DocumentRF

Basic RF measurement quantities: Definitions and measurement

Soc Classification level 1 © Nokia Siemens Networks

LTE OPT TrainingLast Updated: April 10th, 2012Version 1.9

Presentation / Author / Date

Content

• RSRP: Reference Signal Received Power

• RSSI: Reference Signal Strength Indicator

• RSRQ: Reference Signal Received Quality

• SINR / CINR / SNR etc

– Definition and measurement

Soc Classification level 2 © Nokia Siemens Networks Presentation / Author / Date


– Mapping RSRP / RSRQ to SINR

• MIMO variants of the RF quantities and use cases

• 3GPP defined measurement accuracy for UEs

Reference signal from three 2Tx cellsRecap (1/2)

• RS frequency shift based on PCI modulo-3

0R

0R

0R

0R

1R

1R

1R

1R

0R

0R

0R

0R

1R

1R

1R

1R

0R

0R

0R

0R

1R

1R

1R

1R

R0: tx antenna 1R1: tx antenna 2


00

0R

0R

0R

0R

1 1

1R

1R

1R

1R

00

0R

0R

0R

0R

1 1

1R

1R

1R

1R

00

0R

0R

0R

0R

1 1

1R

1R

1R

1R

Reference signal from three 2Tx cellsRecap (2/2)

• RS as seen by UE from cells of a 3-sector site � no RS-RS interference since PCImod3 are different

0R

0R

1R

1R

0R

0R

1R

1R

0R

0R

1R

1R

Sector 1: blueSector 2: greenSector 3: red

R : tx antenna 1


0R

0R

0R

0R

0R

0R

1R

1R

1R

1R

1R

1R

0R

0R

0R

0R

0R

0R

1R

1R

1R

1R

1R

1R

0R

0R

0R

0R

0R

0R

1R

1R

1R

1R

1R

1R


RSRPDefinition in 3GPP TS 36.214

• RSRP is the linear average of received power of RS resource elements• UE measures the rx power of multiple RS REs and takes average of 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 Unit

RSRP_00 RSRP < -140 dBm

RSRP_01 -140 ≤ RSRP < -139 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

RSRP measurement using different devices

-70

-651 31 61 91 121 151 181 211 241 271 301 331 361 391

RSRP measurement, idle cell and fully loaded cell

• Comparison of Samsung UE and R&S TMSW / PCTel EX scanners– UE inside car and scanners using the same antenna mounted on car roof


-90

-85

-80

-75

RS

RP

[d

Bm

]

time, seconds

PCTel RSRPR&S RSRPUE RSRP

Measured RSRP depends on scanner

UE starts downloading

Quiz: RSRP vs CPICH RSCP• Below a measurement of 2.1GHz CPICH RSCP versus 1.8GHz/20MHz

RSRP using a multi-RAT multi-band scanner. • Q: Should RSRP be scaled somehow to make it comparable to RSCP?


Content











0R

0R

1R

1R

RSSI

• RSSI is internal to UE, not reported in uplink to eNB• RSSI measures all power within the measurement bandwidth

– hence it includes thermal noise, interference and serving cell power• Measured over those OFDM symbols that contain RS for antenna port R0


0R

0R

0R

0R

0R

0R

1R

1R

1R

1R

1R

1R

OFDM symbol containing R0

Relation between RSSI and RSRPTheory

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

• Without noise and interference, 100% DL PRB activity:

Where:– N is number of PRBs across the RSSI is measured and depends on the BW

RSSI=12*N*RSRP


• Based on the above, under 100% PRB utilization and high SNR:

Presentation / Author / Date

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

Relation between RSSI and RSRPField measurement in fully loaded 10MHz cell

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

-85

-75

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

RS

RP

[dB

m]


-125

-115

-105

-95

RSSI [dBm]

RS

RP

[dB

m]

Measurement: slope ~28 dB �

agrees with theory (27.8dB)

Relation of RSSI and RSRPImpact in cell load

• RSSI increases about 5dB when PRB activity increases to 100%,10MHz cell -70

-60

-50

-40

1 6 11 16 21 26 31 36

Po

wer

, d

Bm

30

40

50

60

Nu

mb

er o

f R

Bs

SCell-RSSI(Com)

SCell-RSRP(Com)

RB Num(DL)

RSRP


100%,10MHz cell

• RSRP is independent of cell load

-100

-90

-80

-70

Time, seconds

Po

wer

, d

Bm

0

10

20

30

Nu

mb

er o

f R

Bs

RSSI increases

about 5-6dB

RSRP independent of cell load


Content











RSRQ 3GPP TS 36.214

Where:

• RSSI is pure wide band power measurement, including serving cell power, interference and thermal noise

• N: RSSI measurement bandwidth in PRBs

RSRQ = N x RSRP / RSSI



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

• RSRQ reporting range -3…-19.5dB• Some devices report values < -20dB

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_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

RSRQ Theoretical values in idle / fully loaded cell

• Idle cell:When there is no traffic, and assuming only the reference symbols aretransmitted (there are 2 of them within the same symbol of a resource block) froma single Tx antenna then the RSSI is generated by only the 2 reference symbolsso the result becomes:– RSRQ = -3 dB for 1Tx

RSRQ = N x RSRP / RSSI, N = measured bandwidth in PRBs


– RSRQ = -3 dB for 1Tx– RSRQ = -6dB for 2Tx

• Fully loaded cell: Cell with 100% PRB utilizationIf all resource elements are active and are transmitted with equal power then– RSRQ = -10.8 dB for 1Tx– RSRQ = -13 dB for 2Tx (assumption: rx div not used!)

