Maintaining and Troubleshootingcdma2000 Base Stations

Page 2

Module Objectives

At the end of this presentation you will be able to:• Understand why maintenance testing is important• Be familiar with the new variable length Walsh Code Display• Be familiar with key CDMA transmitter measurements• Be able to relate the measurements to solving network problems• Understand the differences in IS-95 and IS-2000 Base Station

measurements

Page 3

Why Do Maintenance Test?

• System performance is a big competitive issue• Periodic maintenance helps prevent shutdown• Equipment problems may only show up as reduced capacity• Monitoring for interference finds problems unrelated to network

equipment• Systems will be stressed as loading increases• Defective components may be hidden by CDMA's soft handoff, power

control, and error correction

Page 4

What We Test - The CDMA Network

Public SwitchedTelephoneNetwork(PSTN)

Mobile TelephoneSwitching Office

(MTSO)

BTS

BTSBTS

Backhaul (T1)

Base StationTransceiverSubsystem

(BTS)

Page 5

CDMA Site Block Diagram

SectorGamma

TxRxRx

SectorBeta

TxRxRx

CDMAController

ChannelElement

Morechannelelements

�D/A & A/DConversion

Modulate& Amplify

DigitalInterface

GPS Rx

System TimeReference

& Distribution

TXRXA

RXB

To BSCor MTSO

T1

Sector Alpha

Page 6

Base Station Parametric Measurements

• Total Power (Average Power & Channel Power) • Waveform Quality (Rho) • Carrier Feedthrough • Pilot Time Tolerance (Time Offset)• Frequency Tolerance (Frequency Error)• Code Domain Power • Antenna and site Measurements

Page 7

Base Station Measurement Choices

Connected Measurements Over-the-Air Measurements

Page 8

Transmitter Test Setup

Even Second Pulse*

Base Station TimingBTS RF SignalFrom BTS Antenna Port

* Only Required for Time Offset

Page 9

Why is Forward Link Power Management Important?

Power management is critical to maximizing the system’s capacity

• Network operators sometimes attempt to set power higher to extend coverage to reduce infrastructure cost; result can be pilot pollution problems

• Initial settings for the sites must be accurate to match settings specified by the RF engineering department

• More power is not necessarily better but can lead to interference and dropped calls

• Too little power for the site may result in dead spots between sites

Page 10

Why is Accurate Power Measurements Important

• An Inaccuracy of ± 1dB can• Equate to affecting power

output by 21% • Affect channel capacity• Interfere with cell site

planning• CDMA system is limited by

interference from other transmitters

Power Set too Low –Dead area

BTS

BTS

BTS

Power Set too High –Added interference

Page 11

CDMA Average Power Measurements

Input Detector A / DConversion

DSP

• Performing cdma power measurements requires a new measurement technique

• High/variable crest factor of cdma forward link makes peak power meters read high

• Base Station Test Set Power Meter works independent of crest factor

• Broadband measurement• Offers the best accuracy (7.5%)• Measurement is triggerable

Page 12

CDMA Channel Power Measurements

• Band-limited measurement (1.23 MHz bandwidth)• Useful when interference sources exist (including other cdma

channels)• Wider dynamic range than average power

• For low-level signals (down to -70 dBm)• Measurement is triggerable

Input A / DConversion

Band PassFilter (DSP) DSP

Page 13

Non-Linearity in the Frequency Domain

• Non-linearity causes intermodulation

• ‘Shoulders’ on the Waveform• Power in adjacent channels

• Causes of intermodulation• Overdrive Power Amplifiers• Bad Mixers

• CDMA signals have high crest factor

• In excess of 12 dB• EXAMPLE: 10 Watts average

power transmitter needs to have an amplifier with enough headroom to produce 158 Watts peak

Energy in Adjacent Channels

Shoulders

Page 14

Base Station Parametric Measurements

• Total Power (Average Power & Channel Power) • Waveform Quality (Rho) • Carrier Feedthrough • Pilot Time Tolerance (Time Offset)• Frequency Tolerance (Frequency Error)• Code Domain Power • Antenna and site Measurements

Page 15

Waveform Quality (Rho)

� Power that correlates with idealTotal Power

=

Signal PowerSignal Power + Error Power� =

� > 0.912

Page 16

Why is Rho Important?

