01 rf measurement_and_optimization

Soc Classification level 1 © Nokia Siemens Networks Presentation / Author / Date

RF measurements quantities and optimization


Content

• LTE RF measurement quantities in field measurements– RSRP– RSSI– RSRQ– SINR– MIMO variants of these– 3GPP defined measurement accuracy for UEs

• Measurement results with different measurement tools• RF optimization


SINR vs. RSSI vs. RSRP and RSRQ


Field measurement parameters

• 3GPP is defining following measurements:– RSRP (Reference Signal Received Power)– RSRQ (Reference Signal Received Quality)

• Scanners and terminals are typically measuring following RF quantities:– RSRP– RSRQ– RSSI, Wideband channel power– P-SCH, S-SCH power– RS SINR, P-SCH/S-SCH SINR

• Understanding of different measurement quantities is very important for field performance analysis.


RSRP, 3GPP definition• RSRP is the average received power of a single RS resource element. • UE measures the power of multiple resource elements used to transfer the

reference signal but then takes an average of them rather than summing them.• Reporting range -44…-140 dBm

Definition Reference signal received power (RSRP), is defined as the linear average over the power contributions (in [W]) of the resource elements that carry cell-specific reference signals within the considered measurement frequency bandwidth. For RSRP determination the cell-specific reference signals R0 according TS 36.211 [3] shall be used. If the UE can reliably detect that R1 is available it may use R1 in addition to R0 to determine RSRP. The reference point for the RSRP shall be the antenna connector of the UE. If receiver diversity is in use by the UE, the reported value shall not be lower than the corresponding RSRP of any of the individual diversity branches.

Applicable for RRC_IDLE intra-frequency, RRC_IDLE inter-frequency, RRC_CONNECTED intra-frequency, RRC_CONNECTED inter-frequency

Note1: The number of resource elements within the considered measurement frequency bandwidth and within the measurement period that are used by the UE to determine RSRP is left up to the UE implementation with the limitation that corresponding measurement accuracy requirements have to be fulfilled.

Note 2: The power per resource element is determined from the energy received during the useful part of the symbol, excluding the CP.


RSRP mapping 3GPP TS 36.133 V8.9.0 (2010-03)

• The reporting range of RSRP is defined from -140 dBm to -44 dBm with 1 dB resolution.

• The mapping of measured quantity is defined in the table below.

Reported value Measured quantity value UnitRSRP_00 RSRP -140 dBmRSRP_01 -140 RSRP < -139 dBmRSRP_02 -139 RSRP < -138 dBm

… … …RSRP_95 -46 RSRP < -45 dBmRSRP_96 -45 RSRP < -44 dBmRSRP_97 -44 RSRP dBm


Reference Signals recap: OFDMA Channel Estimation• Channel estimation in LTE is based on reference signals (like CPICH functionality

in WCDMA)• Reference signals position in time domain is fixed (0 and 4 for Type 1 Frame)

whereas in frequency domain it depends on the Cell ID• In case more than one antenna is used (e.g. MIMO) the Resource elements

allocated to reference signals on one antenna are DTX on the other antennas• Reference signals are modulated to identify the cell to which they belong.

Antenna 1 Antenna 2

subc

arrie

rs

symbols 60 symbols 60

subcarriers

According 3gpp spec, RSRP and RSSI measured at RS symbol instants only


RSSI

• RSSI not reported to eNodeB by UE– Can be computed from RSRQ and RSRP that are reported by UE

• RSSI measures all power within the measurement bandwidth– Measured over those OFDM symbols that contain RS– Measurement bandwidth RRC-signalled to UE


RSSI and RSRP

• RSSI = wideband power= noise + serving cell power + interference power

• Without noise and interference, 100% DL PRB activity: RSSI=12*N*RSRP– RSRP is the received power of 1 RE (3GPP definition) average of power levels received

across all Reference Signal symbols within the considered measurement frequency bandwidth

– RSSI is measured over the entire bandwidth– N: number of RBs across the RSSI is measured and depends on the BW

