1- What are the main KPI to measure the performance of 3G cell

- Accessibility ( RRC , RAB , CSSR)

- Retainability (speech , Video , PS DCR)

- Mobility (SHO , IRAT HO success rate)

2- What resources affect HSDPA throughput in 3G system

- (DL power, DL code and transport capacity)

3- What parameter tuning can be done to improve HSDPA throughput in any 3G cell

- increase the DL channelization codes for HSDPA

- changing the scheduling algorisms

4- How can we reach 21 Mbps in P7

- 15 codes in DL and 64 QAM

5- what is the usage of the following signaling messages in RRC protocol

- Actives setup updates (ADD/Remove/Replace RL in SHO)

- RB reconfiguration (channel switching between Cell_DCH and Cell_FACH RRC stats)

- Physical channel reconfiguration (IF HO)

6- what is the use of GPEH tool in Ericsson system

- tool used to record RAN and internal events in Ericsson system and the tracing files can be analyzed by TEMS visualization

7- what types of congestion can affect the services accessibility in any 3G cell

- DL power ( AMR - Directed retry - reducing High R.99 RAB users SFxx parameters)

- UL/DL CE ( reducing High R.99 RAB users SFxx parameters)

- DL code (reducing static codes for HSDPA –AMR-Directed retry)

- Transport capacity

8- what is the difference between RSCP and EC/No measures for pilot channel

- RSCP is received signal code power for CPICH channel

- Ec/No is The received energy per chip divided by the power density in the band . it reflects the quality of CPICH channel

9- what is the difference of using 2nd carrier and high power amplifier in expanding the capacity for any 3G cell

- 2nd carrier gives capacity in DL power and DL codes

- High power Amplifier gives capacity in DL power only

10- what is the max bit rate that can be achieved in UL when using 10ms EUL and 2 ms EUL

- 1.5 Mbps for 10 ms EUL

- 5.76 for 2 ms EUL

11- how many HSSCCH channel can be configured in HSDPA cell (

- Four that allows four users per TTI

3G KPI’s and optimization:

1. LOW CSSR: 2. When a low CSSR is detected on a cell the first thing to check is if Admission Control is

rejecting the RRC/RAB setup attempt (pmNoReqDeniedAdm) or if it is failing after admission (pmNoFailedAfterAdm). For high pmNoReqDeniedAdm refer to the “Admission Control” sections below. For high pmNoFailedAfterAdm refer to the “Failure After Admission” sections below.Causes:

Admission Control: DL Power

1. Long term solutions are to increase the power capability of the sector

2. re-engineering the site to reduce feeder lengths

3. The short term solution is to reduce the traffic carried by the site

Admission Control: DL Channelisation Codes

1. This will typically affect the PS Interactive R99 (DCH/FACH) CSSR worse than the Speech CSSR as the PS Interactive R99 RAB requires channelisation codes at a lower spreading factor (using more of the code tree).

2. In the P4 software release a cell that supports R99 and HSDPA typically has 5 spreading factor 16 DL channelisation codes reserved for HSDPA. This means that approximately 32% of available codes are reserved for HSDPA. When this is the case it is common for DL channelisation code congestion too occur

3. The long term solution is to add another cell in the coverage area to take some of the traffic; this may be achieved by introducing a second carrier, another sector, or another site

4. The short term solution is to reduce the traffic carried by the site

5. the required solution is sectorisation of the inbuilding antenna system or implementation of a second carrier frequency

Admission Control: Hardware Usage (Channel Elements)

1. too few channel elements available

2. The channel element capacity of an RBS may be software limited (according the software license configured for the RBS) or hardware limited (according to the TXBs and RAXBs installed in the RBS).

3. Always compare channel element usage to the channel element capacity of the Wcell

4. LOW DCR:

1. If a cell has poor retainability it is typically due to either2. missing neighbour definitions (WCDMA and/or GSM)3. overshooting cell(s), PILLOT pollution.4. a misbehaving neighbour site5. a hardware/software fault or a misconfiguration.6. It is also possible that there is some external source of interference (such as a

microwave link on the same frequency) affecting the retainability.

