1

application note

2.048 MbpsTechnology Basics andTesting Fundamentals

2

T he demand for high-quality 2.048 Mbps circuits requires thorough installationtesting and consistent maintenance and circuit analysis. To provide clean, error-freetransmissions, the 2.048 Mbps installation and maintenance personnel who test theperformance of these circuits demand reliable equipment for their testing needs.

TTC supplies a range of test sets that are ideal for 2.048 Mbps installation, acceptance testing,routine maintenance, and fault isolation, all of which are integral to providing a quality2.048 Mbps service.

This Application Note first describes 2.048 Mbps fundamentals and the impairmentsthat can degrade 2.048 Mbps services. It then provides examples of applications for

• In-service monitoring• Out-of-service testing

Each application section includes guidelines on how to accurately interpret the testresults for rapid trouble shooting and fault isolation.

Those familiar with 2.048 Mbps basics may wish to turn immediately to the testingapplications beginning on page 9.

2.048 MbpsTechnology Basics andTesting Fundamentals

3

2.048 Mbps Basics

A 2.048 Mbps circuit provideshigh speed, digital transmissionfor voice, data, and videosignals at 2.048 Mbps.

2.048 Mbps transmission systems arebased on the ITU-T specifications G.703,G.732 and G.704, and are predominant inEurope, Australia, Africa, South America,and regions of Asia. Due to an increase indemand for global communications inrecent years, 2.048 Mbps installations inNorth America have risen sharply, and existalongside the standard T-Carrier systems.

The 2.048 Mbps standards arenow firmly established for transmissionsystems and are used by telecommunicationsnetwork suppliers, international carriersand end users. The primary use of the2.048 Mbps is in conjunction withmultiplexers for the transmission ofmultiple low speed voice and data signalsover one communication path rather thenover multiple paths. Figure 1 shows atypical system.

Figure 1:

A typical 2.048 Mbps transmission system

DTE 3

DTE 2NTE

256kbps

2 MbpsNetworkMUX

DTE 1

128kbps

64kbps

NTEMUX

DTE 3

DTE 2

256kbps DTE 1

128kbps

64kbps

The 2.048 MbpsLine Code

The most common line code used totransmit the 2.048 Mbps signal is known asHDB3 (High Density Bipolar 3) which is abipolar code with a specific zero suppressionscheme where no more then three consecu-tive zeros are allowed to occur. The HDB3line code is recommended for 2.048 Mbpssignals by ITU-T Recommendations G.703,and it is defined in Annex A to Recommen-dations G.703.

In some instances straightforwardbipolar AMI (Alternate Mark Inversion)coding with no zero suppression isalso encountered.

In the following paragraphs, we willfirst review the AMI coding format, whichrepresents the simplest version of bipolarline code. We will then move on toexplaining the 2.048 Mbps HDB3 line code,which essentially is a variation of AMI wherea high density of pulses is ensured byapplying a zero suppression algorithm.

AMI or Bipolar Line CodeIn the AMI coding format, a binary

one (mark) is represented by a square pulsewith a 50% duty cycle and a binary zero(space) is represented by the lack of pulse,i.e., 0 Volts. Since successive pulses (i.e.,marks) alternate in polarity the line code istermed AMI (Alternate Mark Inversion).

HDB3 Line CodeDespite its numerous advantages,

AMI coding has one very significantshortcoming. Since signal transition are theonly way for 2.048 Mbps equipment torecover the timing information, long stringsof zeros with no pulse transition in the datastream may cause the equipment to losetiming. Hence AMI coding puts strictlimitations on the zero content of the datatransmission in the 2.048 Mbps system.

One solution to this problem is touse a coding scheme that suppresses longstring of zeros by replacing them with aspecific sequence of pulses, which can berecognised and decoded as zeros by2.048 Mbps equipment. HDB3 is one suchcoding scheme upon which the 2.048 Mbpsindustry standardised.

4

How HDB3 WorksThe HDB3 signal is a bipolar signal,

where sets of 4 consecutive zeros arereplaced by a specific sequence of pulses andthe last pulse is coded as a violation. Thisensures that the 2.048 Mbps signal has a highdensity of pulses and no more then3 consecutive zeros. Table 1 shows the rulesfor zero substitution using the HDB3 codingscheme.

An example of how these rulesare applied to an AMI signal is shown inFigure 2.

It is important to note that:

1. The 4th zero is always coded as a violationpulse.

2. The 1st zero may be coded as a“balancing” pulse to ensure thatsuccessive HDB3 violation pulses areof opposite polarity, so that the net DCcomponent of the signal remains zero.

Hence the HDP3 code eliminatesall the limitations on the zero content ofthe signal transmitted in the 2.048 Mbpssystem, while preserving all the advantagesof AMI coding.