• Based on the above, under full load and high SNR:

RSRP (dBm)= RSSI (dBm) -10*log (12*N)


RSSI versus RSRP, measurement with Samsung in fully loaded 10MHz cell

RSRP versus RSSI for fully loaded cell, 10MHz system bandwidth (100% of REs active)

-125

-115

-105

-95

-85

-75

-93 -88 -83 -78 -73 -68 -63 -58 -53 -48

RSSI [dBm]

RSR

P [d

Bm

]

Measurement: 95 dBm – 67 dBm = 28 dB agrees with theory

(27.8dB)


RSRP versus RSSI versus number of RBs, Samsung BT-3710 example

• RSSI increases about 5dB when RB activity increases to 100%, 10MHz cell

-100

-90

-80

-70

-60

-50

-401 6 11 16 21 26 31 36

Time, seconds

Pow

er, d

Bm

0

10

20

30

40

50

60

Num

ber

of R

Bs

SCell-RSSI(Com)SCell-RSRP(Com)RB Num(DL)

RSSI increases

about 5-6dB

RSRP independent of cell load


RSRQ

• RSRQ = N x RSRP / RSSI– N is the number of resource blocks over which the RSSI is

measured, typically equal to system bandwidth– RSSI is pure wide band power measurement, including intracell

power, interference and noise

• RSRQ reporting range -3…-19.5dB

Definition Reference Signal Received Quality (RSRQ) is defined as the ratio N×RSRP/(E-UTRA carrier RSSI), where N is the number of RB’s of the E-UTRA carrier RSSI measurement bandwidth. The measurements in the numerator and denominator shall be made over the same set of resource blocks. E-UTRA Carrier Received Signal Strength Indicator (RSSI), comprises the linear average of the total received power (in [W]) observed only in OFDM symbols containing reference symbols for antenna port 0, in the measurement bandwidth, over N number of resource blocks by the UE from all sources, including co-channel serving and non-serving cells, adjacent channel interference, thermal noise etc. The reference point for the RSRQ shall be the antenna connector of the UE. If receiver diversity is in use by the UE, the reported value shall not be lower than the corresponding RSRQ of any of the individual diversity branches.

Applicable for RRC_CONNECTED intra-frequency, RRC_CONNECTED inter-frequency


RSRQ reporting range

• RSRQ = N x RSRP / RSSI– N is the number of resource blocks over which the RSSI is

measured, typically equal to system bandwidth– RSSI is pure wide band power measurement, including intracell

power, interference and noise

• RSRQ reporting range -3…-19.5dB

Reported value Measured quantity value UnitRSRQ_00 RSRQ -19.5 dBRSRQ_01 -19.5 RSRQ < -19 dBRSRQ_02 -19 RSRQ < -18.5 dB

… … …RSRQ_32 -4 RSRQ < -3.5 dBRSRQ_33 -3.5 RSRQ < -3 dBRSRQ_34 -3 RSRQ dB


RSRQ and serving cell powerRSRQ = RSRP / (RSSI/N), N = number of PRBs

• RSSI = noise + serving cell power + interference power during RS symbolRSRQ depends on serving cell power and the number of Tx antennas

• Impact of serving cell power to RSRQ:Example for noise limited case (no interference):If all resource elements are active and are transmitted with equal power thenRSRQ = N / 12N = -10.8 dB for 1TxRSRQ = N / 20N = -13 dB for 2Tx, taking DTX into account(because RSRP is measured over 1 resource element and RSSI per resource

block is measured over 12 resource elements). Remember that RSSI is only measured at those symbol times during which RS REs are transmitted.

When there is no traffic, and assuming only the reference symbols are transmitted(there are 2 of them within the same symbol of a resource block) from a single Tx antenna then the RSSI is generated by only the 2 reference symbols so the result becomes;

RSRQ = N / 2N = -3 dB for 1Tx

RSRQ = -6dB for 2Tx. Quiz: where does this value come from?