However, in the majority of cases the factors that affect the Speech retainability will also affect the retainability of the other RABs. When a high speech DCR is detected on a cell the first thing to check is the type of drops occurring as indicated

Soft Handover Drops

Soft HO failures due to missing neighbour orTwo common reasons are a neighbouring cell that is misbehaving (often due to faulty hardware/software) or a misconfiguration resulting in a failure to perform an inter-RNC SOHO across the Iur interface

Missing Neighbour Drops

UL Synchronisation Drops

1. missing IRAT neighbour relation definitions resulting in the connection “hanging on” to the 3G network until the call is dropped when it would be better served handing the call over to the 2G network. This may be especially true for cells on the border of the 3G coverage area

2. By identifying such areas any missing 2G neighbour relations may be added; or perhaps a misconfiguration discovered, such as having an IRAT neighbour relation defined in an RNC towards a 2G cell that is not defined as an outer cell in the 3G MSC Server

3. . Another means to improve the situation may be to lower the thresholds used to trigger IRAT HOs

4. For example, triggering compressed mode at Ec/No=-11dBm instead of -12dBm may prevent drops as calls are handed earlier to the GSM network that typically has better coverage than the 3G network

5. . It is recommended to resolve any SOHO and missing neighbour drop problems before attempting to resolve UL sync drops as often the cause of such drops will resolve UL synchronisation drops too.

2. Other Reasons for Drops

If none of the above reasons for a poor DCR may be established, then it is likely to be a more complicated problem to resolve; often relating to a software/hardware fault, or perhaps an external source of interference in the area

HSDPA

Considerations for HSDPA: DL Channelisation Codes

1. i.e. 32% of the available downlink channelisation codes are reserved for HSDPA

2. feature called HSDPA Dynamic Code Allocation making it possible for the HSDPA Scheduler to only use the channelisation codes available after R99 usage

3. This may help with downlink channelisation code congestion, but at the expense of reduced throughput for HSDPA users. It is still possible to reserve a limited set of codes for HSDPA

3. Considerations for HSDPA: Iub Bandwidth

1. the capacity of the Iub interface between RBS and RNC becomes an important factor effecting HSDPA performance and customer perception

2. i.e. across the Iub interface the R99 traffic has priority over the HSDPA traffic3. As the R99 traffic on the site increases, so the Iub bandwidth available for

HSDPA decreases4. To alleviate the problem some traffic may be offloaded to GSM5. but ultimately additional Iub bandwidth is required unless an additional site,

perhaps an inbuilding site at a shopping centre, may be commissioned to carry part of the load.

HSDPA throughput counters

Overall Flowchart for Capacity Management

IRAT cell reselection:

From 2g to 3g:

QsearchI(VOICE)/P(GPRS): 7 enabled ( MS will look for 3G neighbours measurements)

FddQualmin= -10 EcNo (threshold: meaning if 3G cell is becoming better with good quality .. then reselect to 3G cell—lower values (-8,-6) means that reselection will be delayed as mush better quality is required on 3G side.)

From 3G to 2G:

qQualmin/ Qrxlevmin: -16/ -109dBm (The minimum required quality level in the cell (Ec/No).)

SsearchRAT=2-4dB

GSM measurements will happen when it is not possible to maintain call with good quality: CPICH EcNo < Ssearch_RAT + Qqualmin (default: EcNo < -14 dB)

IDLE mode trials:

1) Improved DCR

a. Qrxlevmin: -115dBm à -109dBm

b. Qqualmin: -18dB à -15dB

c. (SsearchRAT=2dB; fddQmin=-12dB) (on border cells) EcNo < -10dB

2) Objective: Improve the RRC Connection Access Complete Ratio, and as a result the CSSR, by improving the required radio coverage conditions at the edge of coverage. The initial synchronization will be made easier with a better coverage. Also the possibility of inter-system ping-pong will be reduced with a higher hysteresis between technologies.

3) RRC Connection Access Failures on LAC-border cells decreased

AdjsQOffset2=2dB and 4dB has been applied on the ADJS at the LAC border.The default was 0dB

The objective of the tested parameter changes is to reduce the amount of Location Updates, and to ensure that they are performed in better radio conditions in the target cell. Two parameter values will be tested.

Statistics monitored: RRC Connection Access Complete Ratio, Amount of RRC Connections due to registration and inter-RAT. The RRC Completion Ratio might also improve.

4) Soft Handover Parameters for NRT

Objective: Reduce the Soft Handover Overhead for Non-Real Time Services, and as a consequence reduce the load and congestion occurrence

Statistics monitored: Soft Handover Overhead, DCH Setup Failure due to Iub AAL2 Trans for PS data background, Background RAB Completion Ratio, PS Backg Allocated DL

• In Lugano, the tested change decreased the Soft Handover Overhead for NRT, without degrading the RAB Completion ratio

• Addition Window=4dB à 2.5dB

• Drop window=6dB à 4dB

• In Biel there is a very small decrease

• The RAB Setup Failures were probably due to RNC issues, they have appeared with RAN04 and disappeared with CD2.1

• Proposal

• Create FMCS #6 with Add=2.5dB / Drop=4dB on all RNCs

• Apply NrtFmcsId=6 to de-congestion sites or areas

5) Trial 4: Downlink traffic volume measurement high threshold

Objective: Reduce the number of 128kbit/s à 384 kbit/s upgrades for little data volumes. Avoid inefficient use of Iub and WSP capacity.