Table 1:

HDB3 substitution rules

1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 0 0 1 0 0 0 0 0 0 0 0Binary

AMI

HDB 3

V V

B BV V

V = Pulse violating the AMI sequenceB = Additional pulse ensuring that the consecutive V pulses are of opposite polarity

Number of Bipolar Pulses (Ones)Since Last Substitution

Polarity of Odd EvenPreceding Pulse

- 000- +00+

+ 000+ -00-

Figure 2:

Example of a HDB3 signal

5

Frame 0

Multiframe Alignment Signal (MFAS) pattern - 0 0 0 0X = Spare parts (set to 1 if not used)Y = Remote Alarm (set to 1 to indicate loss of multiframe alignment)A B C D = Signaling bits

NOTE: Even numbered frames contain the FAS pattern in time slot 0

Bits

1 2 3 4 5 6 7 8

Frame 1 Frame 2 Frame 3 Frame 15

0 0 0 0 X Y X X

Bits

1 2 3 4 5 6 7 8

A B C D

Channel 1(TS-1)

A B C D

Channel 16(TS-17)

Bits

1 2 3 4 5 6 7 8

A B C D

Channel 15(TS-15)

A B C D

Channel 30(TS-31)

TS 0 TS-16 TS-31 TS 0 TS-16 TS-31 TS 0 TS-16 TS-31

Figure 4:

The 2.048 Mbps TS-16 multiframe format

Time Slot0

Frame Alignment Signal (FAS) pattern - 0 0 1 1 0 1 1Si = Reserved for international use (Bit 1)Sn = Reserved for notational useA = Remote (FAS Distant) Alarm - set to 1 to indicate alarm condition

Time Slot1

Time Slot2

Time Slot31

One 2.048 Mbps Frame

125 µs

Frame containingframe alignmentsignal (FAS)

Frame notcontaining framealignment signal

Bits

1 2 3 4 5 6 7 8

Si 0 0 1 1 0 1 1

Si 1 A Sn Sn Sn Sn Sn

Figure 3:

The 2.048 Mbps framing format

The 2.048 MbpsFraming Format

The 2.048 Mbps signal typicallyconsists of multiplexed data and/or voicewhich requires a framing structure forreceiving equipment to properly associate theappropriate bits in the incoming signal withtheir corresponding channels. Figure 3shows the framing for the 2.048 Mbps signalas defined in ITU-T Recommendation G.704.

As can be seen in Figure 3, the2.048 Mbps frame is broken up into 32timeslots numbered 0-31. Each timeslotcontains 8 bits in a frame, and since thereare 8000 frames per second, each timeslot corresponds to a bandwidth of8 x 8000 = 64 kbps.

Time slot 0 is allocated entirely to theframe alignment signal (FAS) pattern, aremote alarm (FAS Distant Alarm) indicationbit, and other spare bits for international andnational use. The FAS pattern (0011011)takes up 7 bits (bits 2-8) in timeslot 0 ofevery other frame. In those frames notcontaining the FAS pattern, bit 3 is reservedfor remote alarm indication (FAS DistantAlarm) which indicates loss of framealignment when it is set to 1. The remainingbits in timeslot 0 are allocated as shown inFigure 4.

If the 2.048 Mbps signal carries novoice channels, there is no need to allocateadditional bandwidth to accommodatesignalling. Hence, time slot 1-31 are avail-able to transmit data with an aggregatebandwidth of 2.048 Mbps - 64 kbps (TSO)= 1.984 Mbps.

6

Figure 5:

The 2.048 MbpsCRC multiframe format

FrameNumber

Bits 1 to 8 (TS 0) of the FrameSub-multiframe(SMF) 1 2 3 4 5 6 7 8

01234567

89101112131415

C1

0C2

0C3

1C4

0

01010101

0A0A0A0A

1Sn

0Sn

1Sn

1Sn

1Sn

0Sn

1Sn

1Sn

1Sn

0Sn

1Sn

1Sn

1Sn

0Sn

1Sn

1Sn

0Sn

0Sn

0Sn

0Sn

C1

1C2

1C3

Si

C4

Si

01010101

0A0A0A0A

1Sn

0Sn

1Sn

1Sn

1Sn

0Sn

1Sn

1Sn

1Sn

0Sn

1Sn

1Sn

1Sn

0Sn

1Sn

1Sn

0Sn

0Sn

0Sn

0Sn

I

II

Multiframe

C1, C2, C2, and C4 = Cyclic Redundancy Check BitsCRC Multiframe Alignment Signal 0 0 1 0 1 1Sn = Reserved for notational useA = Remote (FAS Distant) Alarm - set to 1 to indicate alarm condition

If there are voice channels on the2.048 Mbps signal, it is necessary to takeup additional bandwidth to transmit thesignalling information. ITU-T Recommen-dation G.704 allocates time slot 16 for thetransmission of the channel-associatedsignalling information. This is explained inthe next section.

The 2.048 MbpsTS-16 MultiframeFormat

The 2.048 Mbps can carry up tothirty 64 kbps voice channels in time slot1-15 and 17-31.

Voice channels are numbered1-30; voice channels 16-30are carried in time slot 17-31.

However, the 8 bits in time slot 16 arenot sufficient for all 30 channels to signal inone frame. Therefore, a multiframe structureis required where channels can take turnsusing time slot 16.

Since two channels can send theirABCD signalling bits in each frame, a totalof 15 frames are required to cycle throughall of the 30 voice channels. One additionalframe is required to transmit the multiframealignment signal (MFAS) pattern, whichallows receiving equipment to align theappropriate ABCD signalling bits with theircorresponding voice channels. This results inthe TS-16 multiframe structure where eachmultiframe contains a total of 16 2.048 Mbps,numbered 0-15. Figure 4 on the previouspage shows the TS-16 multiframe format forthe 2.048 Mbps signal as defined by theITU-T Recommendation G.704.