SINR definition

• SINR is the reference value used in the system simulation• SINR can be defined:

1. Wide band SINR2. SINR for a specific subcarriers (or for a specific resource elements)

• SINR = S/(I+N), all measured over the same bandwidth

• Most drive test UEs and scanners support SINR or SNR measurement Example: LG supports RS SNR measurement Example: Samsung BT-3710 measures CINR from RS (e-mail info from

Samsung)


SNR vs. RSRP

• RSRP to SNR mapping• RSRP is measured for a single subcarrier

– noisepower_for_15KHz= -125.2dBm Noise figure = 7 dB Temperature = 290 K

• Assumption: RSRP doesn’t contain noise power

powernoiseKHzPPRSRPSNR

REn

REn

__15_

_

RSRP vs. SNR

-15.00

-10.00

-5.00

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

-135 -130 -125 -120 -115 -110 -105 -100 -95 -90 -85 -80 -75 -70

RSRP (dBm)

SNR

(dB

)

SNR

This curve gives upper limit to SINR with certain RSRP. SINR

is always lower than SNR in live network due to

interference.


SNR vs. RSRP, measurement, Samsung BT-3710SINR versus RSRP, measurement

-125

-115

-105

-95

-85

-75

-7 -2 3 8 13 18 23

SINR [dB]

RSR

P [d

Bm

]

Fading channel measurement, drive test.

With Samsung not very strong correlation between CINR and

RSRP


RSRQ to SINR mapping

• RSRQ depends on own cell traffic load, but SINR doesn’t depend on own cell load.– Used Resource Elements per Resource Block (RE/RB) in serving cell is an

input parameter for RSRQ -> SINR mapping– Assumption: RSRP doesn’t contain noise power

RSSIRSRPNRSRQ

PxNRSRPPRSSIRBsN

usedRBRExxNPP

PPNRSRPSINR

Nni

REnxNn

Nni

**

#_/

12*

12_

__

12_

xRSRQ

xNRSRPRSRQ

RSRPNNRSRPSINR

1

12

**12*


RSRQ to SINR mapping

• Equation used:

– x=RE/RB

• 2RE/RB equals to empty cell. Only Reference Signal power is considered from serving cell.

• 12RE/RB equals to fully loaded serving cell. All resource elements are carrying data.

• In practice, mapping from RSRQ to SINR seems difficult– Currently available measurement

UEs and scanners report SINR directly

xRSRQ

SINR

112

RSRP vs. SNR

-15.00

-10.00

-5.00

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

-135 -130 -125 -120 -115 -110 -105 -100 -95 -90 -85 -80 -75 -70

RSRP (dBm)

SNR

(dB

)

SNR

RSRQ vs SINR

-10.00

-5.00

0.00

5.00

10.00

15.00

20.00

25.00

30.00

-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3

RSRQ (dB)

SIN

R (d

B)

2 RE/RB

4 RE/RB

6 RE/RB

8 RE/RB

10 RE/RB

12 RE/RB

Difficult to estimate SINR in this region from RSRQ, SINR very

sensitive to RSRQ and cell load


RSRQ to SINR mapping, scanner measurement • Lab measurements matches well the

calculated results• Measured with Agilent scanner

– RSRP– RSRQ– Reference signal SINR

• Cable connection between BTS and scanner– Attenuator used to reduce signal level– No traffic = only control channels and

reference signals– Full traffic load = data send in each RB

Note: Validity of formulae have been proven in lab under above conditions and with only one cell on air ( i.e. no other cell interference). Measurements from the field will differ as exact load can not be set

SINR vs. RSRQ

-10

-5

0

5

10

15

20

25

30

35

-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0

RSRQ

SIN

R

Measured - full traffic Caculated - no traffic Calculated - full traffic load Measured - no traffic

SNR vs. RSRP

-15

-10

-5

0

5

10

15

20

25

30

35

40

-140 -135 -130 -125 -120 -115 -110 -105 -100 -95 -90 -85 -80

RSRP

SNR

Measured-full trafficCalculatedMeasured - no traffic


RSRP vs. DL throughput

Example measurement•Drive test, 20MHz BW, ~2.6GHz •FTP download, no other interfering traffic in the network

