WCDMA Radio Optimization
WCDMA Radio Optimization
WCDMA Radio Optimization
Contents
Radio planning optimization ................................................ 3-14 Quality of Service .............................................................. 15-22 Measurement and statistics collection ............................... 23-29 KPI .................................................................................... 30-32 Accessibility....................................................................... 33-38 Retainability....................................................................... 39-44 Integrity ............................................................................. 45-47 HSDPA-HSUPA ................................................................ 48-55
-2-
-3-
Antenna height
Since WCDMA performance is interference limited the cell dominance areas should be kept as controlled as possible lf the antenna is located too high (no proper tilting) then The cell gathers more traffic and external interference and thus the effective capacity is decreased Produced interference decreases the capacity of the surrounding network Also surrounding networks service probability is negatively effected
Antenna azimuth
Natural obstacles and buildings should be used to create good dominance areas for WCDMA cells This improves the SHO performance and decrease interference Example of a UMTS cell, that is naturally bordered (wall effect) by buildings
When re-using the GSM sites, analysis should be made whether the UMTS antennas should be positioned lower This analysis is done with simulations and visiting the site locations in practise
Part of network reused few +40meter GSM antenna heights High UMTS antenna positions lowered to 25-35m
Dominance areas become clear, so less interference is introduced and HO performance is better. Capacity is increased and performance enhanced!
-4-
Antenna tilt
ln addition to antenna height, downtilting is very important physical means for interference minimizing in WCDMA Basic rule of designing antenna tilt is that the height of the antenna should be selected with respect to the wanted amount of cell range If the cell range with respect to available antennas and their tilting with a feasible amount of tx-power becomes too large to suit the network plan, then the antenna must be lowered According to experience, the analysis should start with the optimum tilting and not by reducing the tx-powers of the cell, which can be optimized after the tiltings are done Horizontal plane h
Antenna tilt
According to experience even 15 degrees of downtilting is not impossible (lf the radiation pattern of the antenna supports it), although in practice not very often needed.
There has also been lot of discussion of a potential need to change the tilts often during the network lifecycle (even regularly) However practice have not shown such need if the tilts are design well from the start with help from simulations But once WCDMA gets congested this might be given another look (Remote tilts).
-5-
Sectorisation
According to simulations and analysis, sectorisation of WCDMA site helps to improve capacity of the network However, as permissions for additional antennas are quite hard to come by, e.g. 6-sector sites might be very rare
Antenna 3 Other to own dB cell Beam interference width ratio, i Served users Soft handover overhead UL coverage probability (outdoor to indoor) For 8/64/144 kbps 70/32/40% 85/50/59% 87/55/62% 86/59/62% 90/62/68% 92/67/72% 92/70/71% 88/65/64% 95/75/79% 96/80/82% 96/80/81% 93/76/76%
Sectorisation can increase the capacity if correct beamwidth antennas are selected and SHO properly controlled
OMNI CASE
Omni 1200 900 650 1200 900 650 330 120 900 650 330
0
0.79 1.33 1.19 0.88 1.72 1.49 1.09 0.92 2.18 1.97 1.43 1.15
240 441 461 575 489 510 604 691 593 627 758 880
28% 39% 35% 34% 54% 51% 41% 40% 64% 59% 55% 48%
FOUR SECTTORED CASE, 65O antenna SIX SECTORED CASE, 33O antenna
-6-
Increases uplink coverage/capacity in low loaded network Compensates for feeder and combiner losses in the uplink direction, increasing coverage for suburban, rural and road sites where antennas are in very high positions and the feeder lines are long Allows UEs to reduce transmission power level With heavily loaded network (i.e. high interference) the benefit of the mast head amplifier is negligible Also in downlink limited 3G networks (DL oriented traffic, users in cell edge, DL tx-power per user low e.g. in for high bit rate indoor users) the usage of mast head amplifier is not justified Needs extra space in the masts and increase the wind load
-7-
Transmission powers
Default transmission powers are determined by the equipment vendors. In initial phase of the planning Transmission powers of TCHs and CCHs needs to be set Maximum UE transmission power is to be defined In DL the power tuning between TCHs and CCHs has effect on network performance More power to CCHs > better channel estimation, which improves the Eb/No performance and thus improves coverage More power to TCHs > better capacity Rule of thumb: 15-20% of DL total power is used for CCHs Maximum UE transmission power should be set to 21-24 dBm (network operation and battery life) Most important control channel is the common pilot channel (CPICH)
Transmission powers
Also other control channels beside CPICH need power (for example BCH) to enable correct functioning of the system All the other common control channels are powered in relation to the P-CPICH The goal of allocating power to the common channels is to find a minimum power level needed for each channel to secure the network operation and to provide the same cell coverage area as with CPICH, but not to waste any capacity left for the traffic channels.
