Key Techniques in
4G/LTE
(Oct 2011)

google.me/+SendilKumar
Outline
- Wireless Communication
- Multiuser Communication

- Key techniques in 4G/LTE
- Detecting & Latching to LTE n/w
-...
Wireless Channel : Property
Broadcast interface

• Everyone can
listen to the
medium
• One emitting
source for a
region

S...
Wireless Channel - Terrain

Large Scale : Path Loss, Shadowing
Small Scale : Stationary and Mobile objects
Sendil Kumar, T...
Wireless Channel - Impairments
Time Selective

Frequency Selective

• Relative speed of the
transmitter and receiver
• Mea...
Wireless Channel - Constraints
• Transmit power
– Both from base-station and terminal (FDD)
– TDD !!

• Bandwidth
• Divers...
Spectrum

Resource Plane & Elements

RE

RP
Duration

RP dimensions ≈ BW x Duration

Sendil Kumar, TCOE-IIT Madras

7
Signal Received at UE

Y1=H1X1 +N1

UE
RP

Channel Mask

Noise + Interference

Sendil Kumar, TCOE-IIT Madras

8
Multiplexing physical symbols

How many BITS can be
“reliably” transmitted ???
Sendil Kumar, TCOE-IIT Madras

9
• Measure of reliability of Bits
• Efficient usage of the channel

DELIVER RELIABLE BITS

Sendil Kumar, TCOE-IIT Madras

1...
Capacity of a Wireless Channel
Reliable communication rate

R  C = log2(1 + ) b/s/Hz
Post (receiver) processing SINR
Sha...
Capacity vs. SNR

Sendil Kumar, TCOE-IIT Madras

12
Tune power or rate ??
Power Control
Transmit Power

Data Rate
Rate Control

Channel Quality

Power Control
Transmit power ...
• Multiplexing multiple users’ data
• Basis of smart “picking” of users

MULTIUSER COMMUNICATION

Sendil Kumar, TCOE-IIT M...
Characteristics of Multiple access
• Static or Dynamic, equal/unequal sharing of
the resource (BW x Duration)
• Orthogonal...
Scheduler – “Key Ingredient”
• Single User
– Adapt Rate based on Channel

• Multiple Users
– Adapt Rate also select which ...
Scheduler : Other metrics
• Diversity in the users’ channel
• Queue Size of data for the user
• Quality of service
– Mean ...
Scheduling : Intra Cell
User 4
User 3
User 2
User 1

Frequency
Time

Sendil Kumar, TCOE-IIT Madras

Time Unit

18
Scheduling : Inter-cell Interference
Management (1/2)

Sendil Kumar, TCOE-IIT Madras

19
Scheduling : Inter-cell Interference
Management (2/2)
In-cell

Cell-edge UE

Fractional Frequency Reuse
Sendil Kumar, TCOE...
• More than as a 4G standard
• Flexibility to vendors, operators

KEY 4G TECHNIQUES

Sendil Kumar, TCOE-IIT Madras

21
4G systems (1/3)

Packet only
network
(always on
NET)

MIMO
capabilities

Higher order
modulation
64QAM

Efficient
Uplink
...
4G systems (2/3)
Spectral Efficiency
in reuse-1
deployment

HetNET
deployments

Inter-technology
handoff

Cell Edge
interf...
4G systems : Addons (3/3)

Green
technology

M2M
communication

Network
offloading WIFI
n/w

Sendil Kumar, TCOE-IIT Madras...
• 4G timings with 3G, CDMA2000
• Multimode devices

NUMERICAL COMPATIBILITY

Sendil Kumar, TCOE-IIT Madras

25
Numerical Compatibility
Channel Bandwidth [MHz]

1.4

3

5

10

15

20

73

181

301

601

901

1201

128

256

512

1024
...
• Symbols
• Slots, Subframe
• Frame

RESOURCE PLANE

Sendil Kumar, TCOE-IIT Madras

27
Timing definition

Shorter TTI => Fast user switching
Sendil Kumar, TCOE-IIT Madras
Slide 13

28
Time-Frequency view (DL)

Sendil Kumar, TCOE-IIT Madras

29
Physical Resources
Tslot

The grouping of resources allows
reduce the address size when
scheduling data, when indicating
c...
• Align reception with BS
• Determine Cell Info