RSRP vs. throughput

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

-130

-128

-126

-124

-122

-120

-118

-116

-114

-112

-110

-108

-106

-104

-102

-100 -98

-96

-94

-92

-90

-88

-86

-84

-82

-80

-78

-75

dBm

Mbp

s


SINR vs. DL throughput

Example measurement•Drive test, 20MHz BW, ~2.6GHz

SINR vs. throughput

0

10

20

30

40

50

60

70

-4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

dB

Mbp

s


MIMO variants

Terminals used with drive test tool report RSRP, RSSI and RSRQ per receive antenna

• RSRP– RSRP0 measured at rx antenna 0 (avg. power of RS CEs from tx1 & tx2 ?)– RSRP1 measured at rx antenna 1 (avg. power of RS CEs from tx1 & tx2 ?)

• Antenna-based RSSI and RSRP measurements can detect rx branch power imbalance

Some scanners can report RSRP and SINR for both tx-branches by measuring only with single antenna.

• RSRP– RSRP1tx, measured RS CEs from tx1– RSRP2tx, measured RS CEs from tx2


MIMO variants, measurement, Samsung (1)• Example: RSSI measured by UE• Note rx power imbalance between receive antenna branches

-95

-90

-85

-80

-75

-70

-65

-60

-55

-50

-45 02/18/2010

14:13:33.716 02/18/2010

14:15:16.143 02/18/2010

14:16:58.647 02/18/2010

14:18:40.153 02/18/201014:20:22.159

02/18/201014:22:03.167

time

RSS

I [dB

m]

Average of SCell-RSSI(Com)Average of RSSI(Ant0)Average of RSSI(Ant1)

Time

Data


MIMO variants, measurement, Samsung (2)• Example: RSRQ measured by UE• Note that no notable RSRQ imbalance between receive antenna

branches

-20

-18

-16

-14

-12

-10

-8

-6 02/18/201014:13:33.716

02/18/201014:15:16.143

02/18/201014:16:58.647

02/18/201014:18:40.153

02/18/201014:20:22.159

02/18/201014:22:03.167

time

RSR

Q [d

B]

Average of RSRQ(Ant0)Average of RSRQ(Ant1)

Time

Data


MIMO variants, measurement, Samsung (3)• Example: RSRP measured by UE with one external antenna (one

external antenna rx1 + one internal antenna used rx2)• Note rx big power imbalance between receive antenna branches

RSRP from ant1 (external) and ant2

-95

-90

-85

-80

-75

-70

-65

-601 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73

sec

dBm Ant 1

Ant 2


Channel correlation

•Channel correlation impacts on MIMO performance– If the correlation is high, SM mode doesn’t increse throughput even with

high SINR (compared to Tx diversite mode).

•Channel correlation is not usually reported by DT tools or scanners.

– Can be estimated by following Rank Indicator (RI) reported by UE and SINR measurements.

– Some terminals (LG) report correlation matrix to DT tool.


EPA03, change of correlation, MIMO subchannel throughputs, from fading simulator, LG terminal, 20MHz, 2.6GHz, SINR=25dB

50 100 150 200 2500

10

20

30

40

50

60

70

80

90PHY tput for substreams, EPA 3km/h, 2.6GHz, 3GPP low, medium, high correlation

time, seconds

PH

Y tp

ut [M

bits

/sec

]

stream 1stream 2stream 1 + stream 2

High spatial correlation

medium spatial correlation

low spatial correlation

High spatial correlation causes rank-1 transmission even at high SNR (second stream tput almost zero)


Measurement accuracy requirement, 3GPP TS 36.133 (simplified)• RSRP absolute accuracy under normal conditions

– ± 6dB intra-frequency and inter-frequency– Needed for setting random access pre-amble tx power and triggering

coverage-based handover (A5)• RSRP relative accuracy between two cells under normal conditions

– ± 2dB intra-frequency– ± 6dB inter-frequency– For triggering better cell handover (A3)