-8-
PCH: Paging channel initiates the communication from network side SCH: Synchronization channel FACH: Forward access channel carries control information to terminals that are known to be located in the given cell. ls used to answer to the UL RACH message. BCH: Broadcast channel carries network specific information to the given cell (random access slots for UL, antenna configuration etc) PICH: Paging indicator channel is used to provide sleep mode operation for UE AICH: Acquisition indicator channel is used to indicate the reception of RACH CCPCH: Primary and secondary common control physical channels (P-CCPCH and S-CCPCH) are physical channels that carry BCH, FACH and PCH.
Transmission powers
P-CCPCH transmitted with activity factor 0,9 S-CCPCH transmitted with activity factor 0,25 SCHs transmitted with activity factor 0,1 AICH, PICH and CPICH are transmitted continuously
The BCH is transmitted on the P-CCPCH and FACH and PCH on the S-CCPCH The BCH is transmitted on the P-CCPCH continuously expect during the 256 first chips, when the P-SCH and S-SCh are transmitted we can assume 0,1 activity factor for the SCHs and 0,9 for the P-CCPCH
Channel P-SCH S-SCH PICH AICH P-CCPCH S-CCPCH CPICH All common ch. Allocated power 0,331W 0,224W 0,1W 0,126W 0,245W 1,165W 1W 3,191W 31% 100% 5% 16% Power out of the total common channel powers Power out of the maximum Node B transmission power (20W)
-9-
Carrier addition
Adding a carrier to less transmit power per carrier, if no additional PA is installed Additional carrier can also be used for e.g. optimisation of indoor coverage with clever network planning and parametrisation (not with power reduction) Even with less transmit power, there is a capacity gain possible especially for high traffic areas (low cell range) Actual gain produced is heavily dependent on the traffic mix Carrier configuration 1C>2C 2C>3C Dense Urban 350m 92% 41% DL Capacity gain Urban Suburban 550m 1700m 87% 37% 77% 27% Rural 7km 60% 15%
Pilot pollution
Pilot pollution is faced on a certain area when there is no clearly dominant CPICHs over the others. The pilot pollution creates an abnormally high level of interference, which is likely to result in the performance problems Increased interference level Poor service quality, decreased throughput or increased delay Decreased service access Frequent changes in Active Set and potential risk for unnecessary handovers. Higher non-controllable load
Pilot pollution
The yellow dots represent points where 4-5 CPICHs were received within 6dB window As Active Set size is typically 3, in this situation the rest of the Pilots produce unnecessary interference
Pilot pollution
Pilot pollution can be (at least partly) avoided by planning the CPICH powers and SHO parameters so that throughout the network there is only 2-3 CPICHs available for the UEs, strong enough to be included in the Active Set. All CPICH outside Active Set should be clearly weaker Antenna design, height and tilt are selected carefully Balanced UL & DL SCH/DCH power adjustments
- 11 -
Incilude the missing cell to neighbour list if its wanted to active set or change cell plan if FIO
SHO optimisation
Soft/Softer HO planning and correct operation is one of the most important means of optimizing WCDMA networks The importance is high because of the high biterate (pathloss sensitive) and RT (delay sensitive) RABs SHO is measured in terms of probability, the percentage of all connections that are in SHO state The probability is effected by network planning and parameter settings
SHOs have effect to the network performance Advantages Required to avoid near-far effects Coverage increases when more distant users can connect Capacity can be increased if more users can be connected Alongside with PC, SHO is the main interference migitation means in WCDMA Inconvenient Requires more connections, thus eats DL transmission power and decreases capacity Introduces more interference to DL Increases the traffic in lub 40% SHO probability 1.4 times the traffic!