Duration: 2 weeks

Statistics monitored: PS Backg Allocated traffic for 128 kb/s and 384 kb/s in DL, Radio Bearer Reconfigurations, NRT DCH Requests in DL

Changes:

o Downlink traffic volume measurement high threshold (TrafVolThresholdDLHigh): 1024 (1KB) à 4096 (4KB)

HO Between 2G and 3G:

From 2g to 3g:

QsearchC: 7 enabled ( MS will look for 3G neighbours measurements in dedicated mode)

minimum CPICH Ec/Io level (MET) = -8—10 db (threshold: meaning if 3G cell is becoming better with good quality .. then HO to 3G cell—lower values (-8,-6) means that HO will be delayed as much better quality is required on 3G side).

With this parameter you define the minimum CPICH Ec/Io level of an adjacent WCDMA RAN cell for an inter-system handover attempt. The threshold level must be exceeded before the BSC is allowed to trigger a handover attempt towards the adjacent WCDMA RAN cell.

From 3G to 2G:

HHoRscpThreshold =-105 (Normal)/ -100 (Fast ISHO)

HHoEcNoThreshold=-12 (Normal)/ -10 (Fast ISHO)

The parameter HHoRscpThreshold determines the absolute CPICH RSCP threshold which is used by the UE to trigger the reporting event 1F. When the measured CPICH RSCP of all active set cells has become worse than or equal to the threshold in question, the RNC starts inter-frequency or inter-RAT (GSM) measurements in

compressed mode for the purpose of hard handover.

dU setting:

Related Parameters:

For RSCP:

HHoRscpThreshold -105 (Normal)/ -100 (Fast ISHO)

Related parameters  HHoRscpTimeHysteresis 100ms/1280ms(dU)HHoRscpCancel -102 dBmHHoRscpCancelTime 640ms/ 320(dU)   GSMcauseCPICHrscp enabled/disabledIFHOcauseCPICHrscp enabled/disabled

For EcNo:

HHoEcNoThreshold =-12 (Normal)/ -10 (Fast ISHO)Related parameters  HHoEcNoTimeHysteresis 640ms/640ms(dU)HHoEcNoCancel -9 dBHHoEcNoCancelTime 100ms/320ms(dU)GSMcauseCPICHEcNo enabled/disabledIFHOcauseCPICHEcNo enabled/disabled

HHoRscpTimeHysteresis=100ms

The parameter HHoRscpTimeHysteresis determines the time period during which the CPICH RSCP of the active set cell must stay worse than the threshold HHoRscpThreshold before the UE can trigger the reporting event 1F.

GSMcauseCPICHrscp of FMCG

This parameter indicates whether a handover to GSM caused by low measured absolute CPICH RSCP of the serving cell is enabled.

IFHOcauseCPICHrscp of FMCI

The parameter indicates whether an inter-frequency handover caused by low measured absolute CPICH RSCP of the serving cell is enabled.

HHoRscpCancel = -102 dBm (3dB difference is needed between trigger and cancel parameters.)

If the inter-frequency or inter-RAT (GSM) handover caused by low measured absolute CPICH RSCP is enabled, the RNC starts the inter-frequency or GSM measurement in compressed mode when all active set cells have triggered the reporting event 1F for CPICH RSCP. The RNC cancels the event 1F of an active set cell, if the CPICH RSCP measurement result of the active set cell becomes better than or equal to the threshold HHoRscpCancel and the UE transmits the corresponding event 1E triggered measurement report to the RNC.

HHoRscpCancelTime =640 ms

The parameter determines the time period during which the CPICH RSCP of the active set cell must stay better than the threshold HHoRscpCancel before the UE can trigger the reporting event 1E.

AdjsEcNoOffset.( CPICH Ec/No Offset)

The CPICH Ec/No Offset determines an offset value, which the UE adds to the CPICH Ec/No measurement result of the neighbouring cell before it compares the Ec/No value with the reporting criteria.

AdjsDERR(Disable Effect on Reporting Range)This parameter indicates whether the neighbouring cell is forbidden

to affect the reporting range (addition/drop window) calculation, if it belongs to the active set.

