UMTS Planning

©Informa Telecoms

Understanding UMTS

Radio Planning Issues

©Informa Telecoms

Understanding UMTS


1 W-CDMA PLANNING1.1 The Requirements 1

2 CELL PLANNING2.1 Traditional Cell Planning 32.2 W-CDMA Cellular Planning Principles 52.3 Soft Handover Regions 72.4 Cell Breathing 9

3 INTERFERENCE EFFECTS3.1 Own and Adjacent Cell Interference 113.2 Multi-path Effects 133.3 Interference Sharing and Soft Capacity 153.4 Limitations 173.5 User Data Rates and Number of Users 19

4 REDUCING INTERFERENCE/INCREASING CAPACITY4.1 Sectorisation 214.2 Smart Antennas 234.3 Multi-Carrier Cells 254.4 Planning the Frequency Spectrum 274.5 The Layered Architecture 284.6 GSM Co – Planning 314.7 Use Of Existing Sites 334.8 W-CDMA, EDGE and GSM Coverage Areas 35

5 RADIO PLANNING AND SIMULATIONS5.1 Planning Tools 375.2 Monte Carlo Simulations 37

6 SUMMARY – PLANNING CONSIDERATIONS 39

Understanding UMTS

1. W-CDMA PLANNING

1.1 The Requirements

When planning any radio network there are three overall requirements to be satisfied.

Coverage planning should allow services to be provided continuously over the area of operation. Fortunately in UMTS, the UTRAN can exist alongside the GSM radionetwork, or any other compatible (radio or fixed) access network, with handoversallowed between the different systems. This, together with the different W-CDMAmodes of operation (FDD and TDD) allows a great deal of flexibility in coverageplanning.

Sufficient capacity should ensure that calls can be completed, or data transferredwith a high probability of success. The higher the success rate (Grade of Service)planned for, the more equipment that will need to be provided.

Quality of Service must be maintained at an acceptable level. This becomes muchmore of an issue with UMTS due to the range of services (all with varying needs ofdata rate, delay tolerance, error rates etc.) which can be supported.


©Informa Telecoms1

Fig. 1 – Planning Requirements

2©Informa Telecoms

• Coverage

• Capacity

• Quality

Understanding UMTS

2. CELL PLANNING

2.1 Traditional Cell Planning

For second generation Time Division Multiple Access (TDMA) systems, includingGSM, the air interface was organised into equally spaced carrier frequencies, each ofwhich could support a finite number of users separated by the use of recurring timeslots (in GSM, eight timeslots existed per carrier frequency).

The TDMA notation refers to the use of timeslots, but the equal spacing of the carrierfrequencies is described as Frequency Division Multiple Access (FDMA). Hence GSMis actually a TDMA / FDMA system. In addition, for each uplink frequency there is aseparate, but corresponding downlink frequency. This is known as Frequency DivisionDuplex (FDD).

The frequency and time allocation in second generation networks ensures that thedifferent control data and user data can be kept separate within a given geographicalarea. This can only work if the available carrier frequencies are planned to minimiseinterference. This is achieved by careful control of transmitter powers, and stipulatinga minimum distance between transmitters using the same frequency (minimum re-usedistance) for a given quality of signal.

The radius of the nominated coverage area for each base site (cell) is thereforeplanned to be significantly smaller than the stipulated re-use distance. This ratiodepends on the interference that can be tolerated in each system (GSM is fairlytolerant, hence the re-use distance for a given power is relatively small).

Since power is generally set to provide sufficient coverage for the cell in question, thesmaller the cells, the lower the re-use distance. Hence, with careful power control,cell planning becomes purely a geometrical problem.

In general, a tessellating pattern is used to provide coverage over the requiredgeographical area, as shown opposite. The greater the tolerance to interference, thelower the ratio between re-use distance and cell radius need be, and hence thesmaller the number of cells in the pattern. GSM is more tolerant than the analogueTACS system and can cope with a four cell repeat pattern rather than the minimum ofseven needed in TACS. However, as always, quality must be taken into account whendeciding on the repeat pattern.

For a finite set of available carrier frequencies, the larger the pattern, the lower thenumber of available frequencies per cell. Therefore more cells may be needed withinthe required geographical area in order to provide the same capacity. This leads tohigher infrastructure costs.


