####Good_Radio SIG 24.11.11 Brian Collins -Antennas for MIMO

7/31/2019 ####Good_Radio SIG 24.11.11 Brian Collins -Antennas for MIMO

1/25

Antenova 2011Commercial in Confidence Integrated Antenna and RF Solutions

Many aspects of this presentation are protected by UK and International Patents and Patent

Applications

Antennas for MIMO in userequipments

Brian CollinsChief Engineer, Applications & Business Development


2/25

Antenova 2011Commercial in Confidence Integrated Antenna and RF Solutions 2 of 27

Introduction

The objective of employing MIMO is to maximise the userdata throughput in a given spectral resource (b/s/Hz)

This is not the same as providing a single user with avery high data rate

MIMO throughput is a function of

the number of channels

the S/N (C/I) ratio in each channel the correlation between them (less = better)

On a small mobile platform we have:

Very constrained dimensions for antennasSmall physical spacing between antennas

The users head/hand/body.


3/25


Gain or low correlation?

These curves show thatthere is little to be

gained by reducing ebelow about 0.5

For any value of e,everything is to begained by increasing theSNR

We must not

compromise gainwhen we add moreantennas to our UE.

Average channel capacity (b/s/Hz)v SNR and spatial correlation coefficient (e)

for 2 x 2 MIMO [Ohlmer]

Perfecte Poor


4/25


Channel gain equality a reminder

As the SNR decreases, MIMO

performs more and more like n-branch diversity

This graph reminds us thatunless the signals in eachbranch are reasonably similarthe realised diversity gain is

small

If they are to be effective,the antennas need to havesimilar gain

Any idea that the additional

antennas can have lowerperformance than the mainantenna is misguided.Diversity gain at 90% signal reliability in a 2-branch system as

a function of cross-correlation and mean signal level differencebetween branches (maximal ratio combining). [Turkmani et al]


5/25


LTE: Assigned frequency bands

LTEBand

Number

Uplink(MHz)

Downlink(MHz)

Width ofBand(MHz)

DuplexSpacing(MHz)

BandGap

(MHz)

1 1920 - 1980 2110 - 2170 60 190 130

2 1850 - 1910 1930 - 1990 60 80 20

3 1710 - 1785 1805 -1880 75 95 20

4 1710 - 1755 2110 - 2155 45 400 355

5 824 - 849 869 - 894 25 45 206 830 - 840 875 - 885 10 35 25

7 2500 - 2570 2620 - 2690 70 120 50

8 880 - 915 925 - 960 35 45 10

9 1749.9 - 1784.9 1844.9 - 1879.9 35 95 60

10 1710 - 1770 2110 - 2170 60 400 340

11 1427.9 - 1452.9 1475.9 - 1500.9 20 48 28

12 698 - 716 728 - 746 18 30 12

13 777 - 787 746 - 756 10 31 41

14 788 - 798 758 - 768 10 30 40

15 1900 - 1920 2600 - 2620 20 700 680

16 2010 - 2025 2585 - 2600 15 575 560

17 704 - 716 734 - 746 12 30 18

18 815 - 830 860 - 875 15 45 30

19 830 - 845 875 - 890 15 45 30

20 832 - 862 791 - 821 30 41 71

21 1447.9 - 1462.9 1495.5 - 1510.9 15 48 33

22 3410 - 3500 3510 - 3600 90 100 10

23 2000 - 2020 2180 - 2200 20 180 160

24 1625.5 - 1660.5 1525 - 1559 34 101.5 135.5

25 1850 - 1915 1930 - 1995 65 80 15

LTE BandNumber

Allocation (MHz)Width(MHz)

33 1900 - 1920 20

34 2010 - 2025 15

35 1850 - 1910 60

36 1930 - 1990 60

37 1910 - 1930 20

38 2570 - 2620 50

39 1880 - 1920 40

40 2300 - 2400 100

41 2496 - 2690 194

42 3400 - 3600 200

43 3600 - 3800 200

FDD (paired) bands TDD (unpaired) bands

Not pretty not even complete going to get worse

Unrealistic to cover everything How many different combinations?