ISHO process:

Call setup in 3G

RRC CONNECTION SETUP TIMERS:

1.1 Test Procedures

TEST CASES DRIVE TEST METHODOLOGY

Radio Design Performance Check Drive Test to be done using two UEs and a Scanner. (MS1-in idle Mode and; MS2 –in Dedicated Mode). Both UEs are Locked to 3G Only.

Coverage and quality performance check

1. Coverage performance of each sector (Best Server RSCP)

2. Quality performance of each sector(Best Server EcNo)

Call Performance Check. Short calls: Static test on each sector. Call duration is 20 seconds and 10 seconds idle.

Long Calls: Drive around within the coverage objective of each sector.

Handover Performance Check Handover performance to be verified by driving in clock wise and anti clock wise around the Site, for all three sectors. Then perform handover test towards its surrounding neighbors (at least 3 neighbors)

Data Performance Check R99 & HSPA test for the following:

1. PDP context activation (5 attempts)

2. Throughput test (based on 2MB downloaded file from the FTP server.

Supplementary Services Check 3 SMS/MMS per cell at different locations with 10 sec break after each SMS/MMS

Acceptance Criteria:

Using R99 384kbps bearer, average FTP DL throughput should not be less than 340kbps.

Using HSDPA bearer, average FTP throughput should not be less than 3Mbps (Without transmission limitation).

Test method: The data transfer will be from a test FTP server. The file format should be a txt or data file.

Acceptance Criteria:

The SHO between the sectors of same site should be successful

The SHO between the sites should be successful.

Site acceptance performance test should be performed in static mode and in good RF conditions, CPICH_RSCP > -80dBm & CPICH_EcNo > -8dB

Acceptance Criteria:

The scramble configuration is consistent with the planned configuration. No cross TX feeder observed.

1.2 Drive Test Route

The drive to be performed considering that the Scrambling code of the desired WCDMA cell is the primary server for most of the time.

Coverage observation Quality observation

Sector 1 The average coverage within 800m is about -75dbm. No obstruction observed

The average Ec/No is larger than -10

Sector 2 The average coverage within 1Km is about The average Ec/No is larger than -10

-75dbm. No obstruction observed

Sector 3 The average coverage within 1Km is about -75dbm. No obstruction observed

The average Ec/No is larger than -10

Coverage performance – Scanner mode

Acceptance Criteria:

If no special note is for a cell, the related CPICH RSCP should be larger than –70dBm within 100m from the Node B.

1.3 Quality performance – Scanner Mode (EcIo)

Acceptance Criteria:

If no special note is for a cell, the related CPICH Ec/Io should be larger than –10dB within 100m from the Node B.

1.4 Coverage Performance – Dedicated Mode (CPICH RSCP)

Acceptance Criteria:

If no special note is for a cell, the related CPICH RSCP should be larger than –75dBm within 100m from the Node B.

1.5 Quality Performance – Dedicated Mode (Ec/No)

Acceptance Criteria:

If no special note is for a cell, the related CPICH Ec/Io should be larger than –10dB within 100m from the Node B.

1.6 Active Set Size Plot

Acceptance Criteria:

The plots of Active Set (AS) size=3 should not be more than plots of Active Set size=2 or 1.

EVENTS:

hese different events that a UE reports..like:Event 1x: related to intra freq events.Event 2x: related to interfreq measurementEvent 3x: related to inter-RAT measurement where x can be a, b, c, d, e, fthese are the 3 important categories... There description is as follow:

Event 1A:  "A Primary CPICH enters the reporting range", it is defined by the EcNo or can be  RSCP lev; depends on the critera whether it is EcNo or RSCP. In du case it is EcNo and parameter related to event 1a is "ADDITION WINDOW" which is set to be either 3db or 4 db here. Simply we can say that addition of a cell in active set is triggered by event 1a.

Related Parameters EVENT 1A: AdditionWindow (4 dB default) AdditionTime (100ms: 6 internal value) AdditionReportingInterval (0.5 sec : 2 internal value) ActiveSetWeightingCoefficient (0)4 MaxActiveSetSize (3)

are used in case of event 1A 

Event 1B: "A primary CPICH leaves the reporting range" similar to the above but is for deletion of a cell from active set, Drop Window is set to be either 5 or 6db here.

Related Parameters EVENT 1B: DropWindow (6 dB) Drop Time (640 ms: 12 internal value) ActiveSetWeightingCoefficient (0)

are used in case of event 1B. 