As can be seen in Figure 4, timeslot 16 of frame 0 contains the 4-bit longmultiframe alignment signal (MFAS) pattern(0000) in bits 1-4. The “Y” bit is reserved forthe remote alarm (MFAS Distant Alarm) whichindicates loss of multiframe alignment when itis set to 1.

Time slot 16 of frames 1-15 containsthe ABCD signalling bits of the voice channels.Time slot 16 of the nth frame carries thesignalling bits of the nth and (n+15)th voicechannels. For example, frame 1 carries thesignalling bits of voice channels 1 and 16,frame 2 carries the signalling bits of channels2 and 17 etc.

It is also important to note that theframe alignment signal (FAS) is transmitted intime slot 0 of the even numbered frames.

We have thus explained how framealignment and channel associated signallingare achieved in 2.048 Mbps transmission.(Alternatively, time slot 16 may also be usedfor common channel signalling applicationssuch as primary rate ISDN). It must be noted,however, that the 2.048 Mbps framing andTS-16 multiframing structures discussedso far do not provide any built in errordetection capabilities, which could be usedto determine the error performance of the2.048 Mbps system on an in-service basis.This capability is provided by the CRC (CyclicRedundancy Check) multiframe structure asexplained in the next section.

7

Table 2:

Various 2.048 Mbps frame and multiframe formats

*Note:The two multiframe structures are not related, and need not be aligned with each other in any way.

The 2.048 Mbps CRCMultiframe Format

This section describes the specificsof the 2.048 Mbps CRC Multiframe format.To find out how CRCs provide the enhancederror performance monitoring capabilitiesmentioned above, refer to the “CRC ErrorAnalysis” section (page 9) under Application#1, In-Service Analysis of Live Traffic.

The 2.048 Mbps CRC Multiframestructure as defined by ITU-T Recommen-

dation G.704 is shown in Figure 5 on theprevious page.

The CRC Multiframe consists of16 frames (numbered 0-15) which aredivided into two sub-multiframes (SMF-1and SMF-11) of 8 frames each. The 4-bitlong CRC word associated with each sub-multiframe, SMF(N) is inserted into thenext sub-multiframe, SMF(N+1). The CRCbits take up the 1st bit of time slot 0scontaining the 7-bit FAS (Frame AlignmentSignal) pattern. The CRC Multiframe

alignment signal uses the 1st bit of timeslot 0s not containing the FAS pattern.(See Figure 5).

Combining the TS-16 andCRC Multiframe Structures

A 2.048 Mbps signal may comein a number of different formats, dependingon which of the above frame and multiframestructures are implemented in the 2.048Mbps system. Table 2 gives a comparison ofthe possible variations of a 2.048 Mbps signal.

Framing Format

No Framing 2.048 Mbps (32 time slots) Cannot use the publicly switched network.

No Multiframing 1.984 Mbps (31 time slots) No voice transmission with TS-16 signalling possible.

TS-16 Multiframing 1.920 Mbps (30 time slots) No error performance monitoring via CRCs.No CRC Multiframing

CRC Multiframing 1.984 Mbps (31 time slots) No voice transmission with TS-16 signalling possible.No TS-16 Multiframing

TS-16 Multiframing 1.920 Mbps (30 time slots) Voice transmission with TS-16 signalling and error monitoringand CRC Multiframe* possible.

Notes/LimitationsTotal BandwidthAvailable for Data/Voice

8

Causes of 2.048 Mbps Impairments

here are four main causes of2.048 Mbps impairments:

1. Faulty Equipment: Any piece of 2.048 Mbps equipment can cause

Terrors when the components fail oroperate outside of specifications. Errors,which can signal faulty equipment, includecode errors, bit errors, FAS (frame)errors, excessive jitter, and slips. Forinstance, code errors can occur due tofaulty clock recovery circuitry in spanrepeaters. These errors occur as theequipment becomes older and begins todrift out of specifications.

2. Improper Connections: Transmissionerrors are created by improper connec-tions or configurations. For example,

intermittent errors can occur whencomponent or cable connections areloose, and timing errors can occur whenimproper or conflicting timing sources areconnected together. Dribbling errors areoften caused by loose or unconnectedshield ground cables and by bridge taps.Further, upon installation, the circuit maynot work at all due to mislabelled pin-outson terminating cable blocks and to flip-flopped wires: transmit-to-transmit asopposed to transmit-to-receive. Theseerrors are typically discovered uponcircuit installation and possibly duringcircuit acceptance when tests areperformed end-to end.

3. Environmental: Electrical storms, powerlines, electrical noise, interference, andcrosstalk between transmission links cancause logic errors, FAS (frame) errors,CRC errors in addition to code errors.Typically, these conditions causeintermittent, bursty errors, which aresome of the most difficult to locate.

4. Data Specific: Data characteristics, suchas repetitive patterns, can force equipmentto create pattern-dependant jitter andcode errors. These errors may not existwhen testing the transmission path withstandard pseudorandom patterns.