DETECTING & LATCHING TO LTE

Sendil Kumar, TCOE-IIT Madras

31
UE procedures (1/2)

• some signature !!
• Periodic correlation
• Energy detection

Device Reads N/w
parameter (Public
Rea...
UE procedures(2/2)

• All UE UL data arrives
at same time at
eNodeB
• UL Power Control

Listen to
Control/Paging
• Decode ...
Downlink Transmission

Signals

• Pilots
• Synchronization
(Primary &
Secondary)

Control Channels

• System
Information
•...
Synchronization / Cell Search
• Frequency stability
– Within regulatory & 3GPP requirements
– Handover

• Timing Synchroni...
Cell Sync & Read Info
• Scalable Bandwidth operation
– Common Resource Blocks 72 carrier at centre
– Correlation – Energy ...
Primary & Secondary Sync Signals

Twice in a Frame duration
- Centre of the Bandwidth
- Centre of the Time (Frame & Subfra...
Sync & Broadcast channels
Immediate Symbol after SYNC - provides
necessary system information
(like MIB, overall DL BW, Sy...
• Update link Quality
• Measurement Reporting

CHANNEL ADAPTATION

Sendil Kumar, TCOE-IIT Madras

39
Flexible Channel Adaption
Feedback
– Wideband CQI
– Band based CQI
– UE selected Band CQI
– Periodic Reporting, aperiodic ...
E.g. CASES – When & What is used
Wideband CQI

Best Band CQI

• Maps to the
best MCS
supported for
distributed
mapping
• K...
HARQ

Depending on
the PRB
locations ,eB
can decide what
to do
Accumulated energy
Transmitted bits

CRC Insertion, turbo c...
Pilot/Reference Signals (RS)
• Measure Channel Response
– Along time & frequency
– Sufficient density for interpolation

•...
Pilot Symbol physical locations
1 Tx Antenna

2 Tx Antenna

What about adjacent sector
-Distance between two RS
-Freq : Ur...
Inter-antenna Pilot Interference
Management
Transmit Antennas

With Multiple Receive
Antenna – Estimate
Correlated Channel...
Adjacent Sector Pilot Interference
Management

Sector 0

Sector 1

Sector 2

Deployment Planning

Sendil Kumar, TCOE-IIT M...
Informing the UE on PDCCH
• Information : PRBs and MCS to each UE
• UE monitors the PDCCH(s) to find possible
allocation
•...
For reliability
For capacity

MIMO TECHNIQUES

Sendil Kumar, TCOE-IIT Madras

48
MIMO Abilities – Improve SNR

Diversity for improved
system performance

Beamforming for improved
coverage

Rx : Equalizat...
MIMO Abilities – Increase Capacity

Beamforming for improved
coverage

SDMA for improved capacity

Capacity !! :
a) Smart ...
MIMO Data Mapping
Turbo Coded
Block : CW1
Turbo Coded
Block : CW2

K = 1,2 Code Words

M = 1,2,3,4 Layers

N = 1,2, or 4 A...
Y1=H1X1 + H2X2 + H3X3 + H1X4

4 Tx antenna

UE
Receive Antennas
Sendil Kumar, TCOE-IIT Madras

52
Single Antenna Transmission
Y1=H1X1 + H2X2 + H3X3 + H1X4

4 Tx antenna

UE
Receive Antennas
Sendil Kumar, TCOE-IIT Madras
...
Open Loop Transmit Diversity
X1

A1,A2=fn(X1)

Y1=H1A1 + H2A2 + H3X3 + H1X4

4 Tx antenna
UE
Receive Antennas
Sendil Kumar...
BEAMFORMING
A1=P1X1 A2=P2X1

X1

Y1=H1A1 + H2A2 + H3X3 + H1X4

4 Tx antenna
Receive Antennas
Feedback of P1, P2
Sendil Kum...
Closed Loop , Two Streams
(Spatial MUX)

X2
X1
Y1=H1A1 + H2A2 + H3X3 + H1X4
Y2=G1A1 + G2A2 + G3X3 + G1X4

4 Tx antenna
Fee...
Transmission Modes

Open Loop
• Mode 1: Single antenna port
• Mode 2: Tx Diversity (Alamoutilike )
• Mode 3: Spatial multi...
Physical Resource Planes