Event 1C: "A non-active primary CPICH becomes better than an active primary CPICH" it is basically for replacement of a CPICH in AS (active set).  AS size is 3 which means Max no of CPICH in soft hand over can be 3. Now suppose there are 3 cells in AS and a 4th pilot becomes stronger than any of the 3 in AS, only replacement can be a solution which is triggered by event 1c. This is taken care by a 2db replacement window and of course hyst as well.

Related Parameters EVENT 1B:

ReplacementTime (100ms: 6 internal value) ReplacementWindow (2dB: default) ReplacementReportingInterval (0.5 sec:2 internal value)

are used in case of event 1C. Event 1D:  Change of best cell

Event 1E: A Primary CPICH becomes better than an absolute threshold...it is basically related to cancellation of Hard handover..related params in nokia are HHOEcNoCancel or if criteria is RSCP then HHORSCPCancel. 

 Event 1F: "A Primary CPICH becomes worse than an absolute threshold"....basically triggers HHO to GSM in nokia ( wheras huawei is using event 2D to do the same)...related params are HHOEcNoThreshold or HHORSCPThreshold...if event 1f is reported by allcells in AS, UE goes into compressed mode and if all criteria met..triggers HHO to GSM. So this is about event 1 description....all these events can be seen/analyzed in layer 3 messages... 

Huawei Interview questions:

Kind of question asked by Huawei during interviews: 

              HSPDA : High speed downlink packet access (HSDPA), it is like in GSM we have edge/gprs for data service. In 3G HSPA (HSDPA, HSUPA) do provide high speed data.

       2: HSUPA: High speed uplink packet access, it provides high speed PS data in uplink like HSDPA does in the DL.

 

         3:BTS and RNC if possible hardware description ( Boards and elements used):

Huawei is using BTS3900, which is attached here..         4:Capacity calculations what is the SHO overhead, how many channel elements will be needed, how big will be the Iub, based on presented traffic model

SHO overhead is around 30-40%, 384 CEs are configured on each cell, Iub based interface here is IP based with radios upto 21Mbps per sector.          5:Differences between planning based on coverage and on capacity

Main idea is same as is in 2G, here one has to be more careful to control overshoots to have good EcNo and less or no pilot pollution.         6:What does it mean a noise raise of 3db? 50% coverage reduction

This is Noise rise at Node B ie in the UL, which corresponds to 50% loading of the network. the %age is of pole capacity which is basically theoretical max cap.         7:What is the expected noise floor in uplink and in downlink? 

The term noise floor is related to UL, in term of RTWP it is less than -100dbm, with loading noise floor keep rising..

       8:  Site search cycle (from nominal plan site selection, site survey site ranking, final selection, final report, acquisition)

Just similar to 2G...         After integration

        9: Site health checks: (Alarms, site acceptance drive test. Footprint check, SC check) 

HSDPA & R-99 Throughput, sector swap, hand over tests, IRAT HOs, reselection, functionality test that sites coverage is normal, VSWR and other alarm checks.10:Explain criteria’s thresholds and process of cluster acceptance: Example 80% of sites radiating without problems

Cluster acceptance depends on customers mood:)Generally we have to have CDR less than 2%, RAB set up sucess rate is 98% for all types of RABs       11:  What sort of checks to do in a Swap case taking in consideration that you need to have the same KPI’s before and after 

12:Description of used tools

Actix, Tems, M2000 for different reports generation, LMT for changing/config params, U-Net planning tool. Nastar for viewing RNC logs (data) dump and other RNC data can be loaded directly to Nastar and from there all states can be viewed.        13: Elaborate on Monte Carlo simulations and on Okomura Hata propagation model

Okumara Hata Model is on the same lines as is in 2G or in theory.     Monte Carlo Simulations are tool based analysis to analyze access failures under different loading conditions which are configured before simulations are run. 

CHIP:The bits in the spreading code are called chips.

GMSK shifts one bit at the time, QPSK two bits at thetime, and 8-PSK three bits at a time

The effective bandwidth for WCDMAis 3.84 MHz, and with guard bands the required bandwidth is 5 MHz. The guardbands are needed to reduce the interference between different 5 MHz WCDMAcarriers.

Soft HO capacity gain:

As a conclusion it can be stated that soft and softer handovers consume radioaccess capacity because the UE is occupying more than one radio link connectionin the Uu interface. On the other hand, the added capacity gained from theinterference reduction is bigger and hence the system capacity is actually increasedif soft and softer handovers are used.

Codes - What and why?A code is a specific sequence of bits applied to data to scramble the information.

A symbol is an information unit, transmitted via the radio interface.Downlink each symbol represents two bits.

One bit of the code signal used for signal multiplying is called a chip.

