9 Radio link management

11.213GPPBase Station System (BSS) equipment specificationRadio aspectsRelease 1999TS

9.1 General

This clause describes the functions of the BSS which gain, maintain and release access to the radio link, the main objective being to provide a stable link for the higher protocol layers whilst hiding, as far as possible, the properties of the radiopath.

The detailed operation of these functions can be found in GSM 05.08 (EN 300 911) [23] and GSM 05.10 (EN 300 912) [24]. Some requirements are also found in GSM 05.02 (ETS 300 908) [19].

9.2 Synchronization

The BSS provides control information to the MS so that its transmissions arrive at the BSS within the allocated timeslot window and within the correct frequency tolerance.

The following requirements related to synchronization are not tested in this subclause:

1) Absolute Frequency Tolerance.
This is tested in subclause 6.2

2) Relative Frequency Tolerance
It is not possible to verify by testing that the RF-frequencies of all the TRXs in a BSS are all derived from the same frequency source. This may be confirmed by a manufacturers declaration

3) Synchronization of Data Clocks
It is not possible to verify by testing that the BSS clocking and timebase are derived from the same frequency source used for RF generation. This may be confirmed by a manufacturers declaration.

9.2.1 Timing Tolerance

9.2.1.1 Test purpose

Description

The timing tolerance is the relative time between bits of the same bit number (BN) in timeslots of the same timeslot number (TN) and frame number (FN) transmitted simultaneously from two TRXs in one cell of the BSS.

9.2.1.2 Test case

For a BTS supporting only one TRX, this test is not performed.
If SFH is supported by the BSS, it shall be disabled during this test.

The TRX supporting the BCCH shall be used as the reference to measure the relative timing of the transmissions from different TRXs. The results shall be analysed to ensure that the relative timing between any two TRXs which are transmitted simultaneously meets the requirement.

a) The BSSTE shall establish a TCH using timeslot 0 in the TDMA frame structure (TN=0) on a specific frequency.

b) The relative timing between this frequency and the BCCH frequency shall be measured on the training sequence (BN=74). TDMA-frames carrying Frequency Correction bursts or Synchronization bursts (T3=0,1,10,11,20,21,30,31,40,41) shall be excluded. This measurement shall be repeated for at least 100 bursts and the average shall be calculated and recorded.

c) Step a) and b) shall be carried out with the TCH at the RF channels B, M and T, but avoiding the use of the same frequency as the BCCH.

d) Step c) shall be carried out with the BCCH at the RF channels B, M and T, but avoiding the use of the same frequency as the TCH. The BCCH frequency shall be offset to the closest used RF channel in case they coincide.

e) Step d) shall be repeated for all other TRXs in the BSS.

f) The timing measurement shall be used to establish the maximum time difference between any two TRXs operating in one cell.

9.2.1.3 Essential conformance

This test need not to be performed.

9.2.1.4 Complete conformance

Test environment
Normal

Conformance requirement

The timing difference (as established in step f)) between any two TRXs shall be no greater than 1/4 symbol, measured at the BSS antenna connector.

9.2.1.5 Requirement reference

GSM 05.10 (EN 300 912) [24].

9.3 Frame structure

This section verifies that the BSS correctly generates TDMA frames and is capable of receiving transmitted bursts from Mobile Stations generated according to GSM recommendations.

The following requirements relating to frame structure are not tested in this subclause:

1) The delay of three timeslots between uplink and downlink is tested implicitly in subclause 9.6

2) The SACCH multiframe structure is implicitly tested in transmitter bit exactness test of subclause 6.1.

9.3.1 BCCH Multiframe

9.3.1.1 Test purpose

Description

The BCCH multiframe consists of 51 TDMA frames. There is one BCCH logical channel per BSS, which broadcasts general information. The frequency information is carried on the Frequency Correction Channel (FCCH) and the synchronization is transmitted on the Synchronization Channel (SCH).

9.3.1.2 Test case

a) The BSS is configured with one TRX configured to support a BCCH. This is monitored in he BSSTE.

b) The BSSTE shall search for the Frequency Correction burst.

c) The BSSTE shall then search for the Synchronization burst.

9.3.1.3 Essential conformance

This test need not to be performed.