6/25


Handset Antennas

In 20 years, antennas have evolved from single frequency externalmonopoles to 5-band internal designs

824-960MHz and 1710-2170MHz

Typical free space efficiency >50%

Worst case free space ~10%

Typical hand/head efficiency 15%

Now we are looking to add more bands

Anywhere from 698MHz to 2.6GHz and beyond

How can we do this?.


7/25


Present reaction of industry

More of the same

Extended coverage over 698-960MHz is notreally possible in handset

Tuneable or reconfigurable antennas

Switches (existing technology, many vendors,

low cost, high power, low loss)

Digitally controlled capacitors (new technology,low Q, few vendors)

Adaptive tuning is attractiveBut tuning a poor antenna is no solution.


8/25


Tuning/reconfiguring

5 ways of tuning/reconfiguring a PIFA (+ combinations)

and a PIFA is not our only starting point

Basic structure (eg PIFA)


9/25


Hardware options

Switchable matching circuits

Tuneable components in matching circuits

Re-configurable antennasparts of the antenna are electronically

connected/disconnected

Variety of active components:

Conventional RF Switches GaAsFETs, UltraCMOS,

MEMS Capacitors

BST capacitors


10/25


Constraints

High operating TX power

High burst power on GSM

High peak-to-average power ratio (PAPR) for W-CDMA, LTE

High degree of linearity needed to avoid radiation of spurioussignals

Any ancillary circuits must operate from low voltage at low current

Small size, small PCB footprint

Simple interface to processor

May need to match at widely separated uplink/downlink bands

Low cost


11/25


MIMO

Creating two antennas to provide sufficiently de-correlatedsignals is possible, even on a device like a smartphone

Generally easier on a tablet of laptop

Main issue is to obtain sufficient gain (efficiency) in smallantennas that have to cover wide bands

User interaction

Hands and head

Orientation/grip

Issues are easiest to manage in a laptop, where high data

rates are most needed Most difficult in handsets main penalty for failing is loss of

spectral efficiency


12/25


MIMO Tools (think receive mode)

To obtain MIMO performance we need antennas withlow signal correlation

Our tools are:Spatial separation

Antennas receive multipath signals having different

phase and amplitude at two separated locations

Polarisation

Antennas respond to signals with different

polarisations (H/V, 45, RHCP/LHCP)

Radiation pattern Antennas receive signals from different spatial

directions


13/25


Envelope Correlation

The envelope correlation coefficient e is a measure of how theradiation pattern of two antennas differ in shape, polarization andphase.

Low correlation is important for the performance of a MIMO system

e = 1 identical patterns, no MIMO gain

e = 0 orthogonal patterns, optimal MIMO gain

In the upper bands it is easy to excite different radiation modes andachieve low correlation

In the lower bands the wavelength is large compared to the UE size

The only available radiation modes with good efficiency and bandwidthhave currents along the largest dimension of the UE groundplane.


14/25


Correlation limit for LTE700

Same mode:

High correlation

Antenna #1

Antenna #2

e~4 XPR cos()

XPR [ 3 cos()+1] + 3 sin() = cos() ~ 0.4For XPR = 0dB, ~ 50

Assuming good isolation andusing isotropic propagation model:

Radiation patterns are

skewed ~

~ 50

Diagonal modes:

Mid/low correlation


15/25


Placing handset antennas

X SARX HAC

X SARX HACX Hand

X Hand

Browse

Hobsons Choice!