©Informa Telecoms3

2

1

5

6

7

4

3

2

1

5

6

7

R

4

3

D

3

Re use distance D:

D = R 3N

eg: 7 Cell Re-use

Pattern

Indicates cell withset frequency (s)

Cluster sizeN =

=

= Cluster of cells

Interference considerationsdictate that for:

TACS, Cluster size 7GSM, Cluster size 4

Fig. 2 – Traditional Cell Planning

4©Informa Telecoms

Understanding UMTS

2.2 W-CDMA Cellular Planning Principles

The W-CDMA planning concept initially seems simpler than for TDMA systems. It isbased generally on a single cell repeat pattern, where the same W-CDMA carrierfrequency can be used in adjacent cells continually throughout the network. Usersand control data are separated by the use of codes within the spreading anddespreading process. The ability to despread however mainly depends on the datarate and therefore spreading factor / processing gain, and on the overall interferencereceived along with the wanted signal.

The received interference originates from both the same cell (as the wanted signal),and from adjacent cells. In both cases, interference can be minimised by carefulpower control (fast power control in UMTS is performed 1500 times per second) andby controlling the overall loading of each cell (number of users and their aggregatedata rates).

The recovery of the wanted signal depends not only on the despreading process, butalso on the rejection of the interfering signals. So long as different spreading codesare used, and they are orthogonal (with good cross correlation characteristics), theinterfering signals will remain spread during the despreading process of the wantedsignal.

In addition, code allocation and required data rates are linked in W-CDMA by thenature of the code tree (used to ensure orthogonality). For both these reasons, codeplanning is essential.


©Informa Telecoms5

1

1

1

1

Fig. 3 – W-CDMA Cellular Planning Principles

6©Informa Telecoms

• Spread signals from same and adjacent cells contribute to interference levels- GOOD POWER CONTROL REQUIRED- CAPACITY OF EACH CELL NEEDS PLANNING

• This interference remains spread on despreading wanted signal, unless same spreading codes are used, or poor correlation characteristics exist – CODE PLANNING REQUIRED.

Understanding UMTS

2.3 Soft Handover Regions

The soft handover (between cells provided by different Node Bs) and softer handover(between cells provided by the same Node B) allows mobiles to be served by morethan one radio interface connection at the same time.

This is an important concept in W-CDMA in that it allows the mobile to use eachseparate signal to enhance the overall recovered signal. The soft and softer handoverregions occur generally at the edge of cells where power requirements wouldotherwise be at their greatest. However, the gain introduced by the soft and softerhandover allows for lower powers to be used, minimising the contribution to overallinterference levels. The near-far effect is also mitigated, (where a mobile closer to theNode B introduces a level of interference into the system that precludes successfuldespreading of a mobile on the edge of a cell).

When planning the network, the soft and softer handover regions must allowcontinuous coverage, whilst minimising overall interference under all cell loadconditions. The trade off is that more hardware is required if the average time spent insoft or softer handover increases. This includes Rake Receivers in the Node B, andUTRAN transmission links for soft handover (where combining is done at the RNC).

Note that combining techniques for soft and softer handover are different. Softerhandover utilises the Rake receiver to combine the signals in much the same way thatmulti-path signals can be combined. Whilst soft handover combining is achieved inthe RNC, where the different signals are assessed, and the best signal is chosen forinclusion in the combined signal (choice made every 10-80ms).

Realistic figures for connections spent in soft handover may be 20-40%, whilst thosespent in softer handover may be 10%.


©Informa Telecoms7

8©Informa Telecoms

Fig. 4 – Soft Handover Regions

• Mobiles may be served by more than one base station site in soft handover(shaded) areas

• Continuous coverage should be maintained (if required) under all loadconditions

• The combining process enhances the signal in soft handover

1

1

1

1

Understanding UMTS

2.4 Cell Breathing

If the number of users in a particular area increases, or their aggregate data ratesincrease, more interference is introduced into the overall system. Despreading ofsignals for mobiles in that area becomes more difficult, and the effective cell range fora given power reduces. Unfortunately, the adjacent cells, which may also be servingthe same area (soft handover), will encounter the same problem. This means that forhigher loading, the effective radius of all cells serving the area in question will reduce(for the given power), reducing the area of soft handover.

The variation in effective radius with loading is a phenomenon known as cellbreathing, and it must be taken into account when planning the handover regions.


©Informa Telecoms9

2

3

1

xx

xx

xx

x

x

Lower Load

2

3

1

xx

xx

xx

xxx

xx

x

xx

Higher Load

Fig. 5 – Cell Breathing

10©Informa Telecoms

• Effective range of cell is reduced on higher loading due to interference causedby additional channels

• Adjacent cells also breathe• Soft handover region reduces

Understanding UMTS

3. INTERFERENCE EFFECTS

3.1 Own and Adjacent Cell Interference

Since the same W-CDMA frequency is used in each cell, the total interference in asystem is a combination of that generated within the same cell, and that generated inadjacent, or nearby cells.