Analyzing 2.048 Mbps Impairments

Techniques andMeasurements

T o analyse a 2.048 Mbps circuit’sperformance and to isolate thecauses of degraded services, thetest set must perform many

measurements in different scenarios.There are four typical scenarios where2.048 Mbps testing is required:

1. Installation: When installing a2.048 Mbps circuit, out-of-service testingis very useful in verifying equipmentoperations and end-to-end transmissionquality. One starts by testing the equipment(such as NTE’s, channel banks, multi-plexers), and then verifying cable connec-tions, timing source selections, andfrequency outputs.

Application #2 covers this testscenario.

2. Acceptance Testing: In addition to thetest performed during installation, twoother tests—stress tests and timed tests—should be performed to ensure that the2.048 Mbps circuit is operating properlywith respect to the relevant 2.048 Mbpscircuit specifications and tariff. Theequipment may be stressed by verifyingthe transmission frequency around2.048 Mbps equipment. The sameprocedure may be performed end-to-endto stress the entire 2.048 Mbps circuit.Timed tests with printouts should beperformed over a 24- or 48-hour periodusing standard pseudorandom patterns,which simulate live data.

Application #2 is useful for thisscenario.

3. Routine Preventive Measure: Routinemaintenance test are strongly recom-mended once live data is transmittedacross the 2.048 Mbps circuit. Routinemaintenance can alert technicians todegrading service before it disruptsnormal operations. In most instances,this involves monitoring the live data foralarms, code errors, FAS (frame) errors,

CRC errors, and signal frequencymeasurements which provide informationabout the performance of the 2.048 Mbpscircuit. These tests should be performedwith printouts over a 24- or 48-hourperiod to detect time specific orintermittent errors.

Application #1 covers this scenario.

4. Fault Isolation: Fault isolation isrequired once service is disrupted due toexcessive error rates. This can beperformed using both in-service and out-of-service tests. In-service monitoringprovides general information and can beused before out-of-service analysis tolocalise problems and minimise circuitdowntime. By monitoring the circuit atvarious points, technicians are able toanalyse the results and determine whereproblems are originating. By performingstandard out-of-service tests, such asloopback and end-to-end tests, techni-cians are able to stress the equipment,find sources of errors, and verify properoperation once the trouble is repaired.

Application #1 and Application #2are relevant for fault isolation.

9

Application #1:In-Service Analysis of Live Traffic

T he following sections explain howto evaluate the performance of a2.048 Mbps system usingcustomer data. It is useful:

• When performing periodic maintenanceand when looking for transmissiondegradation before it effects service.

• When analysing the span for intermittenterrors, which are caused by faultyequipment or environmental influences.

• For analysis of 2.048 Mbps circuits whichcannot be taken out-of-service.

Figure 6:

Possible 2.048 Mbps circuit monitoring locations

NTELocalExchange

NTEDigital

NetworkMUX MUX

LocalExchange

DCS DCS

Table 3:

Common alarm and error indications(in-service testing)

• Before out-of-service analysis, tolocalise the problem and minimisecircuit downtime.

To achieve all these benefits, theTTC test set may be configured to monitorthe 2.048 Mbps circuit from practicallyany 2.048 Mbps access point. Figure 6shows a typical circuit and possiblemonitoring locations.

Analysis of Alarmand Error Indications(In-Service Testing)

Testing and troubleshooting of a2.048 Mbps signal requires regularmonitoring for alarms and errors. Themonitoring for alarms and errors allows theuser to detect and sectionalize transmissionlines or equipment problems in a 2 Mbpssignal. Errors can also be intentionallyinjected to see the response of the system.

Table 3 highlights some of theimportant alarm and error indications alongwith possible reasons and solutions.

Result Possible SolutionReason

SIGNAL LOSS Indicates history of receiver Check cabling and connections.signal loss Check network equipment.

FRAME LOSS Indicates history of frame Check SIGNAL LOSS and POWER LOSS LEDs. If thesesynchronisation loss LEDs are not on, check FAS Distant alarm and AIS alarm.

FRAME SYNC Signal is unframed, or synchroni- Verify all settings and connections.sation to the specified framinghas not been achieved

FAS Distant Alarm Indicates remote (FAS Distant) Check span equipment downstream from present location.alarm Check local Tx.

AIS Alarm Indicates AIS alarm (Unframed Check span equipment upstream from present location.All Ones)

10

2 MbpsNetwork

EquipmentInput #1Output #1

Output #2Input #2

AIS Alarm Signal LossFrame Sync LossAIS Alarm

FAS Distant Alarm

Trouble lies here ifFAS distant (ALM 1)is received

2 MbpsNetwork

Trouble lies here ifAIS (ALM 2) is received

TTC Test Set

Figure 7:

2.048 Mbps network alarms

Figure 8:

Detection of 2.048 Mbpsnetwork alarms

Figure 9:

Code errors

1 1 0 1 1 0 0 0 0 0 1

Code Error

A. Code violation due to HDB 3 (no code errors counted)B. The bit error on the 2nd bit causes a code error

(the HDB 3 code is recognized and 1 code error is counted)

Bit Error Occurs HDB 3 Code

HDB 3 Code

The AIS and FASDistant Alarms

This section gives a detailedexplanation of the 2.048 Mbps AIS andFAS Distant alarms.

