Frequency Hopping

8/1551-FCU 101 201 Uen Rev B 1996-11-01 1(22)

USER DESCRIPTION,FREQUENCY HOPPING

Contents

1 Introduction ..................................................................... 3

2 Background ...................................................................... 3

2.1 General ........................................................................................... 3

2.2 Multipath fading................................................................................ 3

2.3 Co-channel interference .................................................................... 4

3 What can be achieved ...................................................... 4

3.1 Frequency diversity........................................................................... 4

3.2 Interference averaging ...................................................................... 5

3.3 Conclusions...................................................................................... 6

4 Short technical description ............................................. 7

4.1 General ........................................................................................... 7

4.2 Baseband hopping............................................................................ 7

4.3 Synthesizer hopping.......................................................................... 8

4.4 Configuration.................................................................................... 8

4.5 Algorithm ....................................................................................... 12

4.6 Main differences from CME 20 R5................................................... 14

4.7 Additions in CME 20 R6.1 ............................................................... 14

5 Engineering guidelines.................................................. 14

5.1 Applications ................................................................................... 14

5.2 Impact of FH on frequency planning ................................................. 16

5.3 Channel configuration...................................................................... 18

5.4 MS aspects of frequency hopping.................................................... 20

5.5 Frequency hopping and subjective speech quality.............................. 20

6 Parameters ..................................................................... 21

6.1 Main controlling parameters ............................................................ 21

USER DESCRIPTIONS AND ENGINEERING GUIDELINES, CME 20 R6

2(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

6.2 Value ranges and default values....................................................... 22

7 References ......................................................................22

USER DESCRIPTION, FREQUENCY HOPPING

8/1551-FCU 101 201 Uen Rev B 1996-11-01 3(22)

1 IntroductionFrequency hopping means that multiple frequencies are used for the transmis-sion of speech or data in a single connection. Each burst is transmitted on afixed frequency, but the frequency is changed between bursts. A burst caneasily be lost when the mobile happens to be located in a fading dip for thatparticular frequency, or if it is subjected to interference. The next burst, if on adifferent frequency, has a good chance to come through. The coding andinterleaving scheme in GSM is constructed so that loss of a single burst shallhave minimal influence on the speech quality.

A predefined set of frequencies is used in each cell. The mobile and the basestation change frequency between every burst, i.e. 217 times per second. Thesame sequence is used in the uplink and in the downlink.

2 Background

2.1 GeneralDuring a call, a number of physical effects influence the perceived radio en-vironment between a mobile station and a base station. One such effect ismultipath fading, which means that transmitted signals reach the receiver viamultiple paths. Depending on the difference in path length, constructive or dest-ructive interference can be obtained, resulting in variations in the receivedsignal strength, i.e. fading.

Another effect is various types of interference. The dominating type isnormally co-channel interference, but other types, such as adjacent channelinterference, intermodulation products, military sources etc. must beconsidered as well. The interference varies as the interferers switch on and off.As a worst case, however, the interference can remain during the entire call.

The time intervals of low received signal strength due to multipath fading areoften fairly long for a slowly moving mobile, if a fixed frequency is used. Theduration of persistent co-channel interference is also usually fairly long. Forboth of these cases, the coding and interleaving will not be adequate to protectthe signal information.

These considerations are similar in the uplink and in the downlink.

2.2 Multipath fadingThe destructive interference produced by multipath fading is called “fadingdips”. Fading dips may cause speech quality degradation. For a specificfrequency, multipath fading dips in the radio signal are normally separated inspace approximately by one half wavelength (i.e. 17 cm for GSM 900, and8 cm for DCS 1800). If the mobile antenna happens to be placed in such a dip


4(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

(e.g. when a car stops at red light, the so called red light problem), the speechquality can become unacceptable. For different frequencies, the fading dipswill occur at slightly different positions in space.

2.3 Co-channel interferenceThe interference situation for a mobile is strongly dependent on which frequ-ency and time-slot that the mobile happens to use. An example of one of themost difficult situations, interference-wise, is shown in figure 1. The exampleshows downlink interference, but the situation in the uplink is similar.

A

B

Figure 1 Downlink co-channel interference

Two mobiles, which transmit on the same frequency and time-slot, are situatedin the vicinity of their own cell border. Consider the white mobile. The signalsreceived by the white mobile are relatively weak, making it sensitive tointerference. At the same time, the interfering base station B transmits atmaximum power to the black mobile producing a high and constant co-channelinterference to the white mobile. The same considerations apply to the blackmobile.