The figure opposite shows the relationship between the spread wanted signal, theinterference contribution from own cell and adjacent / nearby cells, data rates (andtherefore the subsequent processing gain), power and range.

The result is that power, coverage area, data rates (and subsequent spreading factor /processing gain) per user, and the overall loading of the cell all need carefulconsideration when implementing a W-CDMA network.

Interference, capacity and coverage are interdependent and must be consideredtogether.


©Informa Telecoms11

WantedSignal

Interference

Interference

Spread Wantedsignal

• Higher data rate, there-fore lower processing gain – insufficient to raise wanted signal sufficiently above noise

Interference

Spread Wantedsignal

• Wanted signal despread and raised sufficiently to recover signal (Processing Gain is sufficient/data rate is low enough)

Interference

Increased powerof spreadwanted signal

• Power of spread wantedsignal is increased by reducing range, or increasing transmitted power (Increasing trans-mitted power adds more interference for other users)

Need to control: • Power• Range (coverage)• Processing Gain/Spreading Factor per user• Overall loading

Fig. 6 – Interference, Capacity and Coverage


Understanding UMTS

3.2 Multi-path Effects

Unlike in most radio systems, the Rake reception of the W-CDMA signals means thatmulti-path effects can be used to enhance the signal.

Any signals that take different paths from source to destination will travel differentdistances and arrive at the destination at slightly different times. In W-CDMAhowever, the different multi-path components of the same original signal can then bedespread separately by the same code, so long as the code has been delayed by anappropriate amount to account for the path differences. Once the multi-pathcomponents have been despread, they can be adjusted in time so that the despreadsignals coincide with each other, and then added together to give a composite signal.



Delay due todifferent paths

Received Components

Multi-pathPropagation

Rake Output

Fig. 7 – Multi-path Effect


• By delaying code generation in the Rake Receiver, multi-path can be used toenhance recovered signal.

Understanding UMTS

3.3 Interference Sharing and Soft Capacity

Another effect of W-CDMA is called interference sharing.

If a W-CDMA network is planned for equal loading of the cells, the interferencecontributed by each cell to the average system interference will be roughly the same.Based on this, each cell will be planned to expect a certain level of interference fromits neighbouring cells. Hence its loading / capacity will be planned with an upper limitwhich ensures that wanted signals can be despread and raised (by the processinggain) sufficiently above the interference (e.g. 7dB).

However, if neighbouring cells are lightly loaded, the adjacent cell interferencecontribution is lower. The cell in question can benefit by providing higher data ratesper user (lower processing gain, but with less total interference to raise the wantedsignal above), or to accept a higher interference contribution from users in its owncell. This effectively means more users. In both cases, the capacity in the cellincreases due to lower usage of (or interference from) adjacent cells.

Since different cells throughout the network may be benefiting from this effect at anyone time, there is an overall gain in capacity, and this can be planned for in thecapacity calculations.

The effect is known as soft capacity, and can only be taken advantage of if there isspare capacity (hardware and processing) available in the base sites. If there isn’tspare capacity, then hard blocking will occur, where calls or data transfer is blockeddue to lack of resources.



Equally Loaded Cells

Interference over whole system reflects dimensioning for average usage per cell

Cell Usage

Higher Capacity is possiblein middle cell due to reducedinterference from neighbouringcells

Interference over whole systemmay remain the same

Known as “Soft Capacity”

Fig. 8 – Interference Sharing


Understanding UMTS

3.4 Limitations

It is useful to note the limitations of the system in terms of coverage and capacity.

A significant factor is that in the downlink, the maximum transmitted power remainsthe same regardless of the number of users, and each user has to share the poweravailable, whilst in the uplink, each mobile has its own power amplifier.

Even with low downlink load, the coverage will depend on the interferencecontributed by the total number of users in the uplink (more users, less coverage),whilst in the downlink, for a given power, the higher the data rate, the lower the rangefor acceptable service.

Range and capacity are, of course, traded off against each other in both the uplinkand downlink. However, it is the plotted graphs of maximum path loss (rangeamongst other considerations) against load for given conditions, includinginterference, which illustrate the limitations. Below about 600kbps in the downlink, theuplink limits the range and hence coverage, whilst in the downlink, anything aboveabout 700kbps can only be provided at low range (low coverage area).