The code signal bit rate, which is hereafter referred to as the chip rate, is fixed inWCDMA, being 3.84 million chips per second (Mcps/s). With this chip rate thesize of one chip in time is 1 / 3 840 000 seconds.

Spreading factor

K = 2k,For instance, if k = 6, the spreading factor K gets the value 64, indicating that onesymbol uses 64 chips in the WCDMA radio path.Another name for spreading factor is processing gain (Gp), and it can beexpressed as a function of used bandwidths:

The scrambling codes are divided into 512 code sets, each of them containing aprimary scrambling code and 15 secondary scrambling codes.

In some cases, we are able to perform so-called puncturing in case the bit rate issomewhat higher than certain allowed bit rate. Puncturing means that we removesome of the redundant bits from the channel coding, thus reducing the bit ratedown to the wanted level.

Rate matching

After the error protection, the baseband data rate is matched to the bearer bit ratesused in the UMTS radio interface. The data rates are given with the availablechannelisation codes, resp. with the given spreading factors.In the figure below, you can see different bit rates that may be applied for userdata. The bit rates range between 15 and 960 kilobits per second.

Please note that these bit rates are not the same as the application or user data rates.To exemplify this, let us imagine a speech call at 12.2 kilobits per second. Thesebits must undergo channel coding in order to enable error correction. After thischannel coding (convolutional coding), the bit rate is around24 kbit/s. Now we need to rate match this data to the closest allowed bit rate inUMTS, which in this case is 30 kbit/s (as you can see from the figure above). Werepeat some of the encoded data bits according to a certain pattern in order toincrease the bit rate to the wanted level.

UMTS channel structure

Radio Resource Control RRC states

When we have a dedicated channel open for a subscriber (for example, if we areusing video), we say that subscriber is in Cell_DCH state. (The DCH is derivedfrom the name of the channel in the air interface). In this state the UE is sendingmeasurement reports to the network, thus the system can control the dedicatedbearer and perform handovers.If the mobile is only sending small pieces of information, for example intermittentInternet based traffic or for signalling, then the RRC can be in a mode known asCell_FACH (the FACH stands for Forward Access Channel) and is different fromthe previous state as no dedicated channel is used. The network does not performhandovers as the mobile moves from one cell to another. The UE just informs the

network of its current location.Depending on the bearer we have and how it is being used, the RNC will move theRRC between the different states. In addition to the Cell_FACH, if the networkfinds that the bearer is not being used for a long time, it can move the connection toa Cell_PCH mode (Paging Channel), where the mobile is still know to a cell levelbut can only be reached via the PCH. In this state the UE is using a DiscontinousRepetition Function (DRX) to save battery. Again, unlike in the Cell_DCH, as thesubscriber moves, the mobile informs the RNC which cell it has moved to. Thefinal state is the URA_PCH. This state is similar to the Cell_PCH. But, instead ofmonitoring the connection on a cell level, it is now on a RNC level. URA standsfor UTRA Registration Area and the UE monitors the broadcast channel for URAidentities.

Admission control

The main task of admission control is to estimate whether a new call can haveaccess to the system without sacrificing the bearer requirements of existing calls.Thus the AC algorithm should predict the load of the cell if the new call isadmitted. It should be noted that the availability of the terrestrial transmissionresources is verified, too, meaning that there is no limiting factor in the rest of theUTRAN either. Based on the admission control, the Radio Network Controller(RNC) either grants or rejects the access.

Planning uplink admission controlPRX_Target (ReceivePower level) value.The area from 0 to this value is known as the planned load. Once the load isapproaching this value, traffic reason handovers (TRHO) are performed.

As UMTS traffic is variable and constantly changing, it is more than feasible that

the traffic admission may exceed the PRX_Target. To handle this situation, there isa second level used called the PRX_TARGET_BS. This is a parameter used bythe BTS to stop situations of congestion. Once this value is reached, the BTS takesactions to reduce the load in the cell.

Code allocation

Both scrambling and channelisation codes used in the Uu interface connections aremaintained by the RNC. In principle they could be maintained by the BTS, but thenthe system would experience lack of radio resource control, namely softhandovers, which will be explained later in this module. When the codes aremaintained by the RNC, it is easier to allocate Iub data ports for multipathconnections.The Uu interface requires two kinds of codes for proper functionality. A part of thecodes used must correlate with each other to a certain extent, and the others mustbe orthogonal (they do not correlate at all). Every cell uses one scrambling code.As you already know, this code acts like a cell ID. Under every scrambling codethe RNC has a set of channelisation codes. This set is the same under everyscrambling code.