The AIS AlarmAn AIS alarm is an unframed

continuous stream of binary ones. However,a signal with all bits except the framealignment in the 1 state is not mistaken asan AIS.

If the network equipment shown inFigure 7 suffers a signal or framesynchronisation loss, or receives an AISalarm at input #1(2), it transmits the AISalarm at output #1(2). Hence, the AIS alarmindicates the presence of an alarm indicationto the equipment farther downstream (awayfrom the source of the trouble).

Therefore if the TTC test set receivesan AIS alarm, this indicates that the troublemust lie somewhere farther upstream in thenetwork. This is illustrated in Figure 8.

The FAS Distant AlarmThe FAS Distant alarm is indicated by

setting bit 3 equal to 1 in time slot 0 of theframes not containing the FAS pattern. (SeeFigure 3 on page 4).

If the network equipment shown inFigure 9 suffers a signal or framesynchronisation loss, or receives an AIS alarmat input #1(2), it transmits the FAS DistantAlarm at output #2(1). Hence, the FAS Distantalarm indicates the presence of an alarmcondition to the equipment farther upstream(back towards the source of the trouble).

Therefore if the test set receives a FASDistant alarm, this indicates that the troublemust lie somewhere farther downstream inthe network. This is illustrated in Figure 8.

11

Code Error AnalysisThe bipolar nature of the AMI

signal allows the detection of single(isolated) errors since single errors onthe line cause a pulse to be either incorrectlyadded or omitted, which in turn results intwo successive pulses of the same polarity.This constitutes a violation of the bipolarcoding scheme.

Recall that due to the zero suppres-sion scheme used in HDB3, the signal mayalso contain intentional bipolar codeviolations representing strings of 4 consec-utive zeros. These intentional code violationsdue to HDB3 must be distinguishing fromcode violations due to the errors occurringon the 2.048 Mbps line.

Since bipolar code violations due toHDB3 follow specific rules, they can berecognised as such by the TTC test set. Thisconstitutes the basis for the code erroranalysis performed by the test set.

A code error is defined as anyviolation of the bipolar code, which is nota code violation due to HDB3’s zero substitu-tion algorithm. For comparison, an illustra-tion of a code error along side an HDB3substitution code is shown in Figure 9.

It is not necessary to receive andtransmit a known pattern to recognise codeerrors. Hence the TTC test set can performcode error analysis on an in-service basiswithout disrupting the traffic on the 2 Mbpsline. To do this analysis, the TTC test setprovides the following key result:

Advantages/Limitationsof Code Error Analysis

Code errors provide an approximateindication of the error performance on ametallic 2.048 Mbps line without the needto disrupt live traffic. Furthermore, they cangenerally be used to sectionalise problemsto the local span in the 2.048 Mbps network.(This will be discussed further under “Corre-lation of Results and Problem Causes”).

It must be noted, however, that codeerror analysis has certain limitations. Codeerrors are useful in identifying local (nearend) metallic span and repeater problems.However they are not a good indication ofend-to-end performance since networkequipment beyond the local span or non-metallic transmission media (e.g. microwaveand fibre) will correct code errors in the farend 2.048 Mbps span.

FAS (Frame AlignmentSignal) Error Analysis

As we explained in our discussion ofthe 2.048 Mbps Framing Format, time slot 0of every other 2.048 Mbps frame contains afixed 7-bit long FAS pattern (See Figure 3on page 4). When doing in-service FAS erroranalysis, the TTC test set takes advantage ofthe fact that even though the data portion ofthe 2.048 Mbps frame is unknown, the FASbits contain a known pattern such that theerrors occurring on these bits can be detectedwithout disrupting the traffic.

Hence the TTC test set counts a FASerror each time one or more bits in the FASpattern are received in error.

Upon synchronisation with the framealignment signal, the TTC test set automati-cally provides the following result:

Advantages/Limitations ofFAS Error Analysis

FAS errors allow in-service errorperformance analysis of the 2.048 Mbpscircuit. Under random (Gaussian) errorconditions, the FAS error rate will closelyapproximate the actual error rate if thetest is performed over a significantly longperiod of time.

Moreover FAS errors can be usedto isolate problems to network equipment(such as digital cross connect systems andhigher order multiplexers) which frame(or reframe) the 2.048 Mbps data.

The limitations of FAS error analysisare threefold.

1. Since the FAS pattern takes up only 7 bitsfor every 512 bits transmitted (2 framesx 32 time slots/frame x 8 bits/time slot= 512 bits), the analysis is performed ona relatively small number of the receivedbits (about 1.4%). As a result, errors notoccurring on the FAS bits will be missed.

2. Bursty error condition are far morecommon than random (Gaussian)error condition.

3. FAS errors are corrected by multiplexersand digital cross-connected systems.Hence, FAS error analysis cannot be usedto determine end-to-end error perfor-mance in networks where this type ofequipment is installed.

FAS Errors Number of FAS(FAS ERR) errors received

since beginningof test.

Code Errors Number of code(CODE ERR) errors detected

since beginningof test.

