of these criteria.

ATM Performance Measurements:

The Information is in the Cell

Network commissioning, system acceptance

or in-service monitoring: The performance

parameters are the same. The key to measuring

them is the ATM cell itself. The idea is simple,

but the technology behind the measurement is


ATM cell structure

The first four bytes of the cell (Fig. 1) contain

information indicating the virtual channel (VC) and

virtual path (VP) to which the cell belongs. The path

is a group of channels which follows a particular

route through the network (Fig. 2). The term ªvirtualº

makes it clear that channels and their grouping

can only be identified by the identifiers VCI, VPI)

in the cell header, since all cells use the same

physical communications path.

As this information is essential, the first four bytes

are protected by a checksum (CRC-8) transmitted

in the fifth byte (Header Error Control, HEC).

The checksum also serves as the synchronization

criterion on the receive side.

The test cell

A virtual channel carrying test cells with special

contents (Fig. 3) is set up in order to test the

performance parameters while the system is in

service. Consecutive sequence numbers and a

timestamp indicating the relative time of trans-

mission are inserted in the information field of the

cell on the transmit side. A large number of test cell

bytes remain unused and therefore contain zero

sequences. To ensure that the cell content is close

to ªliveº traffic, the contents of the information field

are scrambled. A checksum (CRC-16) is used at

the end of the information field to guard against

bit errors.

The cells in the test channel are then monitored on

the receive side, The principle is simple: Using the

sequence numbers, it is easy to see if cells have

been lost. This can happen if a FIFO buffer in the

system overflows or if the system rejects cells

according to a priority rule (CLP, PT) when

overloaded. It is also possible to see whether cells

have been inserted which originally belonged to

another channel. This might happen if a switch in

the path enters an incorrect header or fails to

detect a header error or does not correct such

errors properly.

The timestamp information makes calculation of

cell delay possible as long as the transmit and

receive clocks are synchronized. This condition is

most easily fulfilled by switching a loop at the far

end of the transmission path and measuring the

send and receive directions at the near end with

the same instrument.

Bit errors can be detected instantly by recalculating

the CRC-16 in the information field and comparing

it with the received checksum. The cell error ratio

can be determined in this way.



Fig. 1: ATM cell


Fig. 2: Virtual channel

and virtual path

Cell Loss Ratio

Ratio of lost cells to total number of cells transmitted

or received

Cell Misinsertion Rate

Number of misinserted cells per second

Cell Error Ratio

Ratio of errored cells to total number of cells


Cell Transfer Delay

Transmission delay

Cell Delay Variation

Variation in cell transmission delay between two points in

a network


Information field (payload)

5 bytes

48 bytes

H1 H2 H3 H4














Transmission path











Fig. 3: Test cell

Sequence number




Unused bytes



Cell synchronization, which is the prerequisite for

all subsequent stages, happens to be relatively

easy because ATM cell are usually transmitted in

SDH or SONETsystems that operate with

byte-oriented framing. All that is necessary is to

load the incoming bytes into a five-byte wide shift

register, to calculate the CRC-8 checksum from the

first four bytes and to compare this with the fifth

byte. Synchronization is assumed if after the first

agreement in this comparison a total of six further

checksums agreeing with the expected value are

found at intervals of one cell (53 bytes).

When a single bit error occurs, the HEC provides

for header correction. Multiple errors cannot be

corrected. In practice, bit errors usually occur in

bursts. To stabilize synchronization behavior, all

cells with an uncorrectable header error and all

errored cells following a correctable error are


The header filter identifies the test cells contained

in the cell stream by comparing their header

contents (VCI, VPI) with the corresponding values.

Before evaluation of the sequence number and the

timestamp can proceed, the CRC-16 check must

be performed and the information field content

must be unscrambled (descrambler). An error

detected in the CRC checksum is the criterion used

in counting ªerrored cellsº.

A subsequent function block detects missing or

lost cells. This is based upon usage of the

sequence numbers.

The transfer delay is determined by recording the

ªarrivalº time of the cell and subtracting the value

indicated by its timestamp. The delay variation is

then ªsimplyº a matter of calculation. Well, not so

simple, actually.

The Wandel & Goltermann solution

The two last-named functions are extremely

processor-intensive. They also require a large

amount of memory if all the original data in the

cells is to be saved. To avoid this and to

ensure that the CPU is not loaded excessively,

Wandel & Goltermann instruments

use hardware data reduction. A type of

histogram memory contains class counters

for delay values. Each time a cell is evaluated,

the corresponding class counter is simply

incremented. The time axis can be shifted to the

range of interest by means of a programmable

offset. Zooming is possible thanks to

programmable resolution.

The hardware is selected according to the task.

All function blocks subsequent to cell synchroniza-

tion in the standard ATM module of the ANT-20 are

made up from FPGAs (field-programmable gate

arrays). Compared with ASICs, these have the

advantage that they can be reprogrammed at any

time, for example if the standards or requirements

change. The last FPGA in the chain can thus

determine either the lost / misinserted cells or the

transfer delay depending on the measurement

mode selected.

The broadband analyzer / generator module of

the ANT-20 and ABT-20 uses a Power PC which is

capable of processing up to four test channels


Friedemann Stockmayer

Wandel & Goltermann, Division 1, Germany



W a n d e l & G o l t e r m a n n b i t s 8 2



Asynchronous transfer mode.


Cell loss priority. 1 bit in the cell header which sets two priority classes.

Cells with CLP = 1 may be rejected if the system becomes overload.

Cells with CLP = 0 may not be rejected if at all possible (compare PT).


Cyclic redundancy check. The checksum is generated using a polynomial,

the order of which is indicated by the designating number. For example,

the header error check polynomial is 1 + x + x


+ x


, i.e. 8th order, and is

hence called CRC-8.


First in first out buffer memory.


Header error control.


Payload type. A 3-bit wide field in the cell header indication the type of

information to be transported. This is of importance when determining

priorities. For example, cells containing OAM (Operation & Maintenance)

information may not be rejected by the system.


Synchronous digital hierarchy.


Synchronous optical network.


Virtual channel.


Virtual channel identifier. 16-bit wide field in the cell header containing

information about the channel to which the cell belongs.


Virtual path. A group of virtual channels.


Virtual path identifier. 8- or 12-bit wide field in the cell header containing

information about the path to which the cell belongs.