The cell- and frequency planning has to take this situation into consideration,so that the difficult situations, such as that shown in figure 1, can be providedwith sufficient service quality.

3 What can be achieved

3.1 Frequency diversityFrequency hopping can reduce the influence of signal strength variationscaused by multipath fading. In this document we will refer to this effect asfrequency diversity.


8/1551-FCU 101 201 Uen Rev B 1996-11-01 5(22)

Multipath fading is frequency dependent. This implies that the fading dipsappear at different locations for different frequencies. Thus a mobile utilisingfrequency hopping will not remain in a specific fading dip for a longer timethan one single burst. Thereby signal strength variations are broken up intopieces of a duration short enough for the interleaving and speech codingprocess to correct for errors. Multipath fading dips, causing low signal strength,are thus apparently levelled out, and slowly moving mobiles (and cars stuck ata red light) will perceive a more even radio environment. Frequency hoppingmakes most of the fading dips appear more shallow, see figure 2.

distance

received signal strength

Figure 2 Schematic picture of multipath fading at two different frequenciesand at frequency hopping between the two frequencies for aslowly moving mobile

The thin solid line and the thin dotted line in figure 2 show the received signalstrength obtained at two frequencies. The thick solid line shows schematicallythe smoothing effect when the two frequencies are combined by frequencyhopping. The signal after decoding will in effect be an average between thesignals at the two frequencies.

Fast moving mobiles obtain the same type of improvement as frequencydiversity by their speed alone, since the radio environment can easily changebetween bursts. Frequency diversity thus gives only a minor extra improvementfor fast moving mobiles.

Cyclic frequency hopping (described in section 4.5.1) gives slightly betterfrequency diversity performance than random frequency hopping (section4.5.2).

3.2 Interference averagingFrequency hopping can also break up persistent interference into periodicoccasions of single burst interference. The cell planning margin againstsituations of bad radio conditions can thus be decreased, since the probabilityof encountering these conditions decreases.


6(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

Changing frequency at each burst offers a way to improve the interferencesituation described above. The co-channel interference will change at everyburst, which is beneficial for the connection that otherwise might suffer from asevere interference during the entire connection. Likewise, the interference thatone connection is causing another, is spread out to a number of connections insingle bursts. This effect is called interference averaging.

Occasionally there will be frequency collisions causing strong interference, butwith a very short duration (one burst). Again, the coding and interleaving willget a chance to deal with the situation. The more frequencies that are used inthe hopping, the more rare such frequency collisions will be.

The same considerations apply to any type of interference, co-channel, adjacentchannel, intermodulation products etc.

The radio environment, in terms of interference, will be more even, and the cellplanning margin can be decreased.

The performance of the interference averaging will be dependent of the modeof hopping, cyclic or random. The greatest improvement will be obtained whenthe interferer and the interfered connections use hopping sequences that areindependent of each other, i.e. no correlation between them (see section 4.5.4).The lower the correlation, the higher the averaging gain. If the interferer and theinterfered connection both use cyclic hopping and, in addition, the samefrequencies, they may get in phase with one another. The two connections willthen hop between the frequencies “hand in hand”, and the co-channelinterference will persist as if there was no frequency hopping at all. The effectis total correlation, and the resulting improvement will be very small.

In a fully loaded system, all possible interferers will actually be transmitting,and consequently interfering. Even in this case random frequency hopping willshow a performance gain. However, a requirement is that co-channel cells usedifferent (independent) hopping sequences, see section 4.5.4.

3.3 ConclusionsFrom a subscriber point of view, frequency hopping will give an improvedspeech quality. From an operator point of view, the benefits are:

• a more dependable and predictable radio environment,

• a possibility to give the subscribers a generally uniform speech quality,

• a possibility to decrease the cell planning margin, which might be used toemploy a tighter frequency re-use yielding a capacity increase.

The effects increase with the number of frequencies used for hopping, but therelative benefit by adding yet another frequency will diminish.


8/1551-FCU 101 201 Uen Rev B 1996-11-01 7(22)

4 Short technical description

4.1 GeneralFrequency hopping in the base stations can be implemented in either of twoways, baseband hopping or synthesizer hopping, each with its pros and cons.The parameter FHOP specifies which method to be used in the entire basestation. The feature Frequency hopping as such is switched on and offseparately for each channel group with the parameter HOP, see section 4.4.1.