Code Error Rate Ratio of number of(CER) code errors in last

test interval tonumber of bitsexamined in lasttest interval.

12

CRC Error AnalysisWhen the 2.048 Mbps signal has the

CRC Multiframe format implemented, the TTCtest equipment will automatically perform CRCerror analysis as explained in Table 4.

Table 4:

CRC error analysis procedure

Note:To derive the approximate bit error rate (BER) from the average CRC error rate use the following formula:

Approximate BER = AVG CRC/(# of bits in SMF - CRC bits in SMF)= AVG CRC/(2048 - 4) = AVG CRC/2044

This formula will give a fairly accurate approximation to the actual BER, as long as there is no more than one bit eror per submultiframe(i.e., average BER < 1E-6).

Advantages/Limitations ofCRC Error Analysis

Most data sequences generate a CRCword which can be uniquely associated withthat particular data sequence. Therefore,CRC errors can detect the presence of oneor more bit errors in a submultiframe to avery high degree of accuracy (93.75%)without the need to take the 2.048 Mbpscircuit out-of-service.

However, the following limitation ofCRC error analysis must be kept in mind.

1. A CRC error indicates the occurrence ofone or more errors, but not the totalnumber of errors in a submultiframe.Hence, the BER obtained using theformula above will be somewhat lowerthen the actual error rate if the error rateis so high that there are several errors inthe submultiframe.

1 error per submultiframecorresponds to an averageerror rate of 4.9E-4.

2. CRCs may be recalculated by networkequipment such as digital cross connectsystems. Therefore, CRC error analysiscannot be used to determine end-to-endperformance in networks where this typeof equipment in installed.

Step 1 The 4-bit CRC is calculated for a 2.048 Mbps SMF (submultiframe = 8 frames).

Step 2 The CRC is inserted in the CRC bits of the next SMF.

Step 1 The TTC test set recalculates the CRC for the SMF.

Step 2 The TTC test set compares the calculated CRC to the CRC it receives in the CRC bits of the next SMF.

Step 3 The TTC test set declares a CRC error if the received CRC and the calculated CRC do not match indicatingthe occurence of one or more bit errors in the SMF.

CRC Errors (CRC ERR) Number of CRC errors counted since beginning of test.

CRC Error Rate (AVG CRC) Ratio of number of CRC errors counted to number of CRCs received.

At the Transmitter:

At the Receiver:

For CRC error analysis, the TTC test set provides the following results:

13

Correlation ofIn-Service Results

To find possible problem causes, useFigure 10 to find your location along the2.048 Mbps span, and cross-reference yourlocation with Table 5, which shows variouscombinations of the results discussed in theprevious sections.

NTELocalExchange

NTE2 MbpsNetwork

LocalExchange

Cross-ConnectLocation

Cross-ConnectLocation

A

C

A

C

B

Table 5:

Correlation of results and problem causes

Figure 10:

Possible problem locations

Location forFigure 10

A Code Errors Local problem. Possibly bad cabling connections between test set and circuit,corroded “dirty” cable plugs, or defective NTE.

A, B, or C Received Frequency Offset Frequencies which are out of range may affect jitter tolerance and noisemargins, in addition to causing error bursts and slips.

B or C Code Errors, FAS Errors, or Local 2.048 Mbps span problem. Possible faulty repeater, span line noise,CRC Errors crosstalk, poor cabling, or defective monitor jacks.

C Code Errors, No FAS Errors, Local 2.048 Mbps span problem.or CRC Errors

C No Code Errors, FAS Errors, Typically far-end span line problem. Sectionalise further. Potential for lightor CRC Errors guide, radio, or Violation Monitor Removal (VMR) equipment in network.

Problem/LocationResults

14

Application #2:Out-of-Service Testing of 2.048 Mbps Circuits

T he following sections explain howthe TTC test set is used to evaluatethe performance of a 2.048 Mbpssystem using pseudorandom

data. It is useful:

• When installing 2.048 Mbps circuits andverifying end-to-end continuity.

• When isolating 2.048 Mbps circuit faultsand verifying end-to-end continuity.

• When performing acceptance testingwhich includes timed and stress tests.

Errors found via this analysis maybe caused by faulty equipment, improperconnections, environmental influences, ordata content. To find these errors, use resultssuch as bit errors, bit error rate (BER), FASerrors, pattern slips, received frequency, errorfree seconds (EFS), percentage error freeseconds (%EFS), etc, which are all measuredsimultaneously. These results will help inisolating the cause of the problem.

There are basically two methods ofperforming out-of-service testing: loopbacktesting and end-to-end testing. These methodsare addressed in the following sections.

Analysis of Alarmand Error Indications(Out-of-ServiceTesting)

Testing and troubleshooting of a2.048 Mbps signal requires regularmonitoring for alarms and errors. Themonitoring for alarms and errors allows theuser to detect and sectionalize transmissionlines or equipment problems in a 2 Mpbssignal. Errors can also be intentionallyinjected to see the response of the system.

Table 6 highlights some of theimportant alarm and error indications alongwith possible reasons and solutions.