• Key measure of modulation quality• Analogous to FM accuracy/distortion (AMPS) and EVM (TDMA

systems)• Rho performance affects site/sector coverage area and capacity in

the site/sector• Rho failures can indicate problems in:

• Compression in linear amplifiers• Magnitude and phase errors in the IQ modulator• Phase non-linearity (group delay)Spurious signals in the

transmission path• Carrier feedthrough

Page 17

Base Station Parametric Measurements

• Total Power (Average Power & Channel Power) • Waveform Quality (Rho) • Carrier Feedthrough • Pilot Time Tolerance (Time Offset)• Frequency Tolerance (Frequency Error)• Code Domain Power • Antenna and site Measurements

Page 18

Carrier Feedthrough

• Carrier feedthrough (origin offset)• Should be < -25 dBc• Carrier feedthrough in I/Q Domain and Frequency Domain

Page 19

Pilot Time Tolerance(Time Offset)• Measure of Short Code sequence timing vs. System time

• Checks the “start” of the PN offset as compared tothe even-second clock signal

110 … 011 ... 011 … 100 111 … 010 000 … 001 110 … 011101000..000 101000..000

110 … 011 ... 011 … 100 111 … 010 000 … 001 110 … 011101000..000

110 … 011 ... 011 … 100 111 … 010 000 … 001 110 … 011101000..000 101000..000

101000..000

PN Offset 0

PN Offset 1

PN Offset 2

32768 Chips

32768 Chips

15 Zero’s

64 chips 64 chips 64 chips

Even Second

32768 Chips

Page 20

What is Pilot Time Tolerance?

• Time offsets outside of specifications can affect handoffs between cells - the island effect

• Time offset is one of the parameters that will lead to errors inposition location with the introduction of E911 and network operator services

• Potential causes for failures of pilot time tolerance:• GPS receiver and timing distribution failures• Cells with a propagation delay greater than the PN Offset time period• The timing delay adjustment (used to compensate for time delays

through the sites cabling) may be off

Page 21

Frequency ErrorWhy is it important?• GPS drift or out-of-lock condition can create the island cell effect • Frequency drift can lead to site timing errors which will lead to

errors in position location with the introduction of E911 and network operator services

• Failures point to problems in GPS receiver and timing distribution

Note: This measurement cannot be made with a frequency counter frequency tolerance (Frequency Error) specifications: ±0.05 ppm PCS (99 Hz @ 1980 MHz)

Page 22

Base Station Parametric Measurements

• Total Power (Average Power & Channel Power) • Waveform Quality (Rho) • Carrier Feedthrough • Pilot Time Tolerance (Time Offset)• Frequency Tolerance (Frequency Error)• Code Domain Power • Antenna and site Measurements

Page 23

Code Domain Power is Important

• Analogous to spectrum analyzer on FDMA systems• Verifies network system settings

• Allocation of assigned power to pilot, sync, paging and traffic channels

• Specified: Code Domain noise floor must be at least <27 dB below total power in inactive channels, else capacity suffers

• Code Domain Power failures can indicate problems in• Linear Amplifiers• Channel elements / cards• Settings in network control software

Page 24

The CDMA Concept10 Khz BW 1.23 Mhz BW 1.23 Mhz BW 10 Khz BW

fcfc

CDMATransmitter

CDMAReceiver

BasebandData

0 0

fc fcfcExternal Interference

Interference Sources

fc

Other User Noise

Encoding &Interleaving

Walsh CodeSpreading

Walsh CodeCorrelator

BasebandData

Decode & De-Interleaving

1.23 Mhz BW1.23 Mhz BWSpurious Signals-113 dBm/1.23 Mhz

9.6 kbps 19.2 kbps 1228.8 kbps 9.6 kbps19.2 kbps1228.8 kbps

Background Noise Other Cell Interference

Page 25

Code Domain Power

1.2288 MHz

User1

User3

User2

Pilot Paging Synch0 1 2 3 4 5 6 7 8 9 32 40 63

PilotSynch Frequency DomainPaging

User #3User #2User #1

Code Domain

Walsh Code

Page 26

IS-2000 Code Domain Power ScreenSupplemental channels are clearly visible as wide blocks.

Supplemental Channel

Single SR1 (1xRTT) channel

Page 27

Cdma2000 Code Domain Power Screen Bit Reverse Ordering

Pilot is the same for IS-95 and 1XRTT channel

Marker Primary + Supplemental Channels

Primary channel Walsh 20

Current Display Bit Reverse order

Channel is 4 Codeswide

Page 28

Complex Power Code Domain power display

Q Code

I Code shows combined IS-95 and 1xRTT

Q Codes show 1XRTT only

I Code

Any Channel that is active in I Code and not active in Q Code means it is IS-95 only.

Page 29

Complex Power Code Domain power display

I Code shows combined IS-95 and 1XRTT

Q Codes show 1XRTT only

Any Channel that is active in I Code and not active in Q Code means it is IS-95 only.