The BCCH channel can never hop, since it is a broadcast channel. The TCHsand SDCCHs, however, can hop.

In the following, baseband and synthesizer hopping is explained with the use ofsome simple examples of combiner and antenna configurations. The actualconfiguration however, depends on the base station (RBS200 or RBS2000) andon the type of CDU unit equipped.

4.2 Baseband hoppingAt baseband hopping each transmitter operates on a fixed frequency. Attransmission, all bursts, irrespective of to which connection they belong, arerouted from the TRX to the transmitter of the proper frequency, see figure 3.

TRX1

TRX2

TRX3

TRX4

transmitterf0 ... fntransmitterf0

transmitterf3

transmitterf2

transmitterf1

controller

controller

controller

controller

bus for routing of bursts

filtercombiner

Figure 3 Routing of bursts from the TRX to the transmitter at basebandhopping

The advantage with this mode is that narrow-band tuneable filter combiners canbe used. These combiners have up to 16 inputs. This makes it possible to usemany TRX:s without having to connect several combiners in cascade.

The disadvantage is that it is not possible to use a larger number of frequenciesthan there are transmitters.


8(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

4.3 Synthesizer hoppingSynthesizer hopping means that one transmitter handles all bursts that belong toa specific connection. The bursts are sent “straight on forward” and not routedby the bus, in contrast to baseband hopping. The transmitter tunes to the correctfrequency at transmission of each burst, see figure 4.

TRX1

TRX2

TRX3

TRX4

transmitterf0 ... fn

hybridcombiner





hybridcombiner

hybridcombiner

controller

controller

controller

controller

Figure 4 Sending bursts from the TRX to the transmitter at synthesizerhopping

The advantage is that the number of frequencies that can be used for hopping isnot dependent on the number of transmitters.

The disadvantage is that wide-band hybrid combiners have to be used. Thistype of combiner has approximately 3 dB insertion loss, making more than twocombiners in cascade impractical.

4.4 Configuration

4.4.1 General

At cell configuration, the frequencies in a cell are assigned to one or severalchannel groups. Each channel group can be defined separately as hopping withthe parameter HOP. For example, there can be two channel groups in a cell,where one is hopping and the other is not. Within each channel group, thechannels will hop over the frequencies defined for that particular channelgroup. SDCCHs as well as TCHs can hop. A BCCH will not hop, even if itbelongs to a channel group that is configured as hopping.

The BCCH frequency must always be transmitting. If there are no traffic burststo be transmitted, there still must be transmission of radio energy at the BCCHfrequency. This is provided by the transmitter itself. If it is configured for asingle frequency, it can be set up so that it transmits dummy bursts whenevernothing else arrives from the controllers via the bus. This is called carrier-zero(c0) filling when applied to the BCCH frequency f0. The c0 filling is obtainedautomatically for the channel group containing the BCCH frequency.


8/1551-FCU 101 201 Uen Rev B 1996-11-01 9(22)

For baseband hopping, the c0 filling is straightforward (see section 4.4.2). Forsynthesizer hopping, the c0 filling is more complicated. There are twoconfiguration options, one with the BCCH frequency included in the hoppingset (section 4.4.3), and the other with the BCCH frequency in a separate nonhopping channel group (section 4.4.4).

4.4.2 c0 filling at baseband hopping

As an example, suppose that a base station containing four TRXs shall be con-figured for baseband hopping. In this case, only four frequencies can bedefined. Assume furthermore that only one channel group is defined. One ofthe transmitters is transmitting on f0 only (see figure 3). This transmitter isautomatically set up for c0 filling. The resulting configured channels are shownschematically in figure 5. Hopping channels are circled.

TS0

TS1

TS2

TS3

TS4

TS5

TS6

TS7

BCCHf0f1f2f3

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

TCHTCH

TCH TCHTCHTCH

TCHTCHTCHTCH

TCHTCHTCHTCH

TCHTCHTCHTCH

TCHTCHTCHTCH

TCHTCHTCHTCH

TCH

TCHTCH

SDCCHTCH

Figure 5 Channel configuration for 4 TRXs and baseband frequencyhopping

Figure 5 indicates that there are four frequencies used. On the first time slot,TS0, the BCCH is not hopping, but transmitted on f0 only. The remaining threechannels on TS0, which are all traffic channels (TCH), hop on three frequencies,f1 to f3, indicated by the circle around the TCHs in the positions for f1, f2 andf3. On TS1 all channels, one SDCCH/8 and three TCHs, are hopping on all fourfrequencies, f0 to f3, as indicated by the circle. On the remaining time slots,there are four TCHs hopping on all four frequencies.