9.3.1.4 Complete conformance

Test environment

Normal

Conformance requirements

1) The BSSTE shall detect Frequency Correction bursts at T3 = 0, 10, 20, 30 and 40 and for no other T3 (T3 = FN mod 51, FN = TDMA frame number).

2) The BSSTE shall also detect synchronization bursts at T3 = 1, 11, 21, 31, and 41 and for no other T3.

3) At the SCH the BSSTE shall detect the BSIC set up for the BTS/BSS. This applies to any BSIC. The BSSTE shall also detect the correct RFN for the various T3s.

9.3.1.5 Requirement reference

GSM 05.02 (ETS 300 908) [19] and GSM 05.10 (EN 300 912) [24].

9.3.2 TDMA-frame structure

9.3.2.1 Test purpose

Description

One TDMA frame consists of eight timeslots, with an average length of 156.25 symbol periods. This may be achieved by setting all timeslots to be 156.25 symbol periods or setting timeslots 0 and 4 to 157 symbol periods and the remaining (1, 2, 3, 5, 6, 7) to 156 symbol periods. This section will test that the BSS conforms to the declared frame structure.

9.3.2.2 Test case

The BSS shall be configured to generate multiframes with a combination of logical channels which gives a contiguous stream of normal or dummy bursts as defined in GSM 05.02 for more than one frame. If SFH is supported by the BSS, it shall be disabled during this measurement.

The slot lengths will be measured between the leading edge of the 14th symbol of the training sequence for that timeslot and the leading edge of the 14th symbol of the training sequence for the next timeslot.

9.3.2.3 Essential conformance

This test need not to be performed.

9.3.2.4 Complete conformance

Test environment

Normal

Conformance requirements

The measurements shall conform to the frame structure 1) or 2) as declared by the manufacturer.

1) The length of each timeslot shall be 156.25 symbol periods.

2) The length of timeslots 0 and 4 shall be 157 symbol periods and the length of the remaining (1, 2, 3, 5, 6, 7) shall be 156 symbol periods.

9.3.2.5 Requirement reference

GSM 05.02 (ETS 300 908) [19] and GSM 05.10 (EN 300 912) [24].

9.4 Radio link measurements

Whilst calls are being established and for their duration, the reception quality shall be continuously assessed in the BSS as criteria for handover and RF power control algorithms. The following criteria may be employed in order to perform this assessment:

– Signal strength (RXLEV)

– Signal quality (RXQUAL)

– MS-BSS distance

– Idle channel level.

The handover and power control strategies based on above parameters are up to the operator.

Some test cases in this subclause assume that the manufacturer provides appropriate logical or physical test access to perform all tests in this subclause. The manufacturer may also show compliance to the requirements by other means agreed between the parties.

9.4.1 Signal Strength

9.4.1.1 Measurement Accuracy

9.4.1.1.1 Test purpose

Description

RXLEV is the received signal level measured at the BSS receiver input averaged over a reporting period of length of 1 SACCH multiframe for a TCH and a SDCCH. This test verifies the range and the accuracy of this parameter.

9.4.1.1.2 Test case

If the manufacturer does provide appropriate logical or physical access to perform all the tests in this subclause, the tests shall be performed according to the test cases below.

The manufacturer shall declare how many TRXs the BSS supports:

1 TRX: The test shall be performed on B, M, T

2 TRX: Tests shall be performed on B,M,T and both TRXs shall be tested on at least one frequency.

3 TRX or more: Three TRXs shall be tested, one on B, one on M and one on T.

If Slow Frequency Hopping (SFH) is supported by BSS, it shall be disabled during this test.

a) A test signal with normal GSM modulation originated from the BSSTE shall be applied to the BSS RX antenna connector on one timeslot.

b) The test signal level shall be adjusted over the level range ‑110.5 dBm to ‑47.5 dBm in 1 dB steps and shall be kept stable for one reporting period.

c) The RXLEV measurements shall be performed under static propagation conditions only.

9.4.1.1.3 Essential conformance

This test need not to be performed.

9.4.1.1.4 Complete conformance

Test environment

Normal and extreme temperature
Conformance requirements

1) The RXLEV value shall nominally be mapped to the received signal level as in Table 28

Table 28: Signal level estimation requirements

RXLEV

Power level:

0

less than ‑110 dBm

1

‑110 dBm to ‑109 dBm

2

‑109 dBm to ‑108 dBm

.