Fig. 4: Function blocks

used to evaluate test


Header error,


Header error,


CPU interface

Errored cells

Lost cells

Misinserted cells

CPU interface




Cell synchronization




Header correction

Header filter

Test cell

Descrambler; CRC-16

Lost / Misinserted

Transfer delay



W a n d e l & G o l t e r m a n n b i t s 8 2

The Protek 3201


is a handy, easy-to-use test

instrument suitable for a range of demanding

measurements on the air interfaces of

radio systems at up to 2 GHz. This compact

instrument is designed above all for the kind

of measurements encountered during the

installation and maintenance of mobile

communications systems (analog cellular

systems, GSM, CDMA, DECT), paging systems,

cable and satellite TV systems and antenna


Hand-Held Analyzer for Radio Fields up to 2 GHz

Until now, the only equipment available for

selective measurements on radio fields has been

expensive microwave spectrum analyzers or

special test receivers which are too large and

too heavy to be generally used for on-site

measurements. The 3201 is the first RF analyzer

in a handy, small format (1056220645 mm)

which is suitable for these applications. It can

be used anywhere, as it is battery powered and

weighs just 700 g. Measurements can now be

made easily and without problems even on

antenna arrays situated in high, inaccessible

towers and masts.

The 3201 is also very different from its larger

counterparts, the microwave spectrum analyzers,

in the way that it is operated. Such spectrum

analyzers with their universal features are much

too complicated for the simpler measurements

that occur during installation and maintenance of

radio systems, and they are correspondingly

complicated to use. Correct operation is normally

only guaranteed after lengthy training. For this

reason, special attention was paid during develop-

ment to making the 3201 as easy to operate as


Despite its small size, the 3201 is capable of a

wide range of measurements, including spectrum

analysis, level measurements on single tone or

multi-tone signals, searching for discrete

interference frequencies, demodulating AM and

FM signals (broadband and narrow band FM), as

well as frequency measurements and the recording

of level trends over a period of time. The results

are clearly displayed as traces or bargraphs

together with the numerical values for the

frequency marker and the measurement para-

meters on an illuminated liquid crystal panel

(1926192 pixel). The measurement parameters

can be set from one main menu and just a few

sub-menus. Operation is thus easily and quickly

learnt, requiring no special knowledge of spectrum

analysis on the part of the user. Operation is further

simplified by the facilities for saving up to ten

complete instrument settings for measurements

that occur frequently. It is also possible to save

up to ten results traces so they can be evaluated

later on.

The Protek 3201 is thus the ideal test instrument for

on-site use, where fast, uncomplicated measure-

ments are required without making sacrifices in

accuracy and reproducibility. The instrument is also

equipped with a RS 232 interface and Windows

software to allow data exchange and evaluation

with a PC.

Miguel Waschl

Wandel & Goltermann & Co.



The Protek 3201 is a product of HC HUNG CHANG, Korea,

distributed worldwide by Wandel & Goltermann & Co.

The Protek 3201

RF Analyzer for fast,


RF measurements


It is rare for people to associate a specific

item with the abbreviation used for a form of

telecommunications. DECT is an exception:

For most people, it means cordless telephones.

There is more to it than that, however. In fact,

one of the first uses was for a wireless LAN that

could operate at a data rate of 552 kbit/s.

This article shows why DECT is experiencing

such popularity and the meaning behind the


DECT ± Technology On The Road To Success

The beginnings

Digital Enhanced Cordless Telecommunication

(DECT) is a digital communications method that

was standardized in 1992 by the European

Technical Standards Institute. At that time, a

generally-applicable, comprehensive technical

standard was badly needed because the regulatory

authorities were practically unable to control the

widely differing types of cordless telephone then

being introduced to the market. Particular

problems were caused by the low security against

eavesdropping and the use of restricted frequency


What's the advantage of DECT?

The most important innovations resulting from the

DECTstandard were:

± Improved transmission quality using digital

techniques, interrupt-free handovers and ISDN

voice quality

± Data encryption for high security against


± High data transmission speeds: Up to 2 Mbit/s

available soon!

± Error detection and correction facilities such as

CRC, ARQ and FEC provide high data security

± Comprehensive interworking facilities with other

networks: ISDN, X.25, LAN, GSM

± Can be used for voice, data and multimedia in

the private, business and public sectors

± Operation of picocellular networks with roaming

and extremely high user densities:

10000 ªtalkersº per square kilometer possible

± Standardized radio interface

± Dynamic assignment of bandwidth

± Self-organizing: No frequency planning needed

How is this realized?

The DECTstandard (ETS 300 175-1 to 8) primarily

defines the parameters for the air interface

between the portable part (PP, mobile phone)

and the fixed part (FP, base station). Below is a

summary of the technical details:

Transmission type:


Multiplex procedure:


Modulation procedure:


Frequency range:

1880 to 1900 MHz (Europe)

Bearer spacing:

1.728 MHz

Duplex channels per bearer:


Number of bearers:


Total number of duplex channels:


Traffic density:

approx. 10.000 erl / km


Max. range:

300 m

Max. movement speed:

20 to 50 km/h

Where is DECTused?

Nowadays, most cordless phones operate using

the DECTstandard. Business use of DECTsystems

(Business Cordless Telephony, BCT) is enjoying

growing popularity. Many companies have built

up complete corporate networks using DECT

technology. The main advantage of this is that

one and the same portable phone can be used

anywhere on a given site and on any other site,

thanks to roaming. The high subscriber densities

possible with DECTsystems mean that they are

also ideal for airports, hospitals and trade fairs, as

demonstrated at CeBIT in Hannover every year.

DECT is also capturing the public access network

sector: Cordless Terminal Mobility (CTM) is the

catchphrase for linking users to telecommunica-

tions networks via DECT portables. Such networks

are already being introduced commercially; FIDO in

Italy is just one example.


W a n d e l & G o l t e r m a n n b i t s 8 2




W a n d e l & G o l t e r m a n n b i t s 8 2

A further use of DECT technology that has made

great gains in recent times is the setting up of

wireless local loops (WLL). Here, the emphasis is

on flexibility rather than mobility, and on the rapidity

with which they can be set up, saving costly

cabling. This is of particular advantage in

inaccessible or difficult terrain. The proportion of

the worldwide market assigned to DECT for WLL

system contracts during 1997 was 31%. This puts

DECTright at the front. Large numbers of these

installations are for developing and emerging


Metropolitan network providers are also discover-

ing DECTas a useful method of bridging the last

mile from the customer to the fixed network.

The low costs of operation and maintenance,

resistance to environmental effects and simplicity

of system expansion are the deciding factors here.

DECT is well-suited to data transmission, and thus

also provides an alternative to the analog loop in

the plain old telephone system (POTS):

Payload data rate/channel, duplex:

Voice (unprotected):

32 kbit/s (ADPCM coded)

Data, ARQ-protected:

25.6 kbit/s


24 kbit/s

ISDN interworking, double-slot format,


80 kbit/s


64 kbit/s

Symmetrical channel grouping,


24 to 264 kbit/s per bearer

Asymmetrical channel grouping:

24 to 552 kbit/s per bearer

These possibilities mean that many applications for

cordless data transmission can be covered:

± Fax (ISDN and analog)

± Internet access for PCs via DECTradio modem

± Wireless LANs

± ªDatapointº: Mobile Internet access at specific

crowded centers, e.g. airports and trade fair


± Video telephony

± Industrial control, e.g. automated stock control

± Measurement applications for moving parts

Typical scenario

for business cordless telephony (BCT)

or cordless terminal mobility (CTM)


Distance: 12 timeslots





DECT transmission ±

A combination of


The total number of

120 channels is given

by the ten bearer

frequencies, each of

which provides twelve

channels per trans-

mission direction.