4.4.3 c0 filling at synthesizer hopping: the BCCH frequencyincluded

A transmitter configured for synthesizer hopping cannot perform filling since itis not configured for a fixed frequency. Furthermore, if the number offrequencies that are configured for a channel group is larger than the number ofTRXs, there is no possibility to guarantee that f0 is always transmitted. Allchannels may be busy and transmitting on a frequency different from f0 . Thisis a consequence of the fact that the GSM specifications defines a number offixed frequency sequences (see section 4.5.1).

A method has been provided so that c0 filling can be obtained at the same timeas f0 is included in a set of hopping frequencies. With this method, an extratransmitter operating only as a c0 filler is used. All traffic bursts that are to be


10(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

transmitted on the f0 frequency are routed to this transmitter. If no traffic burstsare transmitted, dummy bursts are sent instead, as usual for a c0 filler. Allother traffic bursts are sent straight on forward, as usual for synthesizerhopping. The method is thus a mixture of synthesizer hopping and basebandhopping.

There are two ways to configure for c0 filling. Figure 6 shows one of themethods when three TRXs are used.

TRX1

TRX2

TRX3


hybridcombiner

transmitterf0




hybridcombiner

hybridcombiner

controller

controller

controller

f0

f0

f0f0

Figure 6 Synthesizer hopping, plus baseband hopping for bursts of theBCCH frequency, 3 TRXs

The resulting configured channels are shown schematically in figure 7, for thecase HOP=ON.

TSTS0

TS1

TS2

TS3 TS4

TS5

TS6 7

TCHTCH

BCCH

TCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

SDCCH

fn

...

f0f1f2f3

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

Figure 7 Channel configuration for 4 transmitters, 3 TRXs and frequencyhopping with the BCCH frequency included

The figure indicates that on TS1 to TS7, all bursts that shall be transmitted onthe f0 frequency are sent to the c0 filler transmitter instead of to their regulartransmitters. On TS1 to TS7, all frequencies defined (f0 to fn) are used in thehopping sequence. On TS0, the frequencies f1 to fn are used. Note that thenumber of frequencies used in the set of hopping frequencies is not limited tothe number of TRXs (three in figures 6 and 7). The total number of trafficchannels will be 22 according to figure 7.


8/1551-FCU 101 201 Uen Rev B 1996-11-01 11(22)

In the other configuration method an entire extra TRX is added (or one of theexisting devices is used). The controller of that TRX is utilized for the BCCHchannel only, i.e. it carries no traffic. Figure 8 shows this.

TRX4

TRX1

TRX2

TRX3


hybridcombiner

transmitterf0




hybridcombiner

hybridcombiner

controller

controller

controller

f0

f0

f0f0

BCCHcontroller

Figure 8 Synthesizer hopping plus baseband hopping for bursts of theBCCH frequency, four TRXs

Thus, for a base station containing four TRXs, this option leaves three TRXsfor traffic, and the fourth for BCCH and filling. The resulting configuredchannels are schematically shown in figure 9, for the case HOP=ON.

TSTS0

TS1

TS2

TS3 TS4

TS5

TS6 7

TCHTCH

TCH

BCCH

TCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

SDCCH

fn

...

f0f1f2f3

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

Figure 9 Channel configuration for 4 TRXs and frequency hopping with theBCCH frequency included

Note that in this case one additional TCH channel is obtained on time slot TS0,as compared to figure 7.

If a cell is configured for synthesizer hopping, and if f0 is included in the set offrequencies defined for hopping, one of these two channel configurations isobtained automatically, depending on the hardware.

4.4.4 c0 filling at synthesizer hopping: two channel groups

The configurations described above can be seen as a waste of hardwareresources. If the resource utilization is of greater importance than the


12(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

possibility to use f0 in the hopping sequences, the cell can be configured withtwo channel groups. One channel group (channel group 0) must be defined tocontain f0 only, and HOP set to OFF. The other channel group is defined tocontain the rest of the frequencies f1 to fn, see figure 10, for the caseHOP=ON in that channel group.