It is worth noting that a 2Mbps service can be provided using three separate codesand combining the results.

As the capacity limits are approached, it becomes much more difficult to increase thecapacity within the coverage area without adding more cells. Increasing downlinkpower to increase capacity is inefficient, whereas splitting the power between two W-CDMA carrier frequencies would be much more efficient, but requires additionalhardware.

The figures quoted are only for illustration, and many factors will affect theperformance, including improved antenna design for increasing the coverage (e.g.receive antenna diversity), and asymmetric services.



Coverage limited in the uplink

Capacity limited in the downlink

Fig. 9 – Limitations


• Note that for a 2 Mbps service, three codes would be used

Understanding UMTS

3.5 User Data Rates and Number of Users

Since acceptable service depends on despreading the wanted signal, which in turndepends on the level of interference in the system, there is generally a trade-offbetween the number of users and their average data rates.

This is because an upper limit on total interference determines the capacity.

For lower data rate users, the processing gain is relatively high and each signal will beraised significantly above the interference for average planned cell load conditions.However, as the number of users increase, so the interference increases, and anupper limit on the number of users will be reached (this is maximised by effectivepower control).

For higher data rate users, the processing gain is not so high, which results in thesignal being raised less effectively above the interference. The maximum interferencelevel will be reached earlier, hence a lower number of users will be tolerated foracceptable higher data rate services to be provided.

The total user data capacity is affected by the control information, in that more lowerdata rate users will require more control information, leaving less user data. However,higher data rate users will have a less even distribution over the system, which makescapacity planning less efficient. Therefore on average, overall cell loading may belower if higher data rate users dominated and had been planned for. In practice, ofcourse, a mix of users would generally be expected. Asymmetric services may alsobe provided.



0

32

64

144

1612.2

10 20 30 40 50 60

Number of channels per cell

Note: Figures are forillustration only

Use

r ra

tes

(kbp

s)

Fig. 10 – User Data Rates and Number of Users


• Total capacity (Users x User Rates) is not constant due to dimensioning andalso the greater control overhead required for more lower rate users

Understanding UMTS

4. REDUCING INTERFERENCE/INCREASING CAPACITY

4.1 Sectorisation

One simple way to reduce interference is to use sectored cells, where a single basesite can support up to six sectors (or cells) at a time. The directional antennas meanthat antenna gain acts to enhance the signals from mobiles in the sector, and rejectthose outside the sector. At the same time, interference in the downlink is reduceddue to the confinement of power within the sector, which ensures a lower contributionto overall interference. In addition, the cells are effectively smaller (for the samenumber of sites) and therefore lower powers can be used.

Softer handover now becomes a factor in that mobiles can be in handover betweensectors of the same Node B, with the signal combined in the Rake receiver (anadvantage in terms of reduced power requirement). The Soft handover region, ratherthan the Softer handover region still provides the greater area.

Effectively, the reduced interference allows greater capacity for the same number ofbase sites, or reduced hardware costs for the same capacity if the system is plannedas a sectored system from initial rollout.



• Reduced Co-Channel Interference

• Increased Capacity• Reduced Hardware

Costs

Base Station Sites withAxis of sectorisation

Soft Handover Region

Softer Handover Region

Fig. 11 – Sectorisation


Understanding UMTS

4.2 Smart Antennas

The sectored cell benefits can be taken to the extreme with the use of smartantennas, where multiple beams can be generated, each serving a specified user. Thebeams are effectively very narrow sectors, with all the benefits in terms of reducedinterference. Antenna gain is generally higher than for the three or six sector sites.The beams are formed and steered by antenna arrays.

The broadcast channel is configured for normal cell coverage since it must beavailable to mobiles in idle mode.

Smart antennas may well be a feature in UMTS networks.



NarrowBeam

User

BroadcastChannel

Fig. 12 – Smart Antennas


• Multiple beams can be generated for multiple users• Co-Channel interference in Uplink and Downlink is reduced• Range increased due to higher antenna gain• Capacity increased due to reduced co-channel interference

Understanding UMTS

4.3 Multi-Carrier Cells

Multi-carrier cells offer an efficient way to increase capacity (limited generally by thedownlink).

Unlike power increases in a single frequency (which provide little improvement incapacity as the higher data rates, or user numbers are approached), a second carrierfrequency allows the data rates, or user numbers to remain in the part of the loadcurve considered efficient in terms of capacity provision. Each frequency can thenuse power control to increase capacity until each now approaches the limit again (as defined by the load curve).