Last but not least, the combined use of DECT

and GSM (dual mode) is a promising concept for

CTM. DECT CTMs can increase the capacity of a

mobile network in areas of high population density.

The advantage to users is that the GSM-specific

features such as SMS or call redirection can

also be used from the DECT portable. A policy of

reasonable pricing could provide additional

impetus for this development.

What Wandel & Goltermann offers

As an active member of the DECT Forum, WG

provides support in the form of professional test

equipment for most fields of application. For

example: The Wavetek ESP 4032 is for the physical

layer (TBR-6/TBR-10). The WG CPM-10 with

Fixed Part and Portable Part Monitor and Simulator

performs protocol analysis at the air interface.

WG also supplies a complete range of test equip-

ment for testing the links between DECTand fixed

networks such as ISDN or 2 Mbit/s PCM.

You can obtain more information from your local

WG sales partner or from our DECTwebpage under

Marc Kahabka

Wandel & Goltermann,

Division 1, Germany


W a n d e l & G o l t e r m a n n b i t s 8 2

ADPCM Adaptive Differential Pulse Code Modulation ± PCM with

data compression


Automatic Repeat Request


Business Cordless Telephony


Cyclic Redundancy Check


Cordless Terminal Mobility ± Special type of WLL with

roaming facilities


Digital Enhanced Cordless Telecommunication


erlang; traffic unit. 0.5 erl means that a channel is occu-

pied for 0.5 hours at peak periods.


European Telecommunications Standards Institute


Frequency Division Multiple Access


Forward Error Correction


Fixed Part


Gaussian Frequency Shift Keying ± digital modulation



Private Branch Exchange


Plain Old Telephone Service


Portable Part

Roaming Movement of the PP to another cell (or another ex-

change) without interruption of the connection


Short Message Service


Time Division Duplex


Time Division Multiple Access


Wireless Local Loop


Mobile radio network

Public network










1.5 Mbit/s Monitor

During the last 20 years, the testing of the PCM

Telephony interface has moved from large

bench-units into hand-held instruments which are

convenient for field use. These instruments have

remained relatively complex and expensive and

only provided for use by specialized engineers.

New integrational engineering techniques were

used in the design of the PENT1 to put part of the

functionality of these testers into a pocket sized

device at a cost which allows mass deployment

throughout the workforce of the telephone


Technicians who come into contact with T1 circuits

can now determine whether a line is active,

avoiding the possibility of interrupting traffic. The

instrument is so small and convenient that it can be

carried in the shirt pocket as if it were a pen and

will be with the technician at all times. This avoids

the need to return to a van or store for conventional

test equipment.


LEDs to indicate LOS, AIS.

LEDs to indicate B8ZS, framed signal, D4, ESF,

SLC96, unframed QRSS pattern.

Latched LEDs for the following errors and alarms:

BPV, CRC error, Yellow alarm,

D4/SLC96 bit 2 of all channels = 2,

D4 S bit of frame 12 = 1,

ESF repeating OOFF


pattern in FDL.

Receiver ANSI T1 403 1544 kbit/s,

sensitivity 0 to ±30 dB.

bits 8206



The DVB group (Digital Video Broadcasting)

has defined a family of standards that were

first applied to direct-to-home broadcasting

(DTH) using satellite or cable links and

which are now largely applied to professional

applications such as company in-house

television or tributary services.

Monitoring MPEG-2/DVB Digital


Tasks and responsibilities

Public digital broadcasting

Digital television based on MPEG-2/DVB

technology today involves more than 5 million

homes around the world as it concerns cable

or satellite broadcasting. Terrestrial network digital

television will become a reality when commercial

services are launched in Great Britain before the

end of this year and as a result of several pilot

schemes operated elsewhere in the world

(Singapore, Eastern Europe, etc.).

TV viewers receive digital television via cable or

satellite by subscription from a service operator

who may also rent out the receiving equipment.

With this receiving terminal, users have an

advanced television at their disposal that

provides an electronic program guide (EPG)

as well as interactive services (weather forecasts,

rapid Internet access, etc.).

The service operator is responsible for providing

the TV viewer with access to all these services on

a round the clock basis, without any interruptions,

which means that all synchronization information

within the MPEG-2/DVB multiplex structure

and the transport stream must be within limits

compatible with correct terminal function, and

that the signaling (including the access rights

assigned to the user) is correctly realized.

The service operator also has a quality of service

obligation towards the providers, i.e. television

networks, data providers, Internet service

providers, etc. Since individual services are

involved, this obligation is based on observing

average and minimum data flow rates which are

the key parameters in the compromise between

cost and service quality.

Professional services

The MPEG-2/DVB technologies are not only used

for direct broadcasting services: They are also

utilized within several professional applications

such as company in-house television (internal

program broadcasting, e.g. multiple-site training

and product launches), data services (remote

loading of information to ISP servers, etc.), tributary

services (point-to-point and point-to-multipoint

transmission of television programs for feeding

cable network feeders, for example), to name just

some applications.

These services are generally set up by Telcos or

their subsidiaries having responsibility for trans-

mission as well as for MPEG-2/DVB multiplexing

and for encoding. The Telcos then guarantee

transmission and quality of service with the aid of

availability and quality parameters (e.g. errored

seconds, etc.).

W a n d e l & G o l t e r m a n n b i t s 8 2

Figure 1 :

Tributary network

Main cable







DVB analysis

DVB analysis



Network feeder

City A

Network feeder

City B

Monitoring MPEG-2/DVB services

The aim of MPEG-2/DVB system operators is to

restrict and, if possible, eliminate breakdowns of

the MPEG-2/DVB system so that customers

remain satisfied.

Alarms indicating the complete failure of active

equipment in the chain (e.g. coder or multiplexer)

are generally monitored by a closed contact relay

alarm included in the equipment itself. This allows

the operator to rapidly switch over to a back up


Nevertheless, the active equipment itself cannot

monitor the derived clock signals, tables of errored

services, rate tolerances and other information

provided within the data flow that is of the utmost

importance for continuity of service. An external

system that checks whether the MPEG-2/DVB

transport stream contains correct data within the

timing constraints defined by the operator is


This system should be based on two comple-

mentary elements: Probes with distributed

intelligence that completely analyze the

MPEG-2/DVB flow and monitoring software that

collects the reports produced by the network of

probes and concentrates the information into a

single display.

Figure 1 shows an example of a Telco operator

transporting video services for a cable service

provider to the terminals of regional networks.

The Telco operator is responsible for transmission

of the MPEG-2/DVB flow, and should also qualify

the MPEG-2/DVB signals at the points of entry to

and exit from the transport network in order to

ensure that the signal at the input is correct for

transmission and that the network does has not

introduced any errors.