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

fillingc0

TSTS0

TS1

TS2

TS3 TS4

TS5

TS6 7

fn

...

f0f1f2f3 TCH

TCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

TCHTCHTCH

SDCCH

BCCH TCH TCH TCH TCHTCH TCH TCH

Figure 10 Channel configuration for 4 TRXs and pure synthesizer frequencyhopping in one of two channel groups

In this case, one frequency less (f1 to fn) can be used for hopping, but the totalnumber of traffic channels will increase to 30. Note that 7 of these are nonhopping and 23 are hopping.

This configuration is obtained automatically if two channel groups are definedfor a cell. If two channel groups are defined, each with two or morefrequencies, the one containing f0 will be configured according to section 4.4.3.

4.5 Algorithm

4.5.1 Cyclic frequency hopping

In cyclic hopping the frequencies are used in a consecutive order. For instance,the sequence of frequencies for cyclic hopping between four frequencies mayappear as follows:

... , f4, f1, f2, f3, f4, f1, f2, f3, f4, f1, f2, ...

A cyclic sequence is specified by setting the parameter HSN (hoppingsequence number) to zero. There is only one cyclic sequence defined in theGSM specifications. The sequence of frequencies goes from the lowestabsolute frequency number in the set of frequencies specified for that channelgroup, to the highest, and over again.

4.5.2 Random frequency hopping

A random hopping sequence is actually implemented as a pseudo-randomsequence. The sequence is stored in a look-up table, in the mobiles as well as


8/1551-FCU 101 201 Uen Rev B 1996-11-01 13(22)

in the base stations. 63 independent sequences are defined (see section 4.5.4).Which of the 63 sequences to be used is specified with HSN. The algorithmfor selecting the frequency for each burst allows hopping on up to 64frequencies. The actual frequency to be used at each instant is obtained by amodulo operation with the available frequencies.

When random hopping is used, the frequencies will be used (pseudo-)randomly, and a hopping sequence for four freuencies may appear as follows:

... , f1, f4, f4, f3, f1, f2, f4, f1, f3, f3, f2, ...

The period for a random sequence is 6 minutes.

4.5.3 Orthogonal sequences

In the baseband hopping example above, figure 5, four channels utilize thesame time slot. They will be given the same HSN. In order not to interfere witheach other, they may not use the same frequency simultaneously. A frequencyoffset1 is automatically assigned to each channel at configuration. In this way,each traffic channel uses the same sequence, but with different frequencies ateach instance in time. The random sequence in the example above will appearas follows for four frequencies:

... , f1, f4, f4, f3, f1, f2, f4, f1, f3, f3, f2, ...

... , f2, f1, f1, f4, f2, f3, f1, f2, f4, f4, f3, ...

... , f3, f2, f2, f1, f3, f4, f2, f3, f1, f1, f4, ...

... , f4, f3, f3, f2, f4, f1, f3, f4, f2, f2, f1, ...

The hopping channels thus never use the same frequency simultaneously. Thisis called orthogonality. All channels in one and the same cell on one and thesame timeslot must always be orthogonal, since non-orthogonal channels willcause co-channel interference to each other.

4.5.4 Independent sequences

For the interference averaging mechanism to work well, the sequence of freq-uencies in co-channel cells must be different. This is even more important ifthe cells use exactly the same frequency set. Connections in these cells willthen use the same frequencies, but not always at the same time. If thefrequencies are independent, they will only interfere with each other now andthen, when two (or more) bursts happen to coincide in frequency. The numberof collisions per second will depend on the number of frequencies in thechannel group.

1Mobile Allocation Index Offset (MAIO), see GSM specification 05.02


14(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

The sequences for three connections in three co-channel cells might appear asfollows for four frequencies in the hopping set. The frequency collisions, i.e.the instances of co-channel disturbance, are indicated with bold type:

Cell 1: ... , f1, f4, f4, f3, f1, f2, f3, f1, f3, f4, f2, ...



Sequences of this type are called independent since the correlation betweenfrequencies is minimal. Since there is only one cyclic sequence, cyclicsequences can be orthogonal (if they have different MAIO), but neverindependent. This is the reason why cyclic frequency hopping can not give fullinterference averaging. In order to obtain efficient interference averaging, co-channel cells must be assigned different HSN, in much the same way as BSICis planned.