.

.

.

62

‑49 dBm to ‑48 dBm

63

greater than ‑48 dBm

2) For any input signals of level x1 and x2 dBm within the range ‑110 dBm to ‑48 dBm, where x1 is above reference sensitivity in table 30, x1<=x2 and (x2-x1)<= 20 dB, the corresponding measured values y1 and y2 shall be such that

(x2-x1) – a<= (y2-y1) <= (x2-x1) + b

where a and b are the tolerances given in table 29 (see GSM 05.08, 8.1.2).

Table 29: Tolerance for relative accuracy of received signal strength measurement

Absolute level of lower level signal x1 in dBm

Tolerance in dB

normal-BTS

DCS1800, PCS 1900 and

MXM 1900

micro-BTS M1

GSM900, GSM 850 and MXM 850

Micro-BTS M1

and

DCS1800, PCS 1900 and MXM 1900

Micro BTS M2

GSM900, GSM 850 and MXM 850

micro-BTS M2

and

DCS1800, PCS1900 and MXM 1900

micro BTS M3

GSM900, GSM 850 and MXM 850

micro-BTS M3

a

b

>=‑90

>=‑88

>=‑83

>=‑78

>=‑73

2

2

>=‑103

>=‑101

>=‑96

>=‑91

>=‑86

3

2

<‑103

<‑101

<‑96

<‑91

<‑86

4

2

NOTE: It is optional for the BSS to be able to report values below the reference sensitivity in table 30.These specifications apply to measurements which are on the same or on different RF channel.

3) The RMS received signal level at the receiver input shall be measured with an absolute accuracy of +/- 4 dB from ‑110 dBm to ‑70 dBm under normal conditions and +/- 6 dB over the full range of ‑110 dBm to – 48 dBm under both normal and extreme temperature conditions.

3) If the received signal level falls below the reference sensitivity level for the type of BSS then the BSS shall report a level within a range allowing for the absolute accuracy given in conformance requirement c) above. In case the upper limit of this range is below the reference sensitivity level for the type of BSS, then the upper limit shall be considered as equal to the reference sensitivity level in table 30.

Table 30: Reference sensitivity level

BTS Type

Reference sensitivity level

GSM 400/GSM900/GSM 850/DCS1800/PCS 1900/MXM 850/MXM 1900 BTS

‑104 dBm

GSM900/GSM 850/MXM 850 micro‑BTS M1

‑97 dBm

GSM900/GSM 850/MXM 850 micro‑BTS M2

‑92 dBm

GSM900/GSM 850/MXM 850 micro‑BTS M3

‑87 dBm

GSM900/GSM 850/MXM 850 pico‑BTS P1

‑88 dBm

DCS1800/PCS 1900/ MXM 1900 micro‑BTS M1

‑102 dBm

DCS1800/PCS 1900/ MXM 1900 micro‑BTS M2

‑97 dBm

DCS1800/PCS 1900/ MXM 1900 micro‑BTS M3

‑92 dBm

DCS1800/PCS 1900/ MXM 1900 pico-BTS P1

‑95 dBm

9.4.1.2 Selectivity of signal strength measurements

9.4.1.2.1 Test purpose

Description

The received signal level (RXLEV) defined in 9.4.1.1 shall be able to discriminate between wanted signal in actual RF channel and interfering signal in adjacent ARFCN´s. This selectivity characteristic is tested in this subclause.

9.4.1.2.2 Test case

If the manufacturer does provide appropriate logical or physical access to perform all the tests in this subclause, the tests shall be performed according to the test cases below.

If Slow Frequency Hopping (SFH) is supported by BSS, it shall be disabled during this test.

a) As a minimum the test shall be performed on one TRX on one timeslot on one ARFCN.

b) One of the following test set-ups is used:

Test set-up A
Two input signals shall be connected to the receiver via a combining network. The test signal with normal GSM modulation shall have a power level 20 dB above the reference sensitivity level. The interfering signal shall be continuous, and have GSM modulation of pseudo-random bitstream without midamble

Test set-up B
The BSSTE shall establish a call set-up with the BSS and the RXLEV of the assigned channel shall be output from the BSS. The test signal with normal GSM modulation shall have a power level 20 dB above the reference sensitivity level.

c) Register the signal strength (RXLEV value) with only the wanted signal present at the RX input port.

d)

i) In test set-up A repeat the measurements with interferer input signal frequency offset and input level increased for each offset according to table 31a

Table 31a: Interferer offset and input level for RXLEV selectivity measurements

Interferer frequency offset

Relative input level

±200 kHz

9 dB

±400 kHz

41 dB

ii) In test set-up B repeat the measurements with the input signal at frequency offsets and signal levels according to table 31b.