Figure 2 shows an example of a cable service

provider who should verify that the signal is

correctly received at the input to the network

feeder, that the local programs (live programs, local

information recorded on servers, etc.) are correctly

inserted, and that the signal at the output of the

network feeder can be properly decoded when it

reaches the subscriber's home.

The advantages of a MPEG-2/DVB monitoring

system are:

± It alerts non-expert operators to ensure rapid

reaction to service interruptions

± It provides information required by engineers for

service action

± It indicates progressive degradations in the

system, allowing preventive maintenance

± It assists operators to locate faults so that

responsibilities can be identified

± It provides reports and records of failures for

maintaining contracts.

Jean-Pierre HeÂnot

Wandel & Goltermann,

Division 2, France



W a n d e l & G o l t e r m a n n b i t s 8 2

Figure 2 :

Cable network feeder




National tributary

Local program





Distribution interconnection







Transport Stream


Single Program

Transport Stream


Multiple Program

Transport Stream

bits 8207

Page 22

Wandel & Goltermann Extends its Digital Broadcast

Product Range to ATSC

The Advanced Television Systems Committee

(ATSC) was formed in the USA to establish

voluntary technical standards for advanced

television systems covering digital terrestrial


On December 24, 1996, the United States Federal

Communications Commision (FCC) adopted the

major elements of the ATSC Digital Television

Standard including terrestrial modulation, as well

as MPEG packetized data transport structure and

audiovisual compression.

Currently, the USA, Canada, South Korea and

Taiwan have adopted the ATSC standards for the

conversion of their terrestrial broadcast network to

the digital technology.

With DVB (Digital Video Broadcast), the ATSC

offers a second flexible standard for digital

broadcast, and is under evaluation by many

countries in South America and Asia.

After a first successful generation of products

for MPEG/DVB, Wandel & Goltermann helps

you to handle this new standard with brand new

MPEG/ATSC solutions.

ISDN Testers and Analyzers Compatible

with QSIG and CorNet



The interconnection of heterogeneous PBXs via

ISDN is now a reality within the field of extended

enterprise networks or operator networks.

In the early 1980s, the various European

manufacturers of PBX equipment worked out

platforms that allowed integration and use of

added value services with their equipment.

However, these proprietary platforms did not

allow implementation of such services on the

basis of an homogeneous network. In other

words, they only worked with PBXs from one

and the same manufacturer. Thus it was that

Siemens PN developed the CorNet


protocol for

interconnection of its own PBX equipment.

The association of the various manufacturers to

form a common working group, the ECMA

(European Computer Manufacturers Association),

has allowed enhancement of these protocols and

their development into a common, open platform

called QSIG.

QSIG is an evolutionary platform currently

standardized by the major international organiza-

tions (ECMA, ETSI




). The protocol is

supported by the majority of PBX manufacturers

and thus allows interconnection of heterogeneous

PBX equipment. In addition, QSIG is capable of

transmitting the proprietary signaling protocols

transparently over the public network using its

ªGeneric Functional Protocolº mode. Users can

thus benefit from the specific functions provided at

both ends of the network by PBXs from the same


Wandel & Goltermann has made it a priority to

provide support for such protocols in its test equip-

ment: All ISDN testers and analyzers offered by

Wandel & Goltermann are compatible with the

principal commercial proprietary protocols. The

improved implementation of the CorNet



family will be available in the new versions of the

IBT-5 (V4.01) and IBT-10 (V9.01).

Furthermore, the IBT-5, IBT-10 and IBT-20 ISDN

testers together with the Windows


PC Detailed

Decoder software already support the universal

QSIG protocol.

Hubert Chartin

Wandel & Goltermann,

Division 2, France

1 : European Telecommunications Standards Institute

2 : International Standards Organization



is a registered trademark of Siemens AG,




W a n d e l & G o l t e r m a n n b i t s 8 2

bits 8207

bits 8209

Page 23

WG IBT-5 Purchased by Telecom Organization of Thailand (TOT)

TOT, the major telecommunication services and

network provider in Thailand, has chosen Wandel &

Goltermann to supply 250 IBT-5 ISDN test sets in

the face of competition from two other potential

suppliers. IBT-5 was selected because of its

numerous features and because of the support

available through the local WG office.

TOT not only provides domestic and network

services, it also provides international tele-

communications services for other countries such

as Malaysia, Laos, Cambodia and Myanmar.

TOT is currently expanding its ISDN network at a

rate of 45000 BRA lines and 800 PRA lines per

year. It is planned that every maintenance

engineer will carry the IBT-5 as a standard part

of his tool kit. The main reason that WG could

win this tender lies clearly in full compliance

of IBT-5 features with the TOT BRA installation

test process:

± TE simulation

± BER test

± Test of services and supplementary services

± X.25 in D channel test

± Loop-back facilities

± Clear display of results.

EASY is the key word: EASY to launch a test, EASY

to understand the results! It is noteworthy that

handling of both the S and U 2B1Q interfaces was

a prerequisite to winning this order. WG is the only

manufacturer to provide such a powerful and cost

effective ISDN tester for field installation.

WG's worldwide presence means that we can offer

TOTon-site technical and after-sales support

through the WG office in Bangkok.

TELSTRA (Australia) visits Wandel & Goltermann CTS in Rennes


Ron Durkin, Director of ISDN test equipment at

TELSTRA (the Australian telecom operator):

ªWhen a telecom operator and a test equipment

supplier share their points of view, the solutions

resulting from the discussions are sure to be


James Bonnec, CEO of Wandel & Goltermann

CTS: ªIt is only possible to define efficient and

effective products if the engineers in our R&D

centers have direct, permanent contact with our


The Wandel & Goltermann Group has always

worked closely with its clients in its determination

to offer products that provide the optimum

answer to market requirements.

That's why Ron Durkin visited Wandel &

Goltermann CTS to share his experience with

the marketing and technical teams. Wandel &

Goltermann CTS, the Group's French division in

Rennes, is the center of expertise for technologies

linked to digital communication services (ISDN

and multimedia).

The visit was very beneficial for both companies.

TELSTRA was able to highlight current and future

requirements in ISDN tester characteristics. Some

of these, such as new supplementary services, will

be integrated very quickly into coming versions of

the IBT-10 and IBT-20, the devices currently used

by TELSTRA. Other characteristics will be covered

by future developments.

The visit was particularly important for Wandel &

Goltermann as it gave the company a clearer

insight into the needs of a leading figure in the field

of ISDN.

The visit was also an opportunity for Wandel &

Goltermann to present the new functions of its two

ISDN basic rate and primary rate testers / analyzers

IBT-10 and IBT-20, especially the real time analysis

function. Ron Durkin also expressed great interest

in the latest ISDN U-interface monitor, the IUM-10,

which is the only device of its kind on the market.