4.6 Main differences from CME 20 R5No differences from CME 20 R5.

4.7 Additions in CME 20 R6.1No additions.

5 Engineering guidelines

5.1 Applications

5.1.1 General

Frequency hopping should always be enabled in every cell. It is the funda-mental feature used to support high capacity networks while maintainingquality. It also increases the gains from DTX and power control.

Frequency hopping gives a better performance by increasing the speech qualityfor slow or non-moving mobiles (frequency diversity, C/N gain) and byincreasing the traffic capacity as the re-use distance between co-channel cellscan be decreased (interference averaging, C/I gain). See also sections 3.1 and3.2.

High capacity networks is the most important application for the frequencyhopping feature.


8/1551-FCU 101 201 Uen Rev B 1996-11-01 15(22)

5.1.2 Number of TRXs per cell.

The choice of hopping mode depends on the environment in which frequencyhopping is used, the traffic intensity, and the number of frequencies. The morefrequencies configured for hopping the greater the improvement in C/I.Regarding the number of frequencies in the hopping group, there is a “law ofdiminishing returns”. For example: increasing the number of frequencies in thehopping group from 7 to 8 does not give as much improvement as increasing itfrom 2 to 3.

Three transceivers per cell results in a substantial interference averaging gain.A four transceiver per cell system will work even better. A three or a fourtransceiver system with frequency hopping can support an average reusesignificantly tighter than 12. A good reuse target is 8 for 3 TRX/cell and 7 for 4TRX/cell or more. The interference averaging gain for a two transceiver systemis zero, no gain at all. Frequency hopping could be enabled in those cellsanyway to obtain the frequency diversity gain for slowly moving mobiles.

Frequency diversity for slowly moving mobiles depends on the coherencebandwidth of the radio link. Even a two transceiver cell will give frequencydiversity if the carrier frequencies are separated by more than the coherencebandwidth. In a dense urban or urban environment with many scatterers therewill be substantial gain if the spread of the carriers is 1 Mhz or more. Thecoherence bandwidth determining the frequency diversity gain is verydependent upon environment. More scatterers are better in this case. A line ofsight connection will have less gain, but probably higher signal strength fromthe beginning which means that a gain is really not needed.

Table 1 summarizes the frequency hopping gains.

Table 1 Gain from frequency hopping

TRXs percell

Interference averaging(C/I gain)

Frequency diversity(C/N gain)

2 not measurable Substantial if carrierfrequencies separated by

coherence bandwidth: aroundone MHz, dense urban and

Urban scenarios.

3 substantial, enough for highcapacity networks.

Same or better than2 TRXs per cell.

≥ 4 most of the gain collected at4 TRXs per cell.

Same or better than2 TRXs per cell.

5.1.3 Cyclic or random hopping

The random hopping mode gives better interference averaging and is superiorfor reducing the co-channel interference. Random hopping is the hopping modeof choice for high capacity networks. If random hopping is applied, co-channelcells (cells which have the same carrier frequencies in the hopping sequence)should to the greatest possible extent use different HSN.


16(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

If a system is planned without using a regular frequency re-use pattern, cyclichopping will give the same interference averaging effect as random hopping.

5.1.4 Baseband or synthesizer hopping

The choice between baseband hopping and synthesizer hopping depends on theavailable hardware. Some properties of the synthesizer and baseband hoppingare summarized below:

• Number of TXs

In a synthesizer hopping system, the required number of TXs can be lessthan the number of frequencies. In a baseband hopping system, a dedicatedTX is needed for each frequency.

• Combiner type

For the synthesizer hopping system, a hybrid combiner is required.Baseband hopping can also be specified if hybrid combiners are used. If afilter combiner is used then only baseband hopping can be specified.

Hybrid combiners have a greater power loss than filter combiners. Theminimum channel separation in a hybrid combiner is 400 kHz, while theminimum channel separation in a filter combiner is 600 kHz (900 MHzband) and 1200 kHz (1800 MHz band).

• Faulty TX

In a synthesizer hopping system, a fault on a TX will affect up to eightBPCs assigned to that TX. In a baseband hopping system, a fault on a TXwill affect every BPC that uses the frequency that the faulty TX wastransmitting on.

A lost transceiver, included in a frequency hopping sequence, is isolatedand hopping is restored without the faulty transceiver, but the transceiverwill not automatically be included in the hopping sequence after a recovery.