The BCH information is coded with a scrambling code value, and thus the UE mustfirst find the correct scrambling code value first in order to access the cell. When aconnection between the UE and the network is established, the channels used mustbe separated. The channelisation codes are used for this purpose. The informationsent over the Uu interface is spread with a spreading code per channel. Spreadingcode by definition is the same as scrambling code x channelisation code.

Channelisation code allocation and handovers

Fragmentation of code treephenomenon where the probability of the blocked branch of the code tree increases too much and thus it starts to prevent new accesses to the system.

For example, if an active call uses high bit rateover the Uu interface, the spreading factor value in use is small. It furthermoremeans that a very high-level branch of the code tree is blocked (see the figurebelow). When this call is finished and simultaneously new calls access the system,the blocked code tree branch is not “released” before the new accesses. In thissituation the system wastes capacity because the code channels allocated for newcalls are not necessarily allocated in the best possible way.

As codes are released in different branches, the tree can become fragmented andthe RNC should always try to reorganise the tree to make the best use of theresources. Therefore in UMTS networks, it is possible that the channelisation codescould change during a connection.Also, if the scrambling code in the uplink (that is, the user) is being used byanother person in another RNC as the subscriber performs a soft handover, the

handover is refused and the serving RNC must allocate a new scrambling code tothe subscriber.

Power control

Inaccuracy in power control immediately increases interference, thusdecreasing the capacity of the network.

Open loop power controlWhen the UE accesses to the network, the initial level for accessing is based on anestimate. This estimate in turn is based on the signal level received from the NodeB when the UE is in idle mode. The basis for the UE estimate is the downlinkpower level that the UE detects from the physical channel CPICH.Closed loop power controlWhen the radio connection is established, the power control method is changed.During the connection, the method used is called the closed loop power control.Within this method, the Node commands the UE either to increase or to decreaseits transmission power with the pace of 1.5 kHz (1500 times per second) in theFDD mode. (In the TDD mode closed loop power control is performed 100 to 800times a second.) The decision whether to increase or decrease the power is basedon the received Signal-to-Interference Ratio (SIR) estimated by the Node B.Outer loop power controlDue to the macro diversity (the UE is simultaneously attached to the networkthrough more than one cell), the RNC must be aware of the current radio linkconditions and quality. The RNC knows the allowed power levels of the cell andtarget SIR.

Packet schedulerPacket scheduler is a general feature, which takes care of scheduling radioresources for non-real-time (NRT) radio access bearers for both uplink anddownlink directions.The gap between real-time (RT) traffic and the load target of the cell canbe filled by the packet scheduler.

An active set is a list of cells,through which the UE has a connection to the network, that is, through which theradio link set-up has been made. This is, the UE may have active radio connectionbetween itself and the network through three cells simultaneously. In softhandover, the UE is connected to (at least) two Node Bs at the same time. In theuplink direction, the two signals come via the base stations to the Radio NetworkController (RNC). In the RNC the signal to be transported forward to the corenetwork is selected. The selection is done frame by frame for the speech, and insmaller blocks for data

slotted/ compressed modeThe possibility to perform an inter-system handover is enabled in the UMTS by aspecial functioning mode, slotted mode. When the UE uses Uu interface in theslotted mode, the contents of the Uu interface frame is “compressed” in order toopen a time window, through which the UE is able to peek and decode the GSMBCCH information.

Intersystem handover from GSM

The handover is described in the figure 44.• The UE/MS creates a measurement report that the BSC evaluates to make thehandover decision.• If the BSC decides to hand over to a UTRA cell resource reservationmessages are sent to the UTRAN.• The UTRAN acknowledges the resource reservation and provides a UTRANhandover command.• The BSC sends the GSM intersystem handover command to the UE. In thiscommand is included a UMTS Handover to UTRAN command whichcontains all the information needed to set up a connection to the UTRA cell.Since the amount of UTRA configuration information might be too large fora GSM message the message actually contains reference number to UTRAparameters not the real values.• The UE completes the procedure by a Handover to UTRAN completemessage to the RNC.• As a last stage the RNC commands resources to be released by the BSC.

Intersystem handover from UTRANUMTS has already been deployed widely but there will still be reasons to performIntersystem handovers to GSM. This could be because of services used, coverageor even traffic reasons.

Based on the measurement report including both UTRAN and BSS values theRNC makes the handover decision.• Resource reservation messages are sent to the BSC.• The BSC acknowledges the resource reservation and includes a GSMhandover command.• The RNC sends an Intersystem handover command message to the UE,included in this message is the GSM Handover command.