• RSRP reporting range in signalling (handovers)Reported value Measured quantity value Unit

RSRP_00 RSRP -140 dBm

RSRP_01 -140 RSRP < -139 dBm

RSRP_02 -139 RSRP < -138 dBm

… … …

RSRP_95 -46 RSRP < -45 dBm

RSRP_96 -45 RSRP < -44 dBm

RSRP_97 -44 RSRP dBm


Measurement accuracy requirement, 3GPP TS 36.133 (simplified)• RSRQ absolute accuracy under normal conditions

– ± 2.5dB intra-frequency and inter-frequency

• RSRQ relative accuracy between two cells under normal conditions– ± 3dB inter-frequency (intra-frequency not defined)

• RSRQ reporting range in RRC signalling (handovers)

Reported value Measured quantity value Unit

RSRQ_00 RSRQ -19.5 dB

RSRQ_01 -19.5 RSRQ < -19 dB

RSRQ_02 -19 RSRQ < -18.5 dB

… … …

RSRQ_32 -4 RSRQ < -3.5 dB

RSRQ_33 -3.5 RSRQ < -3 dB

RSRQ_34 -3 RSRQ dB


Measurement results with different measurement tools


Measurement differences

Scanners and terminals have own specific algorithms for RF measurements.RSRP•Can be measured from the whole bw or from part of the bw.RSRQ•RSSI, used in the RSRQ definition, can be measured from the whole bw or from part of the bw.SINR•Measured from Reference Signal or from Synchronization channel.Other differences•Averaging methods•Sampling rate•Receiver sensitivity•Cell info decoding capabilities


Impact of serving cell traffic, SINR

•Measurement location is in the middle of the dominance area (high RSRP and SINR).•UE SINR is impacted by own cell load.•PCTel RS SINR is impacted slightly by own cell load.•S-SCH SINR is not impacted by own cell load

SINR and throughput

0

10

20

30

40

50

60

70

80

1 31 61 91 121 151 181 211 241 271 301 331 361 391 421

sec

dB/M

bps

PCTel SSYNCPCTel RSR&S SSYNCUEDL Troughput

R&S S-SCH SINR

PCTel S-SCH SINR

PCTel RS SINR

PCTEL, R&S, UE

Neighbour cell RS SINR


Impact of serving cell traffic, RSRQ

•R&S RSRQ is not reacting at all to own cell traffic.

Download started in the serving cell

RSRQ

-16

-14

-12

-10

-8

-6

-4

-2

01 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365 379 393 407 421

sec

dB

PCTelR&SUE

PCTEL, R&S, UE

PCTel RSRQ

R&S RSRQ


Impact of intra eNodeB interference on SINR

Both cell in idle state(no traffic)

Download started in the serving (blue) cell

Download started in the neighbor (red) cell

Download stopped in the neighbor (red) cell

Neighbor (red) cellshut down

R&S S-SCH SINR

PCTel S-SCH SINR

PCTel RS SINR

PCTEL, R&S, UE


SINR

0

5

10

15

20

25

30

35

40

1 15 29 43 57 71 85 99 113 127 141 155 169 183 197 211 225 239 253 267 281 295 309 323 337 351 365 379

sec

dB

UEPCTel SSYNCPCTel RSR&S SSYNC

Impact of intra eNodeB interference, SINR

Both cell in idle state(no traffic)


Download started in the neighbor cell

Download stopped in the neighbor cell

Neighbor cellshut down

Big variance on SINR measurements, depending on:

•Measurement method

•Measurement equipment

PCTEL, R&S, UE


RSRP

-90

-85

-80

-75

-70

-651 14 27 40 53 66 79 92 105 118 131 144 157 170 183 196 209 222 235 248 261 274 287 300 313 326 339 352 365 378

sec

dBm

UEPCTelR&S

RSRQ

-14

-13

-12

-11

-10

-9

-8

-7

-6

-5

-41 14 27 40 53 66 79 92 105 118 131 144 157 170 183 196 209 222 235 248 261 274 287 300 313 326 339 352 365 378

sec

dB

UEPCtelR&S

Impact of intra eNodeB interference,RSRP & RSRQ

Download started in the serving cell Neighbor cell

shut down

PCTEL, R&S, UE

•PCTel and R&S are showing similar average RSRP• UE used internal antenna• scanners were connected to the same external antenna using a power splitter• RSRQ values are quite different for UE, R&S and PCTel• RSRP seems like the most reliable based on this measurement