16/25


A 9-band LTE antenna

Agilis

A10346 9-bandLTE Antenna

A single antenna 40mm x 12mm x3.2mm with switch-selectable LTE-700/W-CDMA2100 and 4-bandGSM operation

Efficiency on 100mm x 40mm evaluation board


17/25


Case study

USB Dongle for use

with a laptop computer

698-2700MHz3

5mm

30.0mm8mm

75mm

87mm

Area available forcircuit (top and

bottom)

USB


18/25


LTE Dongle: Correlation

Correlation coefficient computed from complex 3-D radiation patterns

Isotropic 3D propagation model

XPR between -15dB and +15dB

Stand-alone DongleXPR [dB]

-15

-13

-11

-9

-7

-5

-3

-1

1

3

5

7

9

11

13

15

746 759 772 785 7980

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]

2500 2540 2580 2620 2660 27000

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]

1710 1802 1894 1986 2078 21700

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]

820 855 890 925 9600

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]

698 710 722 734 7460

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]


19/25


USB Dongle MIMO: Correlation

Correlation coefficient computed from complex 3-D radiation patterns

Isotropic 3D propagation model

XPR between -15dB and +15dB

Dongle Connected to a LaptopXPR [dB]

-15

-13

-11

-9

-7

-5

-3

-1

1

3

5

7

9

11

13

15

820 855 890 925 9600

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]

698 710 722 734 7460

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]

746 759 772 785 7980

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]

2500 2540 2580 2620 2660 27000

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]

1710 1802 1894 1986 2078 21700

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

[MHz]


20/25


Testing

It is comparatively simple to measure the 3-D complexradiation patterns of the antennas on a UE and to

compute the correlation coefficient(s) from them

There is currently no single 3GPP-agreed method forover-the-air testing of MIMO

The latest release contains candidate methods, 5methods using an anechoic chamber and 2 methodsusing a reverberation chamber

There is a significant difference in the complexity ofthe tests, the cost of equipment and the resultsprovided by these methods.


21/25


Fully adaptive tuning

Looks like the answer to many prayers

Well suited to TDD systems

But

Low Q-factor (25/50) means losses are high, especially at lowfrequencies where the antenna is smallest

complex control interaction with the processor on the UE,

especially for downlink optimisation FDD systems require simultaneous control of tuning on up-

and down-links

Must support existing GSM/W-CDMA as well as LTE

and what about LTE/A?


22/25


23/25


Practicalities

Sufficient space must be available for antennas

Multiple antennas must be sited where they will work

Reasonably low correlation of signalsNot under the users hand especially in browser mode

Current experience is that antennas have little priority in

design inadequate space

poor location

effects of neighbouring components

Covering ever more frequency bands erodes performance


24/25


Conclusion

MIMO is a valuable and established technique forenhancing spectral reliability

Works well with strong signals and rich multipath

Effective MIMO needs efficient, independent antennas

Practical implementation faces challenges

Low priority for antenna performance in UE ID Conflict of function and fashion

Small volume

Poor placement

Bandwidth limitations of small antennas

Lack of standardised frequency bands


25/25


References & acknowledgement

A M D Turkmani, A A Arowojolu, P A Jefford & C J Kellett, An experimental

evaluation of the performance of two-branch space diversity schemes at1800MHz, IEEE Trans Veh Tech, Vol VT-44 No 2, May 1995, pp 318 326

E Ohlmer et al, Urban Outdoor MIMO Experiments with RealisticHandset and Base Station Antennas, 71st Vehicular TechnologyConference (VTC2010-Spring, 2010 IEEE

http://www.radio-electronics.com/info/cellulartelecomms/lte-long-term-evolution/lte-frequency-spectrum.php (Ian Poole)

3GPP TR 36.912 V10.0.0 Feasibility study for further advancements for E-UTRA (LTE-Advanced) (Release 10)

3GPP TR 37.976 V1.5.0 Measurement of radiated performance for MIMO andmulti-antenna reception for HSPA and LTE terminals (Release 11)

The author acknowledges his use of results provided by his colleagues DevisIellici and Vijay Nahar.

Documents

####Good_Radio SIG 24.11.11 Brian Collins -Antennas for MIMO