5.2 Impact of FH on frequency planning

5.2.1 General

Frequency planning in a system without frequency hopping is done with respectto an acceptable interference level. However, this level in a system withfrequency hopping can be set to a lower value.

5.2.2 BCCH-carrier planning

The BCCH carries information for identifying cells for access evaluation,paging and measurements on neighbour cells for locating evaluations. Becauseof its importance, great care should be taken to protect the BCCH frominterference. Furthermore, in a system with frequency hopping, the TCHs andmaybe the SDCCHs are hopping but the BCCH channel (timeslot 0 on the


8/1551-FCU 101 201 Uen Rev B 1996-11-01 17(22)

BCCH carrier) is not hopping. Hence, the BCCH carrier can not be plannedaccording to a tighter re-use pattern. A traditional 12 reuse is recommended forthe BCCH-transceiver.

5.2.3 High capacity networks

The TCH-transceivers can be planned using a tighter reuse than 12, since theTCH is a very robust channel. A high capacity network of macro and microcells can be based on baseband random hopping in the macro network andrandom synthesised hopping in the microcell network. The baseband hopping iscompatible with existing networks using filter combiners.

The general procedure is to use the macro network as the first option toincrease capacity in the intermediate phases. Deploying microcells in the laterphases enables the operator to perform site hunting and establish proceduresfor the microcell roll-out.

The macrocells use the Multiple Reuse Pattern (MRP) tecnique in order toachieve a tight reuse. A 12 reuse for the BCCH carriers allocated in a separateband is combined with tight TCH-carrier reuse in order to acheive an averagereuse down to 7.

Hot spot micro cells can be introduced by borrowing frequencies from themacro cells. To make the frequency planning and implementation of thesemicro cells easier, a few frequencies can be made available by a furthertightened reuse on the TCH carriers.

An even tighter TCH reuse in the macro network can be applied by using theEricsson MRP technique to make micro cell carriers available as the micro cellnetwork grows. An allocation of 5 to 8 carriers is normally sufficient for acontiguous micro cell network with 2 transceivers per micro cell. This is madepossible by employing the synthesiser frequency hopping capabilities ofRBS2301 in the micro cells. The synthesiser hopping mode according to 4.4.4could be used.

Networks using wideband combiners, such as RBS2000 CDU-A and CDU-Ccan obtain the frequency hopping gain even when configured with only twotransceivers per cell by utilising synthesiser hopping.

5.3 Channel configuration

5.3.1 Baseband hopping

The following example illustrates the channel structure for a cell with basebandfrequency hopping.


18(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

Example:

Consider a cell without subcell structure and with one channel group.Four frequencies are assigned to the cell in the 900 MHz band, definedby ARFCN 75, 78, 81, and 84.

The following parameter setting is defined for the cell:

Table 2 Parameter setting, example

Parameter Value Description

BCCHNO 75 The carrier defined by ARFCN 75 isdefined as the BCCH carrier(f0 in this example).

CHGR 0 Channel group 0 is defined in the cell.

HOP ON Channel group 0 is hopping.

HSN 0 Cyclic hopping mode is specified (see nextsection).

DCHNO 78&81&84 Three frequencies are assigned to CHGR0,defined by ARFCN 78, 81, and 84(f1, f2 and f3 in this example).

In this example the following correspondence between ARFCNs andfrequencies is assumed using baseband hopping:

ARFCN = 75 f0

ARFCN = 78 f1

ARFCN = 81 f2

ARFCN = 84 f3

For the previous parameter setting, the system automatically defines threehopping frequency sets (HFS):

HFS1 {f0},HFS2 {f1, f2, f3},HFS3 {f0, f1, f2, f3}.

HFS 1 contains frequency f0, which is the BCCH carrier and is non-hopping. HFS 1 is assigned to TS0 on the BCCH carrier.

HFS 2 contains all frequencies except the BCCH carrier and is assignedto TS0 on the non-BCCH carriers.

HFS3 contains all frequencies defined in the channel group and isassigned to TS1, TS2, ..., TS7 on all frequencies defined in the cell.

The BPC in TS0 on the BCCH frequency is non–hopping (HFS1). Theother BPCs in TS0 on the other frequencies are hopping (HFS2). Allother BPCs in TS1 to TS7 on all frequencies are hopping, (HFS3). Thisleads to 31 hopping BPCs.