Table 31b: Test signal offset and input level for RXLEV selectivity measurements

Frequency offset

Relative input level

200 kHz

16 dB

400 kHz

48 dB

d) The measurements shall be performed under static propagation conditions only.

9.4.1.2.3 Essential conformance

This test need not to be performed.

9.4.1.2.4 Complete conformance

Test environment

Normal

Conformance requirements

The reported RXLEV value shall in test case d) not exceed

– the value in test case c) with more than 1 for test set-up A.

– the value in test case c) for test set-up B.

9.4.1.2.5 Requirement reference

GSM 05.08 (EN 300 911) [23], subclause 8.1

9.4.2 Signal quality

9.4.2.1 Test purpose

Description

The received signal quality (RXQUAL) is specified in terms of bit error ratio (BER) before channel decoding averaged over a reporting period of length of 1 SACCH multiframe on a TCH or a SDCCH.

This test verifies the range and the accuracy of this parameter

9.4.2.2 Test case

If the manufacturer does provide appropriate logical or physical access to perform all the tests in this subclause, the tests shall be performed according to the test cases below.

As a minimum the test shall be performed on one TRX on one timeslot on one ARFCN

a) One of the following test set-ups is used:

Test set-up A
If Slow Frequency Hopping (SFH) is supported by BSS, it shall be enabled during this test
for hopping on different ARFCN for all timeslots over the specified hopping bandwidth specified by the manufacturer.
A call shall be set up between the BSSTE and the BSS.

Test set-up B
If Slow Frequency Hopping (SFH) is supported by BSS, it shall be disabled during this test.
Two input signals originated from the BSSTE shall be connected to the receiver via a combining network. The test signal with normal GSM modulation shall have a power level 20 dB above the reference sensitivity level. The interfering signal shall be a random, continuous, GSM modulated signal on the same ARFCN. For test in TU50, each signal shall be connected through a multipath fading simulator (MFS) as described in Annex B1.

b) The signal level (set-up A) and the interferer signal level (set-up B) respectively shall be varied such that the BER on the wanted TCH measured at the logical interface point before channel decoding in the BSS are within all the BER ranges for the RXQUAL values in table 32a in turn.

c) The logical reference point before channel decoding may be obtained by using the unprotected class II bits after channel decoding before any extrapolation is applied. Half-rate channels are measured by first establishing a full-rate channel, measuring the error ratio and then establishing a half-rate channel and checking the indicated error ratio.

d) For each BER range, 1000 RXQUAL values shall be recorded, and with and without uplink DTX.

e) The measurement shall be performed under the propagation conditions static and TU50.

9.4.2.3 Essential conformance

This test need not to be performed.

9.4.2.4 Complete conformance

Test environment

Normal

Conformance requirements

1) Table 32a shows the minimum probability that, when on a TCH, a specified value of RXQUAL shall be reported for a BER within the range as indicated in the table under static propagation conditions.

2) Table 32b shows the minimum probability that, when on a TCH, a specified value of RXQUAL or an adjacent value shall be reported for a BER within the range as indicated in the table under TU50 multipath propagation conditions.

Table 32a: Signal quality estimation requirements (static)

RXQUAL:

Range of actual BER:

Probability that correct RXQUAL

band is reported shall exceed

Full rate:

Half rate

DTX:

0

< 0.10%

90 %

90 %

65 %

1

0.26% – 0.30 %

75 %

60 %

35 %

2

0.51% – 0.64 %

85 %

70 %

45 %

3

1.0% – 1.3 %

90 %

85 %

45 %

4

1.9% – 2.7 %

90 %

85 %

60 %

5

3.8% – 5.4 %

95 %

95 %

70 %

6

7.6% – 11.0%

95 %

95 %

80 %

7

> 15.0

95 %

95 %

85 %

NOTE 1: For the full-rate channel RXQUAL_FULL is based on 104 TDMA frames.