In general, the frequencies used will be adjacent channels (adjacent W-CDMAfrequencies), and will therefore cause adjacent channel interference. However, thiscan be minimised by co-locating the transmitter / receivers and using the sameantennas.

This concept can be extended to different network operators who may be usingadjacent frequencies. They could both benefit from reduced adjacent channelinterference if they co-located their transmitter / receivers.

A mobile required to change frequencies for capacity, coverage, or quality reasonswould use the hard handover technique.



Hard Handovercan be employed

Frequency 2

Frequency 1

Fig. 13 – Multi-Carrier Cells


• Adjacent channel interference is minimised by co-locating and using sameantennas.

• Different operators using adjacent frequency bands would benefit from co-location to reduce adjacent channel interference

Understanding UMTS

4.4 Planning the Frequency Spectrum

Within the available frequency band, the actual carrier frequency and spacing can beset in steps of 200kHz. This can be used effectively to decrease the spacing betweenfrequencies within the operators band (where the increased adjacent channelinterference can be minimised by co-locating the transmitters / receivers, and usingthe same antennas). This allows increased separation (at the edge of the operator'sband) with the adjacent frequencies used by another operator, as shown opposite.



Operator 1 (10MHz) Operator 2 (15MHz)

4.6MHz

frequency

4.6MHz > 5MHz

Larger spacing(Minimises interference)

Fig. 14 – Setting the Carrier Spacing


• Carrier spacing can be set in steps of 200kHz within band to minimise adjacentchannel interference

• Operators can minimise adjacent channel interference from their own carriersby co-location (hence spacing can be reduced in order to maximise spacingbetween operators)

Understanding UMTS

4.5 The Layered Architecture

A layered architecture may be employed for coverage, capacity, or quality reasons,with the different layers operating on different frequencies (required in order tominimise the effects of interference in each cell).

At the micro / pico-cell level, TDD mode may be utilised. Its reduced range can beused to advantage in areas requiring higher data rates. The TDD cell power can becontrolled sufficiently to provide coverage only in selected areas (such as offices).This ensures that interference is kept to a minimum in other coverage areas where theTDD frequency is re-used. Hence, higher data rates / loading can be provided for "hotspots", while more general coverage is provided by FDD operation.

If more than one FDD frequency is available, an umbrella layer can also be used forwider area coverage. The loading of each layer needs careful consideration. Differentdata rates can be traded off against the number of users.

Hard handovers must be used between different modes and frequencies (and also,therefore, between layers).



FDD f2 FDD f2

FDD f1

TDD f3

Fig. 15 – The Layered Architecture


• Different Carrier Frequencies may be employed at different layers (hard handovers can be used)

• Lower range, higher bit rate services may be provided by TDD mode (eg: office environments)

Understanding UMTS

4.6 GSM Co – Planning

The way in which UMTS has been specified allows for an evolutionary strategy fromexisting 2G GSM networks. In fact in the first phase of UMTS, the core network isessentially an evolved GSM Phase 2+ network, incorporating circuit switched andpacket switch (GPRS) infrastructures.

It is a general assumption that GSM and UMTS will co-exist in a similar way that firstand second generation systems did. However for UMTS, the core network will initiallybe that of the existing GSM network, and it follows that operators will also want tomaximise use of the existing radio infrastructure.

Re-using existing GSM base transceiver sites will provide a huge cost saving tooperators. The feasibility of this is illustrated opposite in a comparison of effectiveranges for both GSM (900 and 1800MHz frequencies) and UMTS (at varying datarates).

GSM 900 clearly allows for greater range, whereas GSM 1800 and low to moderateW-CDMA data rate ranges are more comparable. Range is more limited for the W-CDMA at higher data rates.



WCDMA 384kbps

WCDMA 144kbps

WCDMA SPEECH

GSM 1800 SPEECH

GSM 900 SPEECH

Relative Coverage (Range)

Fig. 16 – GSM Co Planning – Range Comparisons


Understanding UMTS

4.7 Use Of Existing Sites

Existing GSM sites may therefore be used for UMTS, so long as the cell range /coverage area required for the planned UMTS system can be provided with theexisting base site spacing. If not, then additional sites will be needed.

Remote, or rural areas, where GSM has been planned simply to provide continuouscoverage (with a low capacity requirement), will generally provide more of a problemthan urban areas due to the greater cell sizes. GSM 900 cell sizes will usually begreater than GSM 1800 in these areas.