Table 6:

Common alarm and error indications

Result

PATTERN SYNC Test set is not synchronised to the Check BERT pattern selection and FRM SYNC status. If test set inincoming pseudorandom pattern self loop is operating properly, this indicates 2.048 Mbps circuit

problem.

FRAME SYNC Signal is unframed, or synchronisation Verify all settings and connections.to specified framing has not beenachieved

FAS Distant Indicates remote (FAS Distant) alarm Check span equipment downstream from present location.

AIS Alarm Indicates AIS alarm Check span equipment upstream from present location.

SolutionReason

15

NTELocalExchange

NTE2 MbpsNetwork

LocalExchange

DCS DCS

TTC Test Set TTC Test Set

NTELocalExchange

NTE2 MbpsNetwork

LocalExchange

DCS DCS

TTC Test Set

Figure 11:

Basic setup—end-to-end testing

Figure 12:

Basic setup—loopback testing

End-to-End TestingEnd-to-end testing is performed

with two TTC test sets so that both directionsof the 2.048 Mbps circuit may be analysedsimultaneously. Figure 11 shows the set-upof an end-to-end test. This test method isbetter then the loopback test since the direc-tion of errors can be found more quickly.

Loopback TestingLoopback testing is performed with

one TTC test set. Figure 12 shows the set-upof the loopback test. If NTE loopbacks are

established to perform the test, it is importantto realise that the far end NTE in loopbackwill affect the result. By design, most NTE’sremove received code errors beforetransmitting the data. This will affect theanalysis result, because the near endtechnician will be unaware of code errorsoccurring on the far end metallic loop andmay draw inconclusive results. Furthermoreloopback tests cannot identify incorrecttiming configurations where the customerpremises equipment (connected to the NTE)may not be loop-timed to the network.

The appropriate pseudorandompattern recommended forout-of-service testing at2.048 Mbps is the 215 – 1pattern as specified by ITU-TRecommendation O.151.

16

Analysis of SlipsSlips and Their Causes

A pattern slip is the insertion of databits into or from the data stream. Based onthe source of the slip and its effect on thenetwork, all slips can be placed on any of thefollowing categories.

1. Controlled Slips: Controlled Slips are bitadditions or deletions which do notdisrupt frame synchronisation.

These slips are typically caused bysynchronisation impairments in digitalcross-connect (DCS) equipment. DCSequipment handles buffer overflows orunderflows by deleting or repeating entireframes of data. Since data is added ordeleted by entire frames, frame synchron-isation is not disrupted.

2. Uncontrolled Slips: Uncontrolled slips arebit additions or deletions that cause bothdata and framing bits to be displaced. Themisalignment of framing bits typicallyresults in frame synchronisation loss.

Uncontrolled slips are typically fromsynchronisation problems in equipmentwhich buffer the entire bit stream such as

satellite down link receivers. Since thebuffer in this equipment does not distinguishbetween framing and data bits, bufferunderflows or overflows result in theaddition and deletion of arbitrary blocksof data.

It should be noted that slips can alsoresult from impairments unrelated to networksynchronisation. Low signal level, noise, andexcessive jitter can also cause slips.

Examples of controlled and uncon-trolled slips are illustrated in Figure 13.

Measuring the SlipsThe TTC test sets pattern slip measure-

ments count the number of times data isinserted into or deleted from the pattern.

This measurement is not acount of the actual number ofbits added or deleted, but rathera count of the number ofinstances where a group of bitswere added or deleted from thebit stream.

Interpreting the ResultsTo troubleshoot a problem, which

causes slips, pattern slip results must becompared to other test results.

If an occurrence of a pattern slip isassociated with a frame loss, it can beassumed that the frame loss is caused by anuncontrolled slip. If a pattern slip occurswithout disrupting framing, it can be assumedthat a controlled slip has occurred. Categori-sation of slips can help identify the cause ofthe problem.

A better understanding of theunderlying problems can also be obtained byconsidering the frequency at which patternslips occur.

17

Figure 13:

Controlled anduncontrolled slips

4 B

its D

elet

ed

248

247

246

245

244

243

242

241

240

239

238

237

236

235

234

FAS

FAS

2928

2726

2524

2322

2120

1918

1716

1514

1312

1110

98

76

54

32

1

B U F F E R

248

247

246

245

244

243

242

241

240

239

238

FAS

FAS

3433

3231

3029

2827

2625

2423

2221

2019

1817

1615

148

76

54

32

113

2530

2015

105

024

524

924

023

5B

uffe

rO

verf

low

Occ

urs

Bit

Posi

tion

Rel

ativ

e to

FAS

2530

2015

105

024

524

924

023

5

Bit

Posi

tion

Rel

ativ

e to

FAS

FAS

Expe

cted

249

Bits

Aft

erFA

S, B

ut O

ccur

s4

Bits

Ear

ly

Unc

ontr

olle

d Sl

ip

FAS

FAS

FAS

FAS

FAS

Fram

e 5

Fram

e 4

Fram

e 3

Fram

e 2

FAS

Fram

e 1

Buf

fer

DC

S

FAS

FAS

FAS

FAS

FAS

Fram

e 6

Fram

e 5

Fram

e 4

Fram

e 2

FAS

Fram

e 1

Buf

fer

Ove

rflo

wed

and

Fra

me

3 W

as D

elet

ed

Con

trol

led

Slip

18

TransmissionDelay Analysis

Using the TTC test set’s DELAY canhelp in troubleshooting specific problemssuch as protocol errors due to timeouts.