NOTE 2: For the half-rate channel RXQUAL_FULL is based on 52 TDMA frames.

NOTE 3: For the DTX-mode RXQUAL_SUB is based on 12 TDMA frames.

Table 32b: Signal quality estimation requirements (TU50)

Expected RXQUAL_FULL:

Range of actual BER:

Probability that expected RXQUAL_FULL is reported shall exceed

0/1

< 0.10%

85 %

1/0/2

0.26% – 0.30 %

85 %

2/1/3

0.51% – 0.64 %

85 %

3/2/4

1.0% – 1.3 %

90 %

4/3/5

1.9% – 2.7 %

90 %

5/4/6

3.8% – 5.4 %

90 %

6/5/7

7.6% – 11.0%

90 %

7/6

> 15.0

90 %

9.4.2.5 Requirement reference

GSM 05.08 (EN 300 911) [23], subclause 8.2

9.4.3 Idle channel signal level

9.4.3.1 Test purpose

Description

A procedure shall be implemented by which the BSS monitors the levels of interference on its idle traffic channels. These measurements are used for handover and channel allocation. This test verifies that BSS can measure signal strength including interference with appropriate accuracy on an idle channel. The measured signal strength of each idle channel is, after averaging, classified in one of five interference bands and reported to MSC on request in a RESOURCE INDICATION message. The report method, report period, averaging period and definition of interference band values are defined in a O&M message by the operator and the manufacturer. See GSM 08.08, subclauses 3.1.3, 3.2.18 and 3.2.2.48, and GSM 08.58 subclauses 8.6.1 and 9.3.21.

9.4.3.2 Test case

If the manufacturer does provide appropriate logical or physical access to perform all the tests in this subclause, the tests shall be performed according to the test cases below.

If the BSS is supporting SFH, this shall be disabled during this test.

The manufacturer shall declare how many TRXs the BSS supports:

1 TRX: The test shall be performed on B, M, T

2 TRX: Tests shall be performed on B,M,T and both TRXs shall be tested on at least onefrequency.

3 TRX or more: Three TRXs shall be tested, one on B, one on M and one on T.

a) A GMSK signal modulated with pseudo-random bit sequences connected to BTS RX input.

b) The limits of the 5 possible interference bands, the reporting period and the averaging period are defined by the operator and the manufacturer.

c) The signal from the BSSTE shall be adjusted over the level range ‑110.5 dBm to ‑47.5 dBm in 1 dB steps.

d) The measured signal strength is recorded for each signal level.

e) The interference level expressed as one of the 5 possible interference level bands, included in the RF_RES_IND message, shall be stored for each idle channel.

9.4.3.3 Essential conformance

This test need not be performed.

9.4.3.4 Complete conformance

Test environment

Normal

Conformance requirement

1) The accuracy requirements 1)‑4)for the measured signal strength in subclause 9.4.1.1.4 apply.

2) The measured signal levels shall be mapped into the interference level band defined in b) and the corresponding band value included in the RF_RES_IND message for each idle channel.

9.4.3.5 Requirement reference

GSM 05.08 (EN 300 911) [23] Annex A 3.1 e).

9.5 Adaptive frame alignment

9.5.1 Test purpose

Description

Adaptive frame alignment is the mechanism by which the timeslots transmitted by the MS are initially and dynamically adjusted in time so that the received timeslots in the BSS always fall within the correct time window. This mechanism is controlled by the BSS.

The adaptive frame alignment mechanism is needed since the guard time between timeslots in the timeslot structure is not long enough to cope with MS-BSS propagation delays due to absolute distance. The MS timing is initially adjusted (initial alignment) when accessing the BSS, and is then continuously adjusted for relative distance variations during the call (dynamic alignment).

The BSS continuously monitors the delay in the transmission from the MS relative to the expected signal from an MS at zero range. This is required to give Timing Advance information to the MS. This information of the delay (up to 63 or 219 symbols) may also be used as a criteria for initiating handover at the cell boundary (MAX_MS_RANGE).

This subclause also tests the MS-BSS distance assessment.