Alain Le HeÂnaff

Wandel & Goltermann, Division 2, France


W a n d e l & G o l t e r m a n n b i t s 8 2


Ron Durkin

(at the right) and

James Bonnec

bits 8209

Thanks to the new version of the Windows


PC Detailed Decoder software available

for Windows 95 and NT together with a

software option for the instruments

themselves, users can now analyze the

signaling frames and X.25 frames captured by

the IBT-10 and IBT-20 in real-time on their PC.

An Essential New Function for IBT-10 and IBT-20:

Detailed Real-Time Analysis

Detailed real-time analysis is a new, particularly

powerful function with the following advantages:

. Fast operation: Off-line detailed analysis first

required creation of a capture file in the IBT-10

or IBT-20, followed by transfer of this file to a

PC. Real-time decoding reduces the number

of operations required and significantly

improves productivity when it comes to

analyzing anomalies.

. Long-term recording: The frames are effectively

stored by the PC rather than by the IBT, so

recording capacity is sufficient for resolving the

type of random problem that require extremely

long recording times.

During recording, the user can interrupt the

real-time display at any time in order to look at

previously received frames without interrupting

storage of frames in the background. The real-time

option provides the same facilities as for off-line

analysis: Detailed and summarized displays of

frame content, filters, etc.

This new version of the Windows


PC Detailed

Decoder software, in addition to the real-time

analysis function, retains all the functions that

contributed to the success of previous


. Transfer of analysis files and results from the

IBT-10, IBT-20 and IST-15 (ISDN and LAN

Cable Tester)

. Off-line detailed analysis of ISDN signaling

from IBT-10 and IBT-20 capture files

. Editing and printout of result files (BER tests,

services tests, cabling tests, etc.)

. Export of results to standard office software

such as spreadsheets, word processors, etc.

This new version is an important step in the

development of this software with the aim of

improving the efficiency of maintenance teams.

It provides:

. Innovative functions for investigating


. A universal user interface for all ISDN test

equipment from Wandel & Goltermann.

The new version of the Windows


PC Detailed

Decoder software is provided free with the

real-time analysis option for the IBT-10 and IBT-20


Yann HerveÂ

Wandel & Goltermann,

Division 2, France


W a n d e l & G o l t e r m a n n b i t s 8 2


bits 8210

From d.c. to light is the coverage aimed for

by the WG Safety Test Solutions Group. This

part of the Company is concerned with personal

safety in electromagnetic fields. The new

ETM-1 Tesla Meter is a further step towards

this goal.

Safety in Constant Magnetic Fields:

ETM-1 Tesla Meter Extends the WG Measurement Spectrum

Down to 0 Hz

Constant magnetic fields do not just occur where

permanent magnets are in use ± just refer to the

figure below for evidence. Wherever d.c. flows, it is

surrounded by a constant magnetic field. This is

not a problem in the case of a battery-powered

flashlight. However, such magnetic fields are

most definitely not negligible when it comes to

aluminum smelting or railroad locomotives where

high-ampere currents flow, or in medical equip-

ment such as nuclear spin tomographs, where the

magnetic field is needed for diagnosis.

Wandel & Goltermann now offers a test instrument

for these fields. The ETM-1 Tesla Meter measures

magnetic fields of up to 2 Tesla. The probe, a

tri-axial Hall-effect sensor, is isotropic (non-direc-

tional) so that it does not need to be oriented in

the direction of the field lines. The overall field

strength is displayed by the instrument's LC

display. All functions can be set using just five keys,

making the ETM-1 exceptionally user-friendly.


W a n d e l & G o l t e r m a n n b i t s 8 2


Top: The Wandel & Goltermann range of test equipment

for personal safety in electromagnetic fields. The frequency

spectrum ranges from constant magnetic fields up to

alternating electric fields of up to 60 GHz.

Left: The new ETM-1 Tesla Meter and probe

The ETM-1 is equipped with a serial interface,

allowing measurement results to be transferred

on-line to a computer for further processing. A

further advantage is provided by the zero-field

chamber supplied with the instrument. This can be

used to perform zeroing without the effects of the

earth's magnetic field or other interfering fields.

Holger Schwarz

Wandel & Goltermann,

Safety Test Solutions,


bits 8211


Despite all the advantages of an OTDR

(optical time-domain reflectometer) it is still

not possible to do without an OLTS (optical

loss test set) with generator and receiver. In

branched networks, for example, fiber loss

cannot always be measured with an OTDR.

Checking for crossed connections requires a

ªtwo-endedº measurement using generator

and receiver. Further arguments in favor of

OLTS are the low cost and ease of use.

Wandel & Goltermann offers a complete

range of optical hand-held instruments from

high performance and pocket-size families

designed for such applications.

Measurements on Optical Networks:

Two Paths Leading to Reliable Results

The high performance family:

Outstanding specifications for

outstanding results

This family of instruments is characterized by

versatility, ease of use and accuracy. It includes

power meters, sources and attenuators. The

wide dynamic range means that many appli-

cations are possible. In the power meter

range, the upper limit for level measure-

ment is +26 dBm, ideal for measure-

ments on CATV systems or EDFA.

The lower limit is ±80 dBm, but

no sacrifices are made in terms

of accuracy.

The OLA-15 Variable Attenuator

is useful for checking laser safety

cutoffs and receiver sensitivity in

optical communications systems.

The dual laser in the OLS-15 Source

makes it easy to measure at two wave-

lengths. A single keystroke is all that is

needed to make simultaneous measurements at

1310 nm and 1550 nm using one and the same

output (TWINtest). And there's no need to check

that the wavelength is set correctly, as this is done

automatically. Fiber identification is also automatic,

so it's fast and reliable: As soon as the end of the

correct fiber is placed close to the power meter

input, a signal will be heard.

All these instruments run from dry or rechargeable

batteries or from the AC line.

± Precision guaranteed by quality adapter system

± Dual laser feeding a single output

± Automatic wavelength recognition

± Audible fiber identification without connecting

the fiber

± RS 232 interface for direct printout of results and

for remote control

W a n d e l & G o l t e r m a n n b i t s 8 2







High performance

solutions for single-

mode applications:

Versatile, thanks to

wide dynamic range,

easy to use, thanks

to automatic test












Dual LED

Dual FP laser














Output level


±20 to ±40



Fiber core


50, 62.5,









FC/PC, fixed




Weight (incl.


200 g

500 g

Table right:


specifications of the

Optical Sources

Pocket-size family:

Pocket-size format for a

pocket-size price

This family is characterized by budget price,

rugged construction and automatic features that all

but eliminate operator error. It includes LED and

laser sources and power meters. The unique

push-pull system that eliminates costly and

time-consuming adapter changes and the simple,

three-button operating concept make for fast


The OLS-5 and OLP-5 instruments are designed

for multimode applications in the datacom/LAN

segment. New additions to the range are the

OLS-6 and OLP-6 for single-mode applications.