In practice, UEs seem to report RSRQ ~11dB for fully loaded cell, assuming no other-cell interference

RSRQ/R measurement, 2Tx cell

Idle cell RSRQ=

• Only serving cell on-air, FDD ZTE (QC)



RSRQ= neg6dB

-6

-4

-2

01 31 61 91 121 151 181 211 241 271 301 331 361 391

RS

RQ

[d

B]

RSRQ measurement, idle cell and fully loaded cell

RSRQ measurement using different devices

• Comparison of Samsung UE and TMSW / PCTel EX scanners– UE inside car, scanners on car roof using the same antenna


-20

-18

-16

-14

-12

-10

-8

RS

RQ

[d

B]

time, seconds

PCTel RSRQR&S RSRQUE RSRQ

NOTE: R&S RSRQ independent of cell load!! (a bug?)

Measured RSRQ depends

on deviceUE starts

downloading

Content











SINR / SNR / CINRDefinition and Measurement

• Usually SINR=SNR=CINR unless the receiver is able to separate interference from thermal noise• Example: IRC receiver can separate dominant interferer from noise

• The actual measurement definition must be checked for every measurement device• which signal the ‘S’ measures?

SINR = S / (I+N)


• which signal the ‘S’ measures?• what is the measurement bandwidth?• is the measured SNR instantaneous or average value?• if the average SNR is measured, what is the number of samples averaged?

• Typically UEs measure SNR from RS• Exact measurement method uncertain, since UE chipset vendors don’t typically reveal

technical data

• Scanners measure SNR from PSS/SSS and/or RS

50

60

70

80

SIN

R d

Bm

/ th

roug

hput

Mbp

s

SINR measurement, idle cell and fully loaded cell

PCTel SSS SINR

PCTel RS SINR

SINR measurement using different devices

• Comparison of Samsung UE and TMSW / PCTel Ex scanners

DL starts, PRB

utilization to100%


0

10

20

30

40

50

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

SIN

R d

Bm

/ th

roug

hput

Mbp

s

time, seconds

PCTel RS SINR

R&S SSS SINR

UE SINR

DL Throughput

Secondary Sync Signal SNR much worse than RS SNR

RS CINR and RSRP measured by FDD scanner

RS CINRRS CINR


Content











Mapping RSRP to SNR

• Motivation: Sometimes might be useful to be able to map RSRP to SNR

• Assuming only thermal noise, RSRP can be mapped approximately to SNR– Quiz: How?

• In practical field conditions, this does not seem to work very well:


– Insufficient averaging for fading signal by measurement equipment– Tx signal nonlinearities cause EVM that saturates measured SNR to ~25-

30dB. – Receiver non-idealities, such as carrier frequency offset, Doppler spread,

oscillator, phase noise etc also saturate SNR– Interference

• Example on next slide

Mapping RSRP to SNRSamsung XCAL measurement in lab, non-fading channel

•Lab measurement in shielded box, UE-reported RSRP versus UE-reported SINR


In 100% loaded cell SINR saturates sooner, probably

due to subcarrier leakage or other imperfections.

Mapping RSRP to SNRXCAL field measurement, fading channel

• Fading channel measurement, drive test.

SINR versus RSRP, measurement

-85

-75

-7 -2 3 8 13 18 23

Samsung measurement values are instantaneous

snapshots � high variance due to fading.


-125

-115

-105

-95

SINR [dB]

RS

RP

[d

Bm

]

Mapping RSRQ to SINRTheory (1/2)

• 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

NRSRPSINR

12*

+=


RSSI

RSRPNRSRQ

PxNRSRPPRSSI

RBsN

usedRBREx

xNPP

PPSINR

Nni

REnxNn

Nni

*

*

#

_/

12_

__

12_

=

++==

=

∗=

+=

xRSRQ

xNRSRPRSRQ

RSRPNNRSRP

SINR−

=−

=1

12

**

12*

Mapping RSRQ to SINRTheory (2/2)

• Equation used:

Where x=RE/RB

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

xRSRQ

SINR−

=112

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 (

dB

)

2 RE/RB

4 RE/RB

6 RE/RB

8 RE/RB

10 RE/RB

12 RE/RB


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

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)

SN

R (

dB

)

SNR

Mapping RSRQ to SINRLab measurements

• Lab measurements match well the calculated results

• Measured with Agilent scanner– RSRP– RSRQ– Reference signal SINR

• Cable connection between BTS and scanner

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


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

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

SN

R

Measured-full traffic

Calculated

Measured - no traffic

Mapping RSRQ to SINRField measurement

• Samsung BT-3730 + XCAL

-3

-1

1

0 5 10 15 20 25 30

SINR versus RSRQ, field measurement


-15

-13

-11

-9

-7

-5

RS

RQ

[d

B]

CINR [dB]

RSRQ�SINR mapping unreliable in practice.