SHO optimisation
- 13 -
SHO optimisation
Probability for soft HO should be set to 30-50% and for softer HO to 5-15%, depending on the area Too high SHO% results in excess overlapping between cells > other-cell interference increases > capacity decreases Too high SHO% also leads to poorly utilised network capacity (unnecessary links) With too low SHO% the full potential of network is not utilised and transmission powers cannot be minimized > trouble with interference SHO performance is planned with a planning tool and optimised by measurements in live network. In early stage SHO% can be planned high, since the traffic density is smaller. With increasing traffic coverage decreases and SHO areas become smaller. SHO% can be tuned with related parameters and dominance areas SHO most important in urban areas due to serious shadowing
Summary KPI
Indicator Coverage Interference Cell overlap Qualitative KPI KPI target example Measured RSCP > -88 dBm over 97% of area (value should be adapted based on required margins) Measured Ec/No > -9 dB over 95% of area Cell overlay < 3 cells over 95% of area Cell Overshoot No cell detected above -111 dBm (CPICH RSCP) Integrity of cell No cell fragmentation detected coverage Best server plot Clean boundary without unnecessary change of best server
- 14 -
Quality of Service
- 15 -
QoS (ITU-T): << The collective effect of service performance which determines the degree of satisfaction of a user of the service>>. Network Performance, NP (ITU-T): << The ability of a network portion to provide the functions related to communication between users>>.
User domain: throughput, accuracy, dependability (reliability, availability), Provider domain: delay, loss, utilisation,
User QoS Requirements QoS offered by Provider
- 16 -
Trends
Non-technical QoS
Network performance
Terminal performance
Sales points
Customer care
- 17 -
Main task of the UTRAN is to create and maintain RAB for communication between UE and CN. RAB is build up in order to give for CN elements an illusion about fixed communication path to UE. The network builds up the end-to-end QoS connection from small pieces, which compose a complete chain without bottlenecks These pieces are called Bearers When the connection is set up, the network elements negotiate the QoS requirements of the bearers set up between them The result is a compromise, in which the QoS requirements and networks capacity is taken into account.
Example application Speech and video calls Real-time streaming video Web surfing File downloading, e-mails
- 18 -
Allowed transfer delay Set the limits for delay (>80ms) QoS negotiable QoS of some services are not negotiable (speech), packet data services admit various QoS classes
- 19 -
QoS Negotiation
UE
E2E service request Maximum bit rate Guaranteed bit rate Transfer delay QoS negotiable (y/n)
CN
UMTS bearer service: Request for UMTS QoS Class Maximum bit rate Guaranteed bit rate Transfer delay QoS negotiable (y/n) RAB assignment request RRM: Admission control
QoS in UMTS
In early UMTS Release 99 all conversational and streaming class traffic were offered over the CS bearer Voice RT multimedia (e.g. videotelephony) In early Release 99 only Interactive and background class traffic utilisises the PS bearer Release 4 capable networks introduce some streaming class traffic on PS bearer as well Release 5 brings along a full portofolio of PS bearers also utilised for conversation traffic
- 20 -
QoS in UMTS