They have the following features in common:

± Pocket-size format

± Extremely easy to use

± Auto-l recognizes wavelength


± TWINtest for multimode fibers

± Novel push-pull system for fast connections

without adapter changes.

OMK: Application-oriented complete solutions

Users will find all they need for measurements right

up to a complete loss test set at an affordable price

in one of the OMK Optical Measuring Kits.

MT-32: Made to measure

For those who need to put together their own

combination of test equipment, the MT-32 Kit

Bag is ideal for equipment from both families. The

advantages offered by the fact that both families

operate together perfectly, e.g. for automatic

wavelength recognition or fiber identification, can

be utilized to the full.

Leyla Mora

Wandel & Goltermann,

Division 1, Germany


W a n d e l & G o l t e r m a n n b i t s 8 2










Pocket-size solutions for

multimode and single-mode

applications: The budget

alternative with three-

button operation

Put together your own package

for your own special application:

The instruments from the high

performance and pocket-size

ranges can be combined

as required.

Want to test out
operation? Visit our
Internet site at


Test solutions for fiber

optic networks ± a new

brochure with example

applications and

technical background


Table left:

Comparative specifi-

cations of receivers

and test kits


High performance

Complete solutions





OMK-14 (OLS-/OLP-15)

OMK-15 (OLS-/OLP-/OLA-15)









Single-mode (e.g. CATV, EDFA)

Measuring range (dBm)

±60 to +5

±65 to +10

±70 to +20

±80 to +15

±60 to +26

Display in

dBm, dB

dBm, dB, mW, mW

Power supply

Dry batteries, NiCd cells

AC adapter, dry batteries, NiCd cells (rechargeable internally)

RS 232 interface for

remote control and




Data storage



Audible fiber identification



Frequency modulation

270 Hz, 330 Hz, 1 kHz, 2 kHz

Adapter system


High-precision, exchangeable

Weight (incl. batteries)

180 g

500 g

bits 8212

Page 28

WG RTU-500 is a remote test probe platform

providing both passive and interactive testing

of telecommunications networks across

unmanned remote sites.

WG RTU-500 Remote Test Unit:

A Valuable Business Investment

Cost reduction is a major consideration for service

providers as the remote test market becomes

increasingly competitive, with labor being a cost

for which significant savings can be made by

deploying the latest remote test technology

effectively. In addition to cost-cutting, service

providers are under increasing pressure to improve

customer support.

The days of the expert engineer in the field are

numbered. As roads become more congested, it

takes longer and becomes more expensive to

transport experts to problem sites so that they can

support local technicians ± and once a key

engineer is preoccupied at one site, he is no

longer available to help if critical problems arise

elsewhere on the network.

The future lies with test management software

controlling remote test equipment. Expertise will

be concentrated in national operations centers

(NOCs) where quality-of-service and fault

management will be carried out. The remote test

units (RTUs) located at distant sites may be

robust, portable, handheld instruments under

central management control, but the current trend

is towards static probes (such as WG RTU-500)

which are permanently installed in the network and

have a wide range of fault and service

management applications.



W a n d e l & G o l t e r m a n n b i t s 8 2

Typical system


Test Set Manager software

controlling an RTU-500




Terminal server













Terminal server



Network under test

2 Mbit/s

2 Mbit/s

Page 29

The centralization of expertise and the use of so-

phisticated software has beneficial cost implica-

tions in the long-term since ultimately fewer ex-

perts will be necessary at any one location. Man-

agement software (such as the WG TSM Test Set

Manager) hides the complexity of the testing, so

the knowledge of the technicians and engineers

routinely carrying out network testing no longer

needs to be so specialized, with consequential re-

ductions in operational and training costs.

To appreciate the savings that can be made, we

can look at the way two employees in different test

environments deal with the problems they face in a

typical day.

Without centralized network management, Peter

has to spend time driving to one or more sites to

isolate the cause of a problem, carrying with him a

range of spares so that he can rectify the fault.

With 40-60% of the problems turning out to be

false call-outs ± mostly due to human misjudgment

± much of his day is wasted and very little time is

spent on producing information for better network


Meanwhile, John has made use of the test

management software to pinpoint the exact

location of a fault and to identify the problem before

a field service technician is dispatched. Because

John is able to control the RTUs, he can carry out

tests from the comfort of his PC and establish the

cause of the problems across both national and

international networks. When the field service

technician leaves his office to make site visits, he

can take with him the exact parts he needs to do

the repairs at various locations. His day is spent

more effectively with no wasted travel time since

he drives directly to the problem sites without

having to carry out additional intermediate tests.

The field service technician does not need to

W a n d e l & G o l t e r m a n n b i t s 8 2














Technician using portable test

equipment at various sites

Drives to first site to isolate a reported


No problem found (false call-out), so

returns to main office for next job ticket.

Drives to second site to isolate a

reported problem.

Problem appears to be at the customer

premises, so drives there.

Hurried sandwich lunch in van followed

by further testing to find problem.

Problem identified, but spare part not

in van. Returns to stores.

Collects part and drives back to site of

fault to complete repair.

Sets up end-to-end test and generates

test results for management report.

No further problems found, so drives

back to main office.


Engineer based at an NOC using

a remote test management system

Uses management software to identify

problem, fault location and spares required.

Despatches technician to remote site with

spare parts to rectify problem.

Sends technician mobile job ticket for next


Runs quality-of-service check and

generates management reports.

Lunch in staff canteen followed by

refreshing stroll in NOC grounds.

Identifies fault at a manned site and

advises the local technician by telephone.

Runs performance validation reports for


Executes interactive tests over multiple

customer circuits during commissioning.

Studies test results and prepares

maintenance plan.

Working in different

test environments

carry a selection of spares with him. Savings are

made in the stores, too, where an integrated stock

control system coupled to the job ticketing system

can deliver the required parts for rapid dispatch.

John spends much less of his day finding and

identifying faults than Peter does, which increases

the time he has available to tackle other problems.

Meanwhile the field service technician achieves a

higher productivity through efficiency. John also

produces performance validation information for

customers, with quality-of-service and fault reports

providing improved service management. In the

long-term this leads to a higher-quality service with

further decreases in operating costs. The result?

Happier customers!

The test management software at the NOC

includes a remote operation feature with a visual

faceplate interface. This not only enables John to

see an on-screen GUI representation of the front

panel of the remote instrument, but provides him

with user-friendly interactive control of the instru-

ment so that he can produce instant test results.

More than one instrument of the same type can

be operated with a single version of software from

the same PC workstation by running more than

one instance of the software. This creates multiple

windows, each dedicated to a particular instru-

ment. This capability enables John to connect up

to multiple problem sites and switch his attention

between them effortlessly, resulting in speedier

solutions to customers' problems.