8/1551-FCU 101 201 Uen Rev B 1996-11-01 19(22)

Figure 11, which is based on figure 5, shows this configuration.

TS7TS6TS5TS4TS3TS2TS1TS0

f0

78,81,84

75

f3

f2

f175, 78, 81, 84

Figure 11 Channel configuration for 4 TRXs and baseband frequencyhopping, example

5.3.2 Configuring synthesizer hopping

If the number of frequencies is greater than the number of TRXs, TSs 1 to 7 onthe TX that supports the BCCH frequency cannot be utilized. Thus, the totalnumber of supported BPCs is reduced by seven. This creates a limitation onconfiguring synthesizer hopping.

The following two methods can be used to eliminate this limitation:

1. Define the available frequencies in different channel groups.

Define two channel groups, one channel group which contains only theBCCH frequency and another channel group that contains the otherfrequencies. This configuration is described by figure 10.

Consider a configuration with two TRXs and four frequencies (f0, f1, f2, andf3). Define two channel groups: CHGR0 and CHGR1. CHGR0 containsonly the BCCH frequency (f0 in this example). CHGR1 contains the otherfrequencies (f1, f2 and f3). With this configuration, one TX supports CHGR0with one BCCH and seven non-hopping TCHs. The other TX supportsCHGR1, with 8 BPCs hopping on three frequencies.

2. Add extra hardware

An existing TX can be added to the configuration. BSC supportsconfigurations of the system with an extra TX. This TX supports the BCCHfrequency. These configurations are described by figures 6 to 9.

5.4 MS aspects of frequency hoppingMS aspects should not prevent the operators to take advantage of frequencyhopping.

5.5 Frequency hopping and subjective speech qualityThe mapping of the raw bit error rate estimated in the rxqual measure has avery non-linear relationship to subjective speech quality. Furthermore, this


20(22) 8/1551-FCU 101 201 Uen Rev B 1996-11-01

relationship depends on the mode of operation in the network when it comes tofrequency hopping and DTX.

A field survey team using a test mobile, e.g. TEMS, or a radio networkengineer using Mobile Traffic Recording (MTR) can collect rxqual in order toasses the speech quality of the system.

The average rxqual limit for poor subjective speech quality is:

• rxqual ≤ 4.5 for a system without frequency hopping

• rxqual ≤ 5.5 on average for a system with frequency hopping.

These limits will approximately correspond to speech Frame Erasure Rate,(FER) of 10%.

6 Parameters

6.1 Main controlling parametersHOP is the switch for turning frequency hopping on or off, defined per channelgroup. HOP defines whether all channels except the BCCH hop (HOP =ON),all traffic channels hop (i.e. all channels except the BCCH and the SDCCHs;HOP =TCH), or no channels at all hop (HOP =OFF).

HSN is the hopping sequence number, defined per channel group. Thisparameter specifies which hopping sequence to be used. All timeslots in onechannel group are configured with the same HSN. HSN=0 yields a cyclicsequence. HSN=1 to 63 yields pseudo-random sequences. Due to theprocedure used by the mobile for measurement reporting, the use of cyclichopping with a multiple of 13 frequencies should be avoided when DTX isused, see ref. 1.

FHOP selects which hopping method to be used, baseband hopping(FHOP=BB) or synthesizer hopping (FHOP=SY). It is defined per transceivergroup. At synthesizer hopping, the hopping method including c0 mentioned insection 4.4.3, is obtained by default in the channel group that contains theBCCH carrier.

COMB specifies which combiner type that has been connected, a wide-bandhybrid combiner (COMB=HYB) or a narrow-band filter combiner(COMB=FLT). It is defined per transceiver group. If a filter combiner isconnected, only baseband hopping can be used as hopping method.


8/1551-FCU 101 201 Uen Rev B 1996-11-01 21(22)

6.2 Value ranges and default values

Parametername

Defaultvalue

Recommen-ded value

Valuerange

Unit

HOP OFF ON OFF, TCH, ON

HSN – 0 to 63

FHOP – BB, SY

COMB – HYB, FLT

7 References1. User Description, DTX, 7/1551–FCU 101 201 Uen

Frequency Hopping

Documents

frequency planning

particular frequency

fixed frequency

different frequency

frequency diversity

channel interference

user descriptions

uen rev b