For urban areas, capacity is often the major planning issue. Here, the availablefrequencies must be re-used more often in order to provide the required resources.This is achieved by carefully controlling the transmitted power and reducing the cellsizes. The GSM cell sites will be closer together and therefore more likely to supportUMTS cells with the required coverage areas. In urban areas, GSM 900 and 1800 arelikely to have similar cell sizes, and therefore are more equally likely to be able toprovide UMTS coverage from existing sites.

The UMTS soft and softer handover regions (including under higher load conditions)must be taken into account when planning coverage from existing GSM cell sites.



Remote Areas (GSM Coverage Planned)

GSM Range

WCDMA Range

Urban Areas (GSM Capacity Planned)

Additional BaseSites Required for

Continuous Coverage

Possible to ProvideContinuous Coverageusing Existing Sites

Cell boundary less than GSMor WCDMA maximum range

Fig. 17 – Using Existing 2G Sites


Understanding UMTS

4.8 W-CDMA, EDGE and GSM Coverage Areas

A network may be planned with the different W-CDMA modes, EDGE (Enhanced Datarates for Global Evolution) and GSM providing different areas of coverage. Eachwould be able to support different data rates and Qualities of Service. The reasons forthis approach could include cost, speed of rollout, coverage, capacity, or overallquality.

It makes sense in many ways to leave basic (voice subscribers) on the GSM network,since a large initial migration to UMTS would increase interference, reducing theability to provide higher data rate services for potentially high revenue (usuallycorporate) users. The high value users will usually be concentrated in business areas.

As more users start using higher data rate services, a gradual migration would allowoperators to use GSM sites initially for UMTS coverage, increasing capacity as thenetwork grows.

The cut off between a likely GSM and UMTS subscriber is not clear cut. EDGE couldprovide an extra step, with the flexibility of higher data rate services withoutincreasing the UMTS load.

UMTS services could be sold at a premium, with higher capacity provided initially inurban areas (where GSM co-planning is most feasible). TDD can be used to providehigher data rates in the required "hotspots".



GSMContinuousCoverage

WCDMA FDDBusiness Coverage

WCDMA TDDHot Spots

(Offices etc)

EDGEURBAN

COVERAGE

Fig. 18 – Coverage Areas


Understanding UMTS

5. RADIO PLANNING AND SIMULATIONS

5.1 Planning Tools

The complex nature of the iterative process necessitates a software planning tool.This allows much of the external data to be loaded, including the radio propagationand geographical data for the area in question. The operator can then ask foriterations under various conditions. Various visual representations are used to coveythe coverage and capacity results on screen, as shown.

5.2 Monte Carlo Simulations

The Monte Carlo method is a general technique applied to many systems to modelthe outcomes, from economics to physics. The basis is random numbers andstatistical probability calculations. There are many forms of Monte Carlo simulations.In UMTS, the simulation is applied to the radio network to model the coverage andtherefore capacity of the radio system under various conditions.

The planning tools used at this stage of the planning process use the Monte Carlomethod because it allows complex systems to be analysed by sampling a number ofrandom configurations. It then uses this data to describe the system as a whole.



Fig. 19 – Planning Tools


Supplied courtesy of Ericsson

Understanding UMTS

6. SUMMARY – PLANNING CONSIDERATIONS

The planning considerations for UMTS are summarised in the figure shown opposite.



Fig. 20 – W-CDMA Planning Considerations - Summary


• Cellular structure used• Same frequency can be used in each cell• Codes need to be planned to prevent co-channel interference

(64 groups to choose from)• In general, the greater the bit rate per user, the lower the number

of users per cell• Greater cell range gives reduced capacity (and vice versa)

for a given power• The greater the instantaneous cell usage, the smaller the

effective cell range (cell breathing)• Interference decreases range and/or capacity for a given power• Overall system noise/interference must be kept as low as possible• Interference reducing techniques may be used

– Diversity– Multi user detection– Smart antennas– Repeaters

• Increased downlink power gives increased capacity only to a certain limit• Planning may use:

– more than one frequency for capacity or coverage purposes (eg: hierarchical cells)

– GSM/EDGE to offer more complete coverage around UMTS “Islands”• TDD mode suited to shorter range (and possibly higher data rates)• FDD mode suited to longer range• Asymmetric services must be planned for

UMTS Planning

Documents

transmitter

radio planning

reducing interferenceincreasing

softer handover

spread wanted

higher data

careful power

provide continuous