The QoS over the air interface is implemented by matching each radio bearer with a transport channel whose format set defines the QoS parameters The mapping is performed during the establishment of the RAB RNC performs the mapping of RAB characteristics to actual resource requirements (vendor dependent) Example of mapping for web service, which belongs to the interactive class
Interactive Class
128 kbps 1500 10 -6 NA 64 kbps yes 1% NO
^
Parameters
Maximum bit rate Maximum SDU size Residual BER Transfers Delay Guaranteed bit rate Delivery order SDU Error Ratio Delivery of errorneous
QoS in UMTS
Operators can define the wanted QoS profile (in HLR) per subscriber Users can be categorised (QoS differentiation) for various tariffing schemes Traffic handling priorities can be set (THP)
Business Remote office Basic free time Traffic class All four allowed All four allowed Only converational (voice calls) and background Max bit rate 400 kbps 800 kbps 64 kbps Guaranteed bit 384 kbps 64 kbps 12 kbps rate Allowed THPs THP 1 (e.g. for THP 2 (e.g. for THP 3 e-mail file tranfer) download)
- 21 -
QoS in UMTS
- 22 -
- 23 -
GPS
External antennas
Controler
Energy
Processing
- 24 -
- 25 -
- 26 -
QVOICE
PSTN / ISDN Cellular Network
QVS
QVP-Server
3 parts: QVM (QV Mobile), QVS (QV Stationary) et QVP (QV Post processing).
- 27 -
Analyzer (Tektronix)
BiVision ADC/Metrica, NetAct (Nokia, for 3G) UTRAN Network and service Actix
Analysis tools using these counters (generally they are specific). Example: RNO or NPA of Alcatel, SPOTS from Siemens, etc.
- 28 -
HP: Ovis (data services tests, producers KPIs). RamCom: Network Consultant (A, Gb, Gi, Gn, Iub, Iur, Gi and Gn interfaces) Trafica (NetAct from Nokia) Ipanema: Ipanema (2,5 G and 3G data traffic). Cigale (Astellia): 2 and 3G traffic.
GIS display
- 29 -
K PI
- 30 -
Optimization process
Performance measurements Update of parameters, site configuration Key Performance Indicators (KPI)
Network tuning
Reasons that lead to otimisation: Improve the performance Business reasons (cost-effective) Troubleshooting
Performance analysis
Network statistics
Network statistics are collected from different network elements with counters Different types of counters are used KPIs are needed to provide information of the network performance Raw counter data too detailed to be used in monitoring and optimisation (Some counters can be used as KPIs)
KPI definition
KPIs are composed from several counters KPI categories Accessibility Retainability Integrity Documentation of KPIs is important Same KPI can be defined from different counters or formula can be incorrect Measurement period must be reasonable Too much averaging if too long Not enough statistical information if too short
- 31 -
KPI Example
Optimisation based on KIPs: Optimisation is performed for each category Find the worst performing cells Find the reasons behind the poor performance Make the changes in the network Monitor the performance after the changes
- 32 -
Accessibility
- 33 -
RAB Assignment
Accessibility workflow
Other Modules Performance Measurements
Alarms Cell Availability Counters
Performance Analysis
Idle mode RRC Connection Random Access NAS RAB Assignment
Verification of changes
Squal, Srxlev, qQualmin, qRxLevMin, maxTxPowerUl, t3212, t3312, aichPower, powerOffsetP0, preambleRetransMax, constantValueCprach
100
pmTotNoRab EstablishS uccess < RAB > pmmTotNoRa bEstablish Attempt < RAB >
- 34 -
Circuit-Switched 64
100x pmTotNoRrcConnectReqCsSucc x pmTotNoRrcConnectReqCs pmTotNoRabEstablishSuccessCS64 pmmTotNoRabEstablishAttemptCS64
Circuit-Switched 57