NTE

NTE DCS

DCS

TTC Test Set

DCS

DCS

FEP

FEP

London, UK

Sheffield, UK

Path #1Round Trip Delay: 30 ms

Path #2Round Trip Delay: 75 ms

Figure 14:

Roundtrip delay measurements

As an example consider the2.048 Mbps circuit shown in Figure 14.In this figure, transmission path #1 hasa roundtrip delay of 30 ms, whereastransmission path #2 has a round tripdelay of 75 ms. If we assume a protocoltimeout threshold of 50 ms, switching the

2.048 Mbps circuit from transmissionpath #1 to transmission path #2 wouldcause protocol timeouts not experiencedwhen path #1 was in use. TTC test set’s DELAYmeasurements can identify this problem bydetermining such changes in the transmissionpath of a 2.048 Mbps circuit.

ITU-TPerformance Analysis

Performance Analysis results asspecified by ITU-T Recommendations G.821provide statistical information about theperformance of the equipment or systemunder test. These results are used to checkthe compliance of equipment or circuits withthe specified performance objectives.

Available Time vs.Unavailable Time

According to ITU-T RecommendationG.821, the total test time after the initialpattern synchronisation is broken up intoavailable and unavailable seconds. Every testsecond belongs to either one of thesecategories. This is illustrated in Figure 15.

After initial synchronisation isachieved, seconds are considered to be

available time. When the bit error rate (BER)is worse then 10-3 for 10 consecutive seconds,a transition is made to unavailable time, andthese 10 seconds are considered to beunavailable time. When the BER is better then10-3 for 10 consecutive seconds, the period ofunavailable time terminates, and these 10seconds are counted as available seconds.Hence, a sliding window, 10 seconds inlength, is used to detect transitions fromavailable time to unavailable time and viseversa.

Any second in which a signalloss or pattern synchronisationloss occurs, is also consideredto be a second with BER worsethen 10-3.

AvailableSeconds

Total Test Time fromInitial Pattern Synchronization

UnavailableSeconds

Figure 15:

G.821 available timeand unavailable time

19

Figure 16:

Available time

Degraded MinutesDegraded minutes is a count of the

number of minutes during which an averageBER of 10-6 or worse occurs. The one-minuteintervals are derived by removing unavailableseconds and severely errored seconds fromthe total test time, and then consecutivelygrouping the remaining seconds into blocksof 60. The average BER is calculated for theblock of 60 seconds, and if it is 10-6 or worse,the block is counted as a degraded minute.

Copyright 1991, 1999, TTC, a division of Dynatech, LLC.All rights reserved. TTC is a registered trademark of TTC.All other trademarks and registered trademarks are theproperty of their respective owners. Specifications, terms,and conditions are subject to change without notice.

Available TimeAs shown in Figure 16, available time

(or available seconds) is broken up intofurther categories. These categories areexplained below.

Severely errored seconds aredefined to be part of availabletime. Therefore, severely erroredseconds are likely to account forshort error bursts with a BERworse then 10-3, whereas longererror bursts with a BER worsethen 10-3 are likely to be countedas part of unavailable time.

Error Free Seconds Available seconds(EFS) in which no bit

errors occurred.

Errored Seconds Available seconds(ERR SEC) in which at least

one bit erroroccurred.

Severely Errored Available secondsSeconds (SES) in which the BER

was worse than 10-3.

Error FreeSeconds

(EFS)

ErroredSeconds

(ERR SEC)

SeverelyErroredSeconds

(SES)

20

World Headquarters20400 Observation Drive

Germantown, Maryland 20876-4023 USAUSA 1-800-638-2049 • +1-301-353-1550 • FAX +1-301-353-0234

Canada 1-888-689-2165 • +1-905-812-7471 • FAX +1-905-812-3892www.ttc.com

North American OfficesUnited States

Atlanta, Georgia • Chicago, Illinois • Dallas, Texas • Denver, ColoradoEast Rutherford, New Jersey • Los Angeles, California

Roanoke, Virginia • San Jose, California

CanadaCalgary, Alberta • Laval, Quebec • Toronto, Ontario

Vancouver, British Columbia

International OfficesAustralia

Melbourne +61-3-9563-4800Sydney +61-2-9926-1447

Benelux +32-15-28-7686China

Beijing +86-10-6460-5258Hong Kong +852-2892-0990Shanghai +86-21-6445-8938

France +33-1-39-30-24-24Germany +49-6172-5911-00United Kingdom +44-1189-759696European Freephone +800-TTC-UKTAC

(+800-882-85822)

International DistributorsArgentina • Brazil • Chile • Colombia • Czech Republic

Denmark • El Salvador • Finland • India • Indonesia • IrelandIsrael • Italy • Japan • Korea • Malaysia • Mexico • Norway

Peru • Philippines • Saudi Arabia • Singapore • SlovakiaSouth Africa • Spain • Sweden • Switzerland • Taiwan

Thailand • United Arab Emirates • Venezuela

ITU-T BASICS-A-11/99