Impact of intra eNodeB interference, RSRQ

Download started in the serving (blue) cell

Neighbor (red) cellshut down

PCTel RSRQ

R&S RSRQ

PCTEL, R&S, UE


Impact of inter site neighbor, SINR

PCTEL, JDSU, UESINR serving cell

-5

0

5

10

15

20

1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97 103 109 115 121 127 133 139 145

sec

dB

JDSU RS31PCTel RS31PCTel S-SCH31UE 31

SINR neighbor cell

-5

0

5

10

15

20

1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97 103 109 115 121 127 133 139 145

sec

dB

JDSU RS34PCTel RS34PCTel S-SCH34


•Measurement location is between sites at the cell edge•Serving cell SINR is not impacted by own cell traffic.•Neighbor cell SINR is decreased by serving cell traffic.


Driving, idle vs. download RSRP

No traffic (ping only) on serving cell

FTP download on serving cell

PCTel RSRP

R&S RSRP

PCTel RSRP

R&S RSRP

PCTEL, R&S, UE


Summary

•Absolute SINR measurement values can’t be used as a reliable performance indicator.

– Operators should be educated, not to believe blindly measured SINR values.

– Relative SINR changes can be used as performance indicator, if the same measurement tool is used all the time.

•SINR measured from S-SCH and RS behaves differently depending on the interference situation (intra/inter eNodeB).•Detailed SINR measurement methods of the terminals and scanners are not known.• The most robust and reliable measurement quantity seems to be RSRP


RF optimization


RF optimization

Basic RF planning is important•Clear cell dominance areas•Avoid sites shooting over large areas with other cells

Antenna tilting has big impact on other cell interference, at least in planning tool estimates•No LTE reference measurements available


Example from 3HK trial – reusing 3G sitesImproving performance by blocking excess cells

• Overall SINR is improved due to reduction of inter-cell interference

• Locations with improved SINR are visible on the map

• Improvement in throughput is even more significant (see next slide)

FT_04.1 Mobility DT DL - SINR comparison

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

-11 -9 -7 -5 -3 -1 1 3 5 7 9 11 13 15 17 19 21 23 25 27

SI NR ( dB)

CDF

% All cellsBlocked cells

All cells

Blocked cells

Ave SINR improved from 15.2dB to 17.4dB


Example from 3HK trial – reusing 3G sitesImproving performance by blocking excess cells

FT_04.1 Mobility DT DL - Throughput comparison

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64

Phy DL t put ( Mbps)

CDF

% All cellsBlocked cells

Ave throughput improved from 23.34Mbps to 26.78Mbps, i.e. 14.7%

Key message:• The number of LTE cells when converted

from all existing 3G sites seem to be more than sufficient, and cell overlapping and hence inter-cell interference seems to be excessive in outdoor environment.

• Careful planning and cell/antenna selection process, and initial RF tuning is important to the LTE field performance

All cells

Blocked cells


Detecting interference - SINR

•SINR measurements can indicate interference areas, but it doesn’t necessarily see all interference sources:

– Impacted by network load. Traffic in the neighboring cells will reduce Serving cell SINR.

– Depends on the measurement method (RS or SCH) and tool– Depends on PCI planning (RS SINR)


Detecting interference RSRP

•RSRP measurement with scanner is the most reliable way to detect areas with possible interference problems.

– Not impacted by network load

01 rf measurement_and_optimization

Technology

dbm rsrp

rsrp dbm

power rsrp

corresponding rsrp

rsrp determination

rsrp reference signal

rsrp rsrq rssi

gpp definition rsrp