9.5.2 Test case

If Slow Frequency Hopping (SFH) is supported by the BSS, it shall be disabled during this measurement. The tests shall be performed at least on one TRX for the radio frequency channels B, M and T and for timeslots 0 (initial alignment) and 1 or other timeslot except 0 (dynamic alignment) respectively on the same TRX.

The tested timeslots shall be exposed to static propagation conditions.

For each MS emulated the signal strength shall be according to table 33. If other TCH than TCH/FS is to be used in the test, a input signal 3 dB below the test signal input level specified in subclause 7.3 shall be used.

Table 33: Input signal strength at test of adaptive time alignment

BTS Type

Input signal strength

GSM 400/GSM900/GSM 850/DCS1800/PCS 1900/MXM 850/MXM 1900 BTS

‑107 dBm

GSM900/GSM 850/MXM 850 micro‑BTS M1

‑100 dBm

GSM900/GSM 850/MXM 850 micro‑BTS M2

‑95 dBm

GSM900/GSM 850/MXM 850 micro‑BTS M3

‑90 dBm

DCS1800/PCS 1900/ MXM 1900 micro‑BTS M1

‑105 dBm

DCS1800/PCS 1900/ MXM 1900 micro‑BTS M2

‑100 dBm

DCS1800/PCS 1900/ MXM 1900 micro‑BTS M3

‑95 dBm

To avoid a radio link time-out during the tests, the BSSTE may generate an uplink SACCH with a non-limiting signal strength including MEASUREMENT REPORT messages signalling high RXLEV and low RXQUAL values. This applies to the emulated MS when configured with an SACCH, i.e. configured with a dedicated channel.

a) Initial alignment: Random access bursts shall be input on the RACH (timeslot 0) as often as possible using different random references .

b) The emulated round-trip propagation delay shall be 3 different values corresponding to TA-values from 0 to 63 (0 to 219 for GSM 400) in turn (low, medium and high).

c) The Timing Advance (TA) value reported to the MS shall be monitored and compared.

d) Dynamic alignment: A TCH/FS (other TCH can be used if TCH/FS is not supported) shall be established between the BSSTE and the BSS.

e) The emulated round-trip propagation delay of the MS shall vary corresponding to the vehicle speed of 500 km/h, starting from maximum MS-BSS distance moving close to the BSS and back.

The Timing Advance (TA) value signalled to the Mobile Station shall be monitored and compared.

9.5.3 Essential conformance

This test need not be performed

9.5.4 Complete conformance

Test environment

Normal

Conformance requirements

The difference DELTA between the emulated round-trip propagation delay and the signalled TA-value for the applicable timeslot shall be evaluated in symbols rounded to the nearest integer for at least 1000 pairs of timeslots, and shall have the following properties:

1) For initial alignment, the mean of DELTA shall be not greater 0 +/- 1 symbol.

2) For dynamic alignment, the mean of DELTA shall be not greater than 0 +/- 5/4 symbol.

3) Under all conditions the standard deviation of DELTA shall be less than 1 symbol.

NOTE: +/- 1 symbol tolerance is +/- 1/2 symbol for assessment error and +/- 1/2 symbol for quantization error.

NOTE: Requirement 1 above may need 1 additional symbol of tolerance if the BSS has an RX-TX delay tolerance of +/- 1 symbol.

The maximum allowed TA-value signalled to the MS is 63 symbols except for GSM 400 where it is 219 symbols.

9.5.5 Requirement reference

GSM 05.10 (EN 300 912) [24]

Annex A (informative):
Testing of statistical parameters

When measuring statistical parameters like Bit Error Rates (BERs) or Frame Erasure Rates (FERs), the statistical nature of the error events may result in a natural variance in the observed test results. This variance will depend on the number of events observed. Consequently, due to such statistical limitations with the aim to reduce the test time to a minimum, some overall requirements should be met, indicating a certain confidence in the observed results.

Defining a "good" BSS as a BSS which on a long term basis (tested over an infinite time) meets the system requirement for an individual test, and a "bad" BSS as a BSS which on a long term basis fails the system requirement for an individual test, the overall requirements are the following:

1) The probability of passing a "good" BSS should be as high as possible.

2) The probability of passing a "bad" BSS should be as low as possible.