I Code

Q Code

Page 30

IS-2000 SR1, RC3 (9.6 kbps)

Long CodeDecimator

Interleaver

38.4 ksps

1/4 Rate Conv.Encoder

38.4 ksps

9.6 kbps

38.4 kbps

Walsh 64Generator

1228.8 kcps

1228.8 kcps

1228.8 kbps

1228.8kbps

Q

I

S -PPC

Q

I

PCDec

1228.8 kcps

Q Short Code

I

Q

1228.8 kcps

ComplexScrambling

FIR

FIRI Short Code

OrthogonalSpreading

1228.8 kcps

1228.8 kcps

+

+

+

-

38.4 ksps

19.2 ksps

19.2 ksps

P.C. Bits

Gain

Gain

PunctureTiming

Add CRC andTail Bits

Short Code Scrambler

Walsh CodeGenerator

1228.8 kcps

Q Short Code

Q

I

FIR

FIR

I Short CodeIS-95B

Full RateData Bits Power

ControlPuncture

8.6 kbps

I

800 bps

OptionalCan be Carried by F-DCCH

1228.8 kbps

Long CodeGenerator

User LongCode Mask

800 bps

Q

Decimate by Walsh Length/2

Page 31

Non-Linearity in Walsh Code ChannelsMixing Products

• Non-Linearity can cause Walsh Code mixing

• Upper display shows Code Domain Power Display for a cdma signal in a linear system

• Lower Plot shows the same signal through an amplifier driven into compression

• Non-linearity causes power from one Walsh code to bleed into others.

• Walsh 1 mixed with Walsh channel 32, creating power in Walsh 33

• Walsh channel 17 mixes with Walsh 32 creating power in Walsh 49

MixingProducts

Page 32

CW Interference in the Code Domain

• PN spreading distributes CW power over all Walsh codes

• CW tones look like white noise in the Walsh code domain

• Example: CW spur with 200 kHz offset and the same signal level as the cdma signal

• Noise floor is –21 dBc. Standard calls for minimum of –27 dBc

• Spur shows up in Spectrum Analyzer trace in a single location while noise is evenly spread in Code Domain

Page 33

Noise in the Code Domain

• All of the power in White Gaussian Noise (WGN) that falls inside the 1.23 MHz BW becomes interference

• Contributes to the code domain noise floor

• WGN is Walsh code white or equally distributed over all 128 Walsh codes

• Example: AWGN with the same power spectral density as a cdma signal

• Equivalent to tone example• Code Domain floor at –21 dB

Page 34

AWGN in the Code Domain

• All sources of uncorrelated power behave similarly

• Example:• Signal power = -10 dBm/1.23 MHz• Noise power = - 13 dBm/1.23 MHz

• White noise over the 1.23 MHz BW shows up in the frequency domain

• White noise again raises the noise floor in the Code Domain

Page 35

Base Station Over-Air Measurements

Page 36

Pilot, Sync, and PagingCode Domain Display

Page 37

Quick Paging Channel in cdma2000

Page 38

Pilot Dominance Required for Over-Air Tests

Dominant Pilot

Page 39

Over-Air Base Station Measurement Statistics

Page 40

Base Station Troubleshooting

Page 41

Base Station Parametric Measurements

• Total Power (Average Power & Channel Power) • Waveform Quality (Rho) • Carrier Feedthrough • Pilot Time Tolerance (Time Offset)• Frequency Tolerance (Frequency Error)• Code Domain Power • Antenna and site Measurements

• Antenna Return Loss (VSWR)• Cable Fault Location• Insertion Loss• GPS Timing Distribution Measurements

Page 42

Antenna Return Loss

•Checks the health of the feedline and antenna network

•Excessive return loss results in decrease in transmitted power

•Antennas and cables are sometimes damaged due to vandalism and environment

•Test periodically and compare to benchmarks taken at installation time - can store the sweep results

Page 43

Antenna Return LossTest Set-up

Source Signal

Reflected Signal

Directional Coupler(Bridge)

Antenna, Feedline and Connectors

RF Source

SpectrumAnalyzer

Agilent Test Set

Page 44

Distance to Cable Fault Measurement Test Set-up

Agilent Test Set Fault Connector

RF Source

SpectrumAnalyzer

Cable under test

PowerDivider

Antenna orTermination

Results FaultIndication

Fau

lt M

a gn i

t ud e

Distance

Page 45

Cable Fault Measurements

• Cable Fault tests are a good complement to the return loss test (test SWR; if it fails, use cable fault to isolate the problem)

• Allows the technician to better isolate the fault• Can be used to determine if a “tower crew” is needed to fix the

problem• Accuracy: 4 feet for a 500’ cable• Maximum cable length: 500 feet

Page 46

Testing the GPS receiver and distribution

Spectrum Analyzer verifies presence and level of the 16X chip clock

• External GPS RX verifies frequency accuracy of 16X chip clock

Page 47

Summary

• Why we test• Why Power is a Critical Base Station Parameter• What Tests are Performed on a Base Station• What the Test Results mean