(depends on meas device!)

RSRQ�SINR mapping unreliable in practice.

(depends on meas device!)

Content











MIMO variants

• Terminals and scanners may report RSRP, RSSI, RSRQ, SINR per receiver antenna

• Example: RSRP– RSRP0 measured at rx. antenna 0 – RSRP1 measured at rx. antenna 1

• Measurement capabilities should be checked case-by-case


• Measurement capabilities should be checked case-by-case

• Some devices measure even more MIMO quantities– Altair chipset measures four SINR values for a 2Tx BTS, one value per Tx-Rx

pair– The old LG chipset measures antenna correlation coefficient

• Question: How can these multi-antenna measurements be used in practical field testing?

Use cases for multi-antenna RF quantities

• MIMO throughput is degraded by:

– Too large power imbalance between receiver branches– Tx/Rx antenna correlation

• In practice, only power imbalance can be measured by commercial devices


Rx antenna power imbalanceField measurement• Note RSSI power imbalance between receive antenna branches

– This should be eliminated, if possible

-60

-55

-50

-45 02/18/2010

14:13:33.716 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


-95

-90

-85

-80

-75

-70

-65

time

RS

SI [

dB

m]

Average of SCell-RSSI(Com)

Average of RSSI(Ant0)

Average of RSSI(Ant1)

Time

Data

Rx antenna power imbalance: Impact on rank Field measurement

• Too large antenna power imbalance can reduce channel rank � tx falls to single stream even at high SNR

-85

-80

-75

-70

-65

RSSI0

RSSI1

Rx power imbalance


50 100 150 200 250 300 350 400 450 500 550-90

-85

50 100 150 200 250 300 350 400 450 500 550-5

0

5

10Rank fluctuation at high RSSI (SNR)

Rx power imbalance dB

Rank indicator, {0,1}

seconds

seconds

Channel correlation

• High channel correlation degrades multi-stream throughput performance– Even high SNR won’t improve throughput if channel correlation is too high

• Typical problem case: Two vertical antennas on car roof � SNR is excellent but MIMO throughput is bad– Solution: use cross-polarized antennas, or take antennas inside car!


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

Channel correlation, drive test example• The same dt route driven twice, 2.6GHz@10MHz, Huawei E398, tm3

– vertical: two vertical scanner antennas mounted on car roof, mean tput = 29Mbps– x-pol: custom-made cross-polarized omni antenna on car roof, mean tput = 33Mbps

0.7

0.8

0.9

1x-pol versus two vertical

x-pol

verticalx-pol has

better peak


0 10 20 30 40 50 60 700

0.1

0.2

0.3

0.4

0.5

0.6

0.7

MAC DL tput, Mbps

CD

F

verticalbetter peak tput

Vertical has better tput in

bad RF

Impact of channel correlation on MIMO Tput Measurement example, SINR=25dB• SINR=25dB, constant over the measurement• High spatial correlation causes rank-1 transmission even at very high SNR!!

(second stream tput goes to almost zero)

70

80

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

stream 1

stream 2stream 1 + stream 2

High spatial correlation

low spatial correlation


50 100 150 200 2500

10

20

30

40

50

60

time, seconds

PH

Y t

put

[Mbi

ts/s

ec]

correlation

medium spatial correlation

Impact of UE orientation Measurement in excellent RF

0

20

40

60

80

100

1

19

37

55

73

91

10

9

12

7

14

5

16

3

18

1

19

9

21

7

23

5

25

3

27

1

28

9

30

7

32

5

34

3

36

1

37

9

39

7

41

5

43

3

45

1

46

9

48

7

50

5

52

3

54

1

55

9

57

7

59

5

61

3

63

1

64

9

66

7

PH

Y t

pu

t [M

bp

s], C

INR

[d

B]

Time

PHY DL Throughput [Mbps]

CINR 0 [dB]

CINR 1 [dB]

Main message 1: MIMO throughput is very sensitive to UE orientation!

Main message 2: Eliminating power UE


Time

-85

-80

-75

-70

-65

-60

-55

1

19

37

55

73

91

10

9

12

7

14

5

16

3

18

1

19

9

21

7

23

5

25

3

27

1

28

9

30

7

32

5

34

3

36

1

37

9

39

7

41

5

43

3

45

1

46

9

48

7

50

5

52

3

54

1

55

9

57

7

59

5

61

3

63

1

64

9

66

7

RS

RP

[d

Bm

]

Time

Antenna0 RSRP [dBm]

Antenna1 RSRP [dBm]

Eliminating power imbalance or channel correlation alone is not sufficient for good MIMO throughput. Both should be eliminated! In practice, this is a trial and error process in stationary measurements.

UE orientation changes

Documents

DocumentRF