RF

(1)

Basic RF measurement quantities:

Definitions and measurement

Soc Classification level

1 © Nokia Siemens Networks

LTE OPT Training

Last Updated: April 10

th

, 2012

Version 1.9

(2)

Content

•

RSRP: Reference Signal Received Power

•

RSSI: Reference Signal Strength Indicator

•

RSRQ: Reference Signal Received Quality

•

SINR / CINR / SNR etc

–

Definition and measurement

2 © Nokia Siemens Networks Presentation / Author / Date

–

Mapping RSRP / RSRQ to SINR

•

MIMO variants of the RF quantities and use cases

(3)

Reference signal from three 2Tx cells

Recap (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 1 R1: tx antenna 2

0 0 0R 0R 0R 0R 1 1 1R 1R 1R 1R 0 0 0R 0R 0R 0R 1 1 1R 1R 1R 1R 0 0 0R 0R 0R 0R 1 1 1R 1R 1R 1R

(4)

Reference signal from three 2Tx cells

Recap (2/2)

•

RS as seen by UE from cells of a 3-sector site no RS-RS interference since

PCImod3 are different

0R 0R 0R 1R 1R 1R 0R 0R 0R 1R 1R 1R

Sector 1: blue

Sector 2: green

Sector 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 R0: tx antenna 1 R1: tx antenna 2

(5)

RSRP

Definition 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. R0: tx antenna 1

R1: tx antenna 2

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.

(6)

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

(7)

RSRP measurement using different devices

-70 -65

1 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 R S R P [ d B m ] time, seconds PCTel RSRP R&S RSRP UE RSRP Measured RSRP depends on scanner UE starts downloading

(8)

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?

(9)

Content

•

• –

–

•

(10)

0R 1R 0R 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

(11)

Relation between RSSI and RSRP

Theory

•

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

(12)

Relation between RSSI and RSRP

Field 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 R S R P [ d B m ]

-125 -115 -105 -95 RSSI [dBm] R S R P [ d B m ] Measurement: slope ~28 dB agrees with theory (27.8dB)

(13)

Relation of RSSI and RSRP

Impact 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 P o w e r, d B m 30 40 50 60 N u m b e r o f R B s SCell-RSSI(Com) SCell-RSRP(Com) RB Num(DL) RSRP

13 © Nokia Siemens Networks 100%,10MHz cell

•

RSRP is independent of cell load -100 -90 -80 -70 Time, seconds P o w e r, d B m 0 10 20 30 N u m b e r o f R B s RSSI increases about 5-6dB RSRP independent of cell load UE starts downloading

(14)

Content

•

• –

–

•

(15)

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

R0: tx antenna 1

(16)

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

(17)

RSRQ

Theoretical values in idle / fully loaded cell

•

Idle cell:

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 = -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 utilization

If 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

(18)

RSRQ/R measurement, 2Tx cell

Idle cell RSRQ=

•

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

UE starts downloading RSRQ=

(19)

-6 -4 -2 0 1 31 61 91 121 151 181 211 241 271 301 331 361 391 R S R Q [ 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 R S R Q [ d B ] time, seconds PCTel RSRQ R&S RSRQ UE RSRQ

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

Measured RSRQ depends

on device UE starts

(20)

Content

•

• –

–

•

(21)

SINR / SNR / CINR

Definition 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

(22)

50 60 70 80 S IN R d B m / t h ro u g h p u t M b p 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%

22 © Nokia Siemens Networks 0 10 20 30 40 50 1 31 61 91 121 151 181 211 241 271 301 331 361 391 S IN R d B m / t h ro u g h p u t M b p s time, seconds PCTel RS SINR R&S SSS SINR UE SINR DL Throughput

Secondary Sync Signal SNR much worse than RS SNR

(23)

RS CINR and RSRP measured by FDD scanner

RS CINR RS CINR

(24)

Content

•

• –

–

•

(25)

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

(26)

Mapping RSRP to SNR

Samsung 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

(27)

Mapping RSRP to SNR

XCAL 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] R S R P [ d B m ]

(28)

Mapping RSRQ to SINR

Theory (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

N RSRP

SINR *12

+ =

RSSI RSRP N RSRQ P xN RSRP P RSSI RBs N used RB RE x xN P P P P SINR N n i RE n xN n N n i * * # _ / 12 _ _ _ 12 _ = + + = = = ∗ = + = x RSRQ xN RSRP RSRQ RSRP N N RSRP SINR − = − = 1 12 * * 12 *

(29)

Mapping RSRQ to SINR

Theory (2/2)

•

Equation used:

Where x=RE/RB

•

2RE/RB equals to empty cell.

Only Reference Signal power is

x RSRQ SINR − = 1 12 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) S IN R ( d B ) 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 ) S N R ( d B ) SNR

(30)

Mapping RSRQ to SINR

Lab 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 S IN R

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

– 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 S N R Measured-full traffic Calculated Measured - no traffic

(31)

Mapping RSRQ to SINR

Field 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 R S R Q [ d B ] CINR [dB] RSRQ SINR mapping unreliable in practice. (depends on meas device!) RSRQ SINR mapping unreliable in practice. (depends on meas device!)

(32)

Content

•

• –

–

•

(33)

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

(34)

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

(35)

Rx antenna power imbalance

Field 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 02/18/2010 14:15:16.143 02/18/2010 14:16:58.647 02/18/2010 14:18:40.153 02/18/2010 14:20:22.159 02/18/2010 14:22:03.167

35 © Nokia Siemens Networks -95 -90 -85 -80 -75 -70 -65 time R S S I [d B m ] Average of SCell-RSSI(Com) Average of RSSI(Ant0) Average of RSSI(Ant1) Time Data

(36)

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 10 Rank fluctuation at high RSSI (SNR) Rx power imbalance dB Rank indicator, {0,1} seconds seconds

(37)

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:

(38)

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 1

x-pol versus two vertical

x-pol vertical

x-pol has better peak

0 10 20 30 40 50 60 70 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 MAC DL tput, Mbps C D F vertical better peak tput Vertical has better tput in bad RF

(39)

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 90

PHY tput for substreams, EPA 3km/h, 2.6GHz, 3GPP low, medium, high correlation stream 1 stream 2 stream 1 + stream 2 High spatial correlation low spatial correlation

50 100 150 200 250 0 10 20 30 40 50 60 time, seconds P H Y t p u t [M b it s /s e c ] correlation medium spatial correlation

(40)

Impact of UE orientation

Measurement in excellent RF

0 20 40 60 80 100 1 ₁9 3 7 5 5 7 3 9 1 1 0 9 1 2 7 1 4 5 1 6 3 1 8 1 1 9 9 2 1 7 2 3 5 2 5 3 2 7 1 2 8 9 3 0 7 3 2 5 3 4 3 3 6 1 3 7 9 3 9 7 4 1 5 4 3 3 4 5 1 4 6 9 4 8 7 5 0 5 5 2 3 5 4 1 5 5 9 5 7 7 5 9 5 6 1 3 6 3 1 6 4 9 6 6 7 P H Y t p u t [M b p s] , C IN R [ 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 1 0 9 1 2 7 1 4 5 1 6 3 1 8 1 1 9 9 2 1 7 2 3 5 2 5 3 2 7 1 2 8 9 3 0 7 3 2 5 3 4 3 3 6 1 3 7 9 3 9 7 4 1 5 4 3 3 4 5 1 4 6 9 4 8 7 5 0 5 5 2 3 5 4 1 5 5 9 5 7 7 5 9 5 6 1 3 6 3 1 6 4 9 6 6 7 R S R P [ d B m ] 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