The RTU-500 remote test unit can be readily

integrated into existing or new test systems. The

unit comprises a sub-rack which can hold up to

eight Remote Test Application Modules (RTAMs),

two power supply modules, and an optional alarm

module. The RTAMs provide the test capability of

the system, with functions ranging from passive,

non-intrusive alarm monitoring to interactive,

out-of-service testing. Initial applications include

remote interactive testing of primary rate digital

circuits (E1), network link performance analysis,

remote network end-to-end BER testing, digital

PCM frame simulation and alarm generation, audio

signal generation, traffic generation, and signaling


Depending on the circumstances, the initial capital

outlay for portable test equipment or RTUs and

test management software could be similar.

However, the notable savings that can be made

by the use of the latter become apparent within

6 months of installation. When existing tele-

communications operators are considering

upgrading their existing test systems or new Telcos

are looking at state-of-the-art management

systems, they should remember that their preferred

remote test solution should not only be seen as a

necessary operational investment COST, but more

importantly as a business VALUE investment.

Kate Fereday

Remote Test Solutions

Wandel & Goltermann,

Division 4, UK


W a n d e l & G o l t e r m a n n b i t s 8 2




New approach





Network management


Element management


Element layer

Traditional approach

Seen as business VALUE investment

Quality of Service

test solution

Test data acquisition

Fault and maintenance

test solution

± Directly enhances customer's

business portfolio

± Enables support revenue


± Creates marketable


± Core building block

± Test engine

± Enabling value solution

± Improves operational process

± Reduces operational costs

± Improves efficiency of


Seen as a necessary operational investment COST

bits 8213

Page 31

PA/PFA Family ± A Worldwide Success!

1998 has seen the release of the 5th generation

PA / PFA instruments, the most powerful 2 Mbit/s

test solutions to date, capable of even more ad-

vanced testing than ever before including jitter

generation and measurement. First released five

years ago, customer feedback has been used to

evolve the instrument, building on the strengths of

flexibility, power, quality and reliability. This quickly

saw the PA/PFA Family of five instruments

established in the market to meet the varied

requirements of our customers all over the world.

With each generation, more new test solutions and

user benefits have been added and this, together

with a growing list of downloadable software

options and accessories, has ensured that the

family stays at the forefront of test technology. With

almost all PTTs and Telcos using 2 Mbit/s pipes as

the standard bearer for services such as Frame

Relay and GSM, the importance of our software

options, which allow PA/PFA instruments to test

these and many more services, is underlined.

Together with the extensive range of upgrades that

we also provide, this means older generation

instruments can be brought up to the same

standard as the latest generation, so customers

can actually enlarge their PA/PFA family of testers

as their business grows and develops.

Sold in over 70 countries worldwide and originally

targeted at major PTTs and Telcos, deregulation in

the global communications market has seen the

demand for this type of test solution increase

dramatically. Almost all the major PTTs and Telcos

have now standardized on the PA/PFA family and

the new Telcos, like these traditional PTTs, have

made the PA/PFA family their preferred choice.

Continuing satisfaction amongst these customers

has ensured that shipments have increased year

on year, a trend which is set to continue as we

approach the next millennium.

Colin Chapman

Product Marketing Engineer,

with thanks to the many contributors.

Wandel & Goltermann,

Division 4, UK



W a n d e l & G o l t e r m a n n b i t s 8 2

1st Generation PA-10,

circa 1993

5th Generation PFA-35,

July 1998

bits 8214


W a n d e l & G o l t e r m a n n b i t s 8 2


Digital signal processing (DSP) is a flexible and

powerful system for manipulating and

measuring signals. DSP algorithms allow us to

measure signal parameters which would be

extremely difficult or even impossible with

conventional analog signal processing


Wandel & Goltermann were quick to realize the

potential benefits of DSP and the DLA-9D was

one of the world's first data line analyzers to

harness the power of DSP. The TSA-40 builds

on this success.

Digital Signal Processing in Analog Products

DSP technology is now becoming commonplace

in everything from hi-fi systems to motor

controllers. These products all implement simple

algorithms to filter and process signals over a

narrow bandwidth. A high performance, multi-

function data line analyzer set like the DLA-9D

places great pressure on DSP and a large amount

of development was required to produce the

specialist algorithms required for the various


The choice of DSP device used in the instrument is

also extremely important. WG instruments are

battery powered and portable and yet provide a

high degree of accuracy. The specifications of the

DLA-9D are based on that of the well established

DLA-9; a signal range of over 100 dB and a

bandwidth of 300 kHz with filters specified to over

50 dB places great demands on a DSP processor.

Many of the measurement modes (such as events)

require the DSP to be doing several things at once

(such as measuring the level and frequency of a

tone and looking for glitches while implementing a


The more 'bits' the better when it comes to DSP,

every bit gives an extra 6 dB of range. Engineers

have a choice of 16, 24 or 32 bit floating point DSP

devices. 16 bit parts are used in portable

applications while 32 bit parts are used when

extremely high performance is required. WG used

a 16 bit part in the DLA-9D design as it gives the

best blend of power, accuracy and cost

effectiveness. The 16 bit word size coupled with

powerful DSP algorithms can handle large

dynamic ranges over a wide bandwidth.

It is also important to realize that the DSP device

cannot be chosen in isolation. Careful

consideration needs to be given to other

components in the DSP subsystem. The data

conversion (primarily the ADC) sets the level of

performance that can be achieved.

The DLA-9D instrument uses a high performance

14 bit ADC coupled with a Texas Instruments C50

DSP device. This is a 16 bit fixed point processor

with a large amount of on-chip RAM. This allows

programs to run at high speed and also reduces

EMC emissions. The DSP and ADC are closely

coupled with a logic array (a FPGA) which

generates all the timing signals required by the

circuitry. A high performance, low noise, switched

gain amplifier scales the ADC input signal to give

maximum dynamic range. This is all done

automatically under DSP control.

Block diagram of

DLA-9D DSP system








to digital










Control signal




to main


The circuitry was also designed for low power

operation and parts of the circuitry can be

switched off when not required. Internal

diagnostic and calibration procedures ensure

reliable operation.

All the code for the signal measurements and line

characterization was developed in-house by WG

engineers and is hand coded to give compact and

fast running programs. Techniques such as

extended precision and error-spectral shaping

have been used to give a high degree of accuracy.

The flexibility of the design means that future

options can be developed which allow specialist

measurements to be developed quickly without

any changes to hardware. In fact, many can be

done as a closed case field upgrade!

Engineers at WG are continuing to develop new

and more powerful algorithms for the

characterization of analogue lines. The continued

increase in performance of DSP devices means

that detailed analysis of a line over a wide

bandwidth can be made in real time. We have

already developed systems which use extremely

high performance floating point DSPs, which has

allowed us to develop new techniques that may be

used in future analog products.

Paul Brooks

Wandel & Goltermann,

Division 4, UK


What is DSP?

DSP performs operations such as filtering and level measurement

using mathematical operations on digital data. This data can come

from any source and after conversion through an ADC it is

manipulated by algorithms which use addition, multiplication and

other operations to process the signal.