100x pmTotNoRrcConnectReqCsSucc x pmTotNoRrcConnectReqCs pmTotNoRabEstablishSuccessCS57 pmmTotNoRabEstablishAttemptCS57
Where Y =
pmTotNoRrcConnectReqPsSuccess PmTotNoRrcConnectReqPs
Where
Yx pmTotNoRrcConnectReqPsSuccess pmTotNoRrcConnectReqPs
- 35 -
Successful First and Repeated Page attempts of total number of first attempts, Paging success rate in aMSC
100 x NPAAG1RESUCC + NPAG2RESUCC NPAG1GLTOT + NPAG1LOTOT
Paging intensity per cell in a RNC (if RNC, LA and RA consist of exact same cells):
pmCnlnitPagingToldleUeLa + pmCninitPagingToldieUeRa + pmCnlnitPagingToldleUe Measurement period x total number of cells in that RNC
Number/percentage of false detections, which is the case that preamble is detected but there is no enough energy in message part, due to noise on the random access channel for a carrier (it could be due to loss of AICH, wrong recognition of preamble or loss of RACH message part after the UE sends message out):
No of AICH_ACK-No of RRC connection setup-No of cell (re)selection during RRC establishment No of AICH_ACK
x100%
- 36 -
pmTransmittedCarrierPoweri x
i 2
pmTransmittedCarrierPoweri
pmAverageRssii
pmSumOfSampleAseUI pmNoOfSampleAseUI
pmSumOfSampleAseDI pmNoOfSampleAseDI
- 37 -
Code allocation failure for SFn, where n is the spreading factor for a cell could be found in the following formula (as an example the SF 128 was used):
How many users are in compressed mode? Well the average number of users in compressed mode for a cell:
Ratio between RRc connection returning and redirection due to load sharing for a cell:
pmNoOfReturingRrcConn pmNoLoadSharingRrcConn
The failures can be observed by the successful rate of directed retry to GSM for a cell:
pmNoDirRetrySuccess pmNoDirectionRetryAtt
x100%
- 38 -
Retainability
- 39 -
Verification of changes
pmSystemRabRelease<RAB> pmNormalRabRelease<RAB> pmNoSysRelSpeechULSynch pmNoOfTermSpeechCong pmNoSysRelSpeechSoHo UL out of Synch Congestion control, SHO functions IFHO functions IRAT Handovers
Circuit-switched 64
100x
pmNoSystemRabReleaseCs64 (pmNoNormalRab ReleaseCs64 + pmNoSystemRab ReleaseCs64)
Circuit-switched Streaming
100x
pmNoSystemRabReleaseCsStream (pmNoNormalrRab ReleaseCsStream + pmNoSystemRab ReleaseCsStream)
- 40 -
Circuit-switched 64
100x Cs64_U_User pmNoSystemRabReleaseCs64 x number of minutes
Circuit-switched Streaming
100x
Cs57_U_User
The following formula shows the failure rate for RL addition/replacement to active set
pmNoTimesCellFailAddToActSet (pmNoTimesCellFailAddToAct + pmNoTimesRlAddToActSet
- 41 -
The following metric measures the success rate for HS Cell Change in target cell
(pmNoNormalRabReleaseSpeech + pmNoSystemRabReleaseSpeech)
Shows fraction of speech drop due to HO action when a valid or non-valid cell cannot be added to active set. This includes also drop due to missing neighbour.
100x pmNoSysRelSpeechNeighbr (pmNoSystemRabReleaseSpeech + pmNoNormalRabReleaseSpeech)
Shows fraction of speech drop due to missing neighbour reason when a non-valid cell cannot ne added to active set.
- 42 -
Drop due to IHO failure for speech: Outgoing IFHO failure when UE failed to return to present active set.
100x
pmFailNonBlindInterFreqHoFailRevertCsSpeech12 pmAttNonBlindInterFreqHoCsSpeech12
Drop due to IFHO failure for PS greater than 64 kbps: Outgoing IFHO failure when UE failed to return to present active set.
100x pmFailNonBlindInterFreqHoFailRevertPsInteractiveGreater64 pmAttNonBlindInterFreqHoPsInteractiveGreater64
Drop due to IFHO failure for PS streaming and others: Outgoing IFHO failure when UE failed to return to present active set.