• The UE switches to GSM RR protocol and sends the GSM handover accessmessage to the BSC.• The BSC finally initiates resource release with message to the UTRAN

Micro diversityReferring to the soft handover and active set, there are two terms describing thehandling of the multipath components: micro diversity and macro diversity. Microdiversity means the situation where the propagating multipath components arecombined in the Node B.Macro diversityBecause of the fact that the UE may use cells belonging to different Node Bs oreven different RNCs, the macro-diversity functionality also exists on the RNClevel. The following picture presents a case in which the UE has a 3-cell active setin use and one of those cells is connected to another RNC. In this case, the NodeBs do signal summing concerning the radio paths of their own. In the RNC level,the serving RNC evaluates the frames coming from the Node Bs and chooses thebest signal to send towards the CN domains.

Summary of code usage in the uplink and downlink direction:In the downlink direction, we need to be able to make a difference betweendifferent cells. Therefore the scrambling code is used for this purpose. Since wealso must make a difference between different users within the cell, thechannelisation code is used for this purpose. It also means that we will only use onededicated physical channel in the downlink direction, and both signallinginformation (such as power control commands) and application data must bemapped onto this physical channel.In the uplink direction, we do not need to separate between cells. It means that wecan utilise the scrambling code for separating between the different users.Following the same logic as earlier, it means that we can use the channelisationcode to separate between different channels.

CDMA sequencing – a way to spread informationSequencing as a term refers to how the information to be transferred over the radiopath with CDMA technology is spread over the defined frequency band. There aretwo basic alternatives: Frequency Hopping (FH) and Direct Sequencing (DS).

In UMTS Release 99, there are two WCDMA modes:• FDD modeFDD stands for frequency division duplex. Two separate 5 MHz frequencybands are used – one for uplink transmission and another one for downlinktransmission.• TDD modeTDD stands for time division duplex. Hereby, one frequency band is usedboth for uplink and downlink transmission.

RNC tasksThis subsection gives a summary of the main tasks and functions:WCDMA radio resource management• Radio resource management of channel configurations, that is, how manytraffic channels and signalling channels can be used in the RAN. This isdone in connection with the radio network planning.• Management of traffic channels and stand-alone dedicated control channelscan be further divided into radio resource management that attends to codeallocation, admission control, channel release, load control, power controland handover control.• Handovers are controlled by the RNC, but can be initiated by the mobilestation (MS) or the RNC. A handover can be one or more of the followingtypes:− Soft, softer and hard HO with intra- and inter-RNC handovers− Code switching and code type switching HO.Telecom functionalityThe telecom functionality includes tasks that are much related to the mobility andsession management of subscribers and their connections:• Location and connection management• Ciphering• Indication of blockage on the channels between the RNC and the MSC• Allocation of the traffic channels between the RNC and the base stations• ATM switching and multiplexing• ATM transmission on SDH or PDH• GPRS tunnelling protocol (GTP) towards the packet core network• Security functions

MaintenanceOperation

High Speed Downlink Packet Access (Rel.5)

High Speed Downlink Packet Access (HSDPA) is a feature based on a downlinkshared channel reserved for transfering only data. This feature allows data rates upto 10 Mb/s.The new high speed access is based on the new High Speed-Downlink SharedChannel (HS-DSCH) transport channel, which keeps some of the characteristics ofthe Release 99 DSCH. It is defined for FDD and both TDD modes. It is a timeshared channel, mapped to one or more physical data channels. A new physicaldownlink data channels is defined (HS-PDSCH), together with an associateddownlik control channel for layer 1 signalling (HSSCCH). An uplink signallingchannel is also required, HS-DPCCH, based in the standard DPCCH. One of themain characteristics of HSDPA is the advanced link adaptation: the transmissionscheme changes every Transmission Time Interval to adapt to the radio linkconditions.HSDPA uses link adaptation AMC, adaptative modulation and coding withseveral predefined combinations of modulation and channel coding. The Node Bselects the modulation and the coding for each TTI for each user based on anestimate of the downlink. The UE reports in the uplink signalling a measurement ofthe downlink. Higher order modulations (16 Quadrature Amplitude Modulation)will be used in good radio link conditions and lower schemes (Quadrature PhaseShift Keying ) are used in poor radio conditions to maintain the error rate.Automatic Retransmission Query (ARQ) is an error detection mechanism usedin the link layer that brings controlled efficient retransmissons.The HSDPA benefits are:• support for services requiring high data rates in downlink, e.g. Internetbrowsing and video on demand.• High data rates up to 10Mbit/s