Once the signal has been converted into digits it can be processed

to a specified degree of accuracy and is not subject to errors

associated with ordinary `analog' filters such as component

tolerance, drift and noise. Furthermore, one piece of DSP

hardware can perform many different filtering and processing

operations by simply changing the code. Analog circuitry would

need to be redesigned each time a new requirement arose.

Many types of processing would simply be impossible in circuitry

realized from op-amps and other analog parts. DSP can

perform any signal processing operation that can be described


Typical DSP parts can perform over 50 million operations per

second while state-of-the-art parts often operate at 1000 million

operations per second.

For further information see ªA Simple Approach to Digital Signal

Processingº by C Marvem and G Ewers, TI Mentors series,

ISBN 0-904047-00-8.

Application Notes from

Wandel & Goltermann

± free of charge!

Order your copies

using the fax from

or via

bits 8260

bits 8259

bits 8258

bits 8255

bits 8257

bits 8256


W a n d e l & G o l t e r m a n n b i t s 8 2


When using the telephone, echoes of the

talker's voice exist for both national and own-

exchange calls. This article outlines the effect

of speech echo within telecommunications

networks and how its effect may be controlled.

The Disturbing Effect of Echo

Local access network cable tends to be a two-wire

(2W) metallic pair between the serving digital

exchange and the user's telephone. However,

when a connection is established, four-wire (4W)

operation is used and to interface the local

access 2W to the 4W switched circuit, the local

exchange line card uses a two-four wire

(2W/4W) conversion circuit. Hence, for both

own-exchange and national calls an element of the

connection includes a 4W section as illustrated in

Figure 1.


Sidetone describes the receipt of the talker's

voice at their own telephone and is a desirable

effect because it makes the instrument appear

ªaliveº. However, sidetone loudness can

affect transmission performance as perceived by

the user:

Low Sidetone Loudness:

Talkers may perceive the connection to be poor

and raise their voice. This may cause listener to

believe the talker to be too loud, and so hold the

telephone receiver away from their ear. When the

listener starts to talk, their mouth may be too far

from the microphone resulting in a low level of

speech being received by the other party who

starts to talk even louder!

High Sidetone Loudness:

Talkers will tend to lower their voice, resulting in

the listener considering the talker to be too quiet.

When the listener starts to talk, their voice will likely

be raised, resulting in a high level of speech being

received by the other party who starts to talk even

more quietly!

Sources of Echo

For an established call, when the talker speaks,

signals carrying the speech travel through the

connection making the transition between 2W and

4W operation. The loss through the far-end line

card's 2W/4W converter across the 4W section's

ªgoº and ªreturnº transmission directions is

critical. If this loss is not of a high order, then a

significant level of the talker's speech may be

introduced from the ªgoº into the ªreturnº

direction. This spurious signal then travels back

down the connection to the talker's telephone

where it may enhance the near-end telephone

sidetone. This is undesirable because it may result

in the user perceiving and reporting poor

transmission quality. Figure 2 illustrates this

concept further.

The prime source of talker echo is the far-end

2W/4W converter. The loss presented by a 2W/4W

converter to the 4W circuit (termed transhybrid

loss) should be kept above 20 dB within the 300 to

3400 Hz bandwidth. ITU-Trecommend that a

weighted measurement is made within the

telephone bandwidth; this loss is referred to as

echo loss (G.122 and G.131).

Signal Propagation Delay and Echo

Signals conveyed from the one end to the other

end of a telephone connection experience a finite

time delay as they travel through its

interconnected circuits. This time delay is referred

to as the signal propagation delay or one-way


If significant talker echo is present, then the

signal propagation delay within the echo signal

path may further contribute towards the

perception of poor performance. If the

propagation delay of the echo signal exceeds

more than a few milliseconds, the talkers may

notice echoes of their own voice as the echo

starts to become distinct from the sidetone.

If propagation delay is significantly increased,

Figure 1: Example of a

national call involving

several switching

stages. For local

exchange calls only

one digital switch is






Digital switches

Long 4-wire switched section

within a national call











Page 35

speech echo may become too disturbing,

making it impossible to use the telephone


The subjective effect upon the talker is a function

of speech echo level and the time delay for the

speech echo to be returned to the talker (round-trip

delay). This time delay is composed of the one-way

delays to and from the point where the speech

signal is reflected to cause the echo. Figure 3

illustrates the concept of echo.

For telecommunications networks where the echo

loss of line card 2W/4W converters can be

expected to be above 20 dB, the one-way delay

should typically be under 23 ms (46 ms round-trip)

to prevent users from experiencing significant

disturbance due to speech echo. This assumes

the use of telephone instruments that conform to

ITU-T recommended Send and Receive Loudness

Rating values.

Sources of One-Way Delay

The sources of time delay within

telecommunications networks include:

Digital switches (exchanges):

± 1.7 ms for a local exchange

± 0.75 ms for a transit exchange


± Metallic at 4 ms/km

± Optical fiber at 5 ms/km

± Loaded cable 45.5 ms/km

It is possible that a national connection could

involve several switching stages and diverse

routings leading to significant one-way delays.

Echo and Its Effect on Other Services

This article has concentrated on telephony and

the effect of echo and delay. However,

increasingly the effects of echo in association

with round-trip delay have become relevant for

facsimile (FAX) transmission and data transfer

using modems. For handshaking purposes, some

types of FAX machine and modem calculate the

round-trip delay for the connection that has been

established. If the time delay becomes too

excessive, it is possible that some machines may

refuse to work or become ªlockedº.


Transhybrid losses and propagation delays are not

easy to assess unless access for measurement is

available on an end-to-end basis. However, if

access is available at the 2 Mbit/s level (0 dBr

point) within an established connection, a WG

PFA-35 could be used to provide an indication of

the far-end's transhybrid loss (level and frequency


Concerning round-trip delay, this is difficult to

measure in most cases for a ªliveº connection.

However, if the transmission systems and

switches that are used to collectively form a

connection can be permanently connected for the

purposes of testing, the WG PFA-35 delay

mode could be used to provide an indication of the

round-trip delay from the point where the

instrument is connected (0 dBr, 2 Mbit/s point). If

this test point is located at the serving local

exchange, then an estimate of the round-trip

delay could be measured.

Mike Parkin

Wandel & Goltermann,

Sales, UK


W a n d e l & G o l t e r m a n n b i t s 8 2






Speech signal path

Speech echo path

4-wire switched section

Echo may be

perceived by










Figure 2: Concept of

speech and echo


Figure 3: Concept of

echo loss and round

trip delay

Speech signal path

0 dBr

Speech echo and

round trip delay

(546 ms) paths

Echo may be

perceived by




(420 dB)

VOL. 3



bits 8234

bits 8221

bits 8222

bits 8231

bits 8232

bits 8235

bits 8233

bits 8236