100x pmFailNonBlindInterFreqHoFailRevertStreamingOther pmAttNonBlindInterFreqHoStreamingOther
- 43 -
IRAT handover
The following metric measures hard handover success rate between UtranCell and target GSM cell for speech calls. The formula is considering the GsmRelation.
100x pmNoSuccessOutIratHoSpeech pmNoAttOutIratHoSpeech
The following metric measures hard handover success rate between UtranCell and target GSM cell for streaming calls. The formula is considering the GsmRelation.
100x pmNoSuccessOutIratHoCs57 pmNoAttOutIratHoCs57
IRAT handover
The following metric measures hard handover success rate between UtranCell and target GSM cell for Multi-RAB calls. The formula is considering the GsmRelation.
100x pmNoSuccessOutIratHoMulti pmNoAttOutIratHoMulti
The following metric measures cell change failure rate between UtranCell and target GSM cell for PS calls when the UE successfully returns to UtranCell. The formula is considering the GsmRelation.
100x pmNoOutIratCcReturnOldCh pmNoOutIratCcAtt
Congestion
100x
- 44 -
Integrity
- 45 -
The method for finding worst performing cells is based on top to down analysis. Initial worst 10-15 performing cells can be identified based on the Uplink Block Error rate before combining. 100x pmFaultyTranspoertBlocksBcUL pmTransportBlocksBcUI
BLER
Throughput
Throughput =
Average throughput per cell and RAB in the DL, excluding HSDPA:
pmDlTrafficVolume<RAB> pmSum<RAB>RabEstablish pmSamples<RAB>RabEstablish *ROPsec
- 46 -
Payload counters
Radio UL Payload counter DL Payload counter Connection type Speech pmUITrafficVolumeCs12 pmUITrafficVolumeCs12 PS64/64 pmUITrafficVolumePs64 pmUITrafficVolumePs64 PS64/128 pmUITrafficVolumePs128 pmUITrafficVolumePs128 PS64/384 pmUITrafficVolumePs384 pmUITrafficVolumePs384 CS 57.6 pmUITrafficVolumeCs57 pmUITrafficVolumeCs57 (streaming) CS 64 (UDI) pmUITrafficVolumeCs57 pmUITrafficVolumeCs57 Speech/PS pmUITrafficVolumeCs12Ps64 pmUITrafficVolumeCs12Ps64 64 multirab PS pmUITrafficVolumePsCommon pmUITrafficVolumePsCommon Common
- 47 -
HSDPA-HSUPA
- 48 -
Channel Quality Feedback (CQI) UE measures channel quality (SNR or Ec/No) and reports to Node B every 2ms or longer time. Node B chooses modulation scheme, transport block size and code effective rate based on CQI
Good coverage
Bad coverage
AMC could improve radio bandwidth and fit for high speed radio transmission.
DCH Code HSPA Maximum For R99 and HSPA access and traffic channel Code
Power resource
Threshold for R99 load control, which should not be allocated too much to avoid no power for HSPA user
- 49 -
UE3
Mobility management
- 50 -
Bode B
RNC
UTRAN decides to stop HSDPA but keep DCH for the UE due to: Low Downlink data activity High UE mobility
Radio link reconfiguration prepare Radio link reconfiguration ready Radio link reconfiguration commit Radio Bearer Reconfiguration message
- 51 -
Cell 2
Time-to-
Cell change
HS-DSCH on Cell 1
HS-DSCH on
Time
- 52 -
CQI
Make sure that CQI is distributed as appropriate proportion. Cell edge throughput requirement could be fulfilled in the door coverage area.
- 53 -
Performance optimization
Modify the load balance policy Carry out smart admission Power and lub bandwidth
algorithm such as DRD or Downsize Access
Problem Analyze
configuration by UE
- 54 -
Available lub backhaul for HSPA The Data Power to be transmitted Node B RNC
Check point
DATA
SGSN/GGSN
- 55 -
- 56 -