Data Layer X1,X2, X3..
Phy Ant 1 : RP

Precoding based on
CodeBook (Looktable )
Closed Loop
Or
O...
MIMO Implementation
A very flexible and Vendor specific Implementation ,
- Cell Capacity
- Multiple UE serving
- Slow mobi...
BACKUP SLIDES

Sendil Kumar, TCOE-IIT Madras

60
Upcoming SlideShare
Loading in …5
×

Brief presentation about Key Techniques in 4G/LTE

1,412 views

Published on

Brief presentation about Key Techniques in 4G/LTE and where there is space for proprietary way of tuning it.

Note it was presented about 2years back !!

Published in: Technology, Business
0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
1,412
On SlideShare
0
From Embeds
0
Number of Embeds
8
Actions
Shares
0
Downloads
65
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide

Brief presentation about Key Techniques in 4G/LTE

  1. 1. Key Techniques in 4G/LTE (Oct 2011) google.me/+SendilKumar
  2. 2. Outline - Wireless Communication - Multiuser Communication - Key techniques in 4G/LTE - Detecting & Latching to LTE n/w - Channel Adaptation Techniques - MIMO techniques - Miscellaneous Developments
  3. 3. Wireless Channel : Property Broadcast interface • Everyone can listen to the medium • One emitting source for a region Shared resource • All share the resources • Need resource management to optimize the best utilization Sendil Kumar, TCOE-IIT Madras Dynamic properties compared to wired medium • Adapt to the nature of the channel • Need feedback of the user link quality 3
  4. 4. Wireless Channel - Terrain Large Scale : Path Loss, Shadowing Small Scale : Stationary and Mobile objects Sendil Kumar, TCOE-IIT Madras 4
  5. 5. Wireless Channel - Impairments Time Selective Frequency Selective • Relative speed of the transmitter and receiver • Measured in terms Doppler frequency • Multipath channel • Measured in terms of the delay spread of the channel Sendil Kumar, TCOE-IIT Madras 5
  6. 6. Wireless Channel - Constraints • Transmit power – Both from base-station and terminal (FDD) – TDD !! • Bandwidth • Diverse topography Sendil Kumar, TCOE-IIT Madras 6
  7. 7. Spectrum Resource Plane & Elements RE RP Duration RP dimensions ≈ BW x Duration Sendil Kumar, TCOE-IIT Madras 7
  8. 8. Signal Received at UE Y1=H1X1 +N1 UE RP Channel Mask Noise + Interference Sendil Kumar, TCOE-IIT Madras 8
  9. 9. Multiplexing physical symbols How many BITS can be “reliably” transmitted ??? Sendil Kumar, TCOE-IIT Madras 9
  10. 10. • Measure of reliability of Bits • Efficient usage of the channel DELIVER RELIABLE BITS Sendil Kumar, TCOE-IIT Madras 10
  11. 11. Capacity of a Wireless Channel Reliable communication rate R  C = log2(1 + ) b/s/Hz Post (receiver) processing SINR Shannon-Hartley capacity relation Sendil Kumar, TCOE-IIT Madras 11
  12. 12. Capacity vs. SNR Sendil Kumar, TCOE-IIT Madras 12
  13. 13. Tune power or rate ?? Power Control Transmit Power Data Rate Rate Control Channel Quality Power Control Transmit power inversely proportional to channel conditions Good for circuit switched mode of communication Rate Control Transmit at fixed power level Good for data traffic , Long term data rate Sendil Kumar, TCOE-IIT Madras 13
  14. 14. • Multiplexing multiple users’ data • Basis of smart “picking” of users MULTIUSER COMMUNICATION Sendil Kumar, TCOE-IIT Madras 14
  15. 15. Characteristics of Multiple access • Static or Dynamic, equal/unequal sharing of the resource (BW x Duration) • Orthogonality within intra-cell downlink • Control overhead with increasing number of users Sendil Kumar, TCOE-IIT Madras 15
  16. 16. Scheduler – “Key Ingredient” • Single User – Adapt Rate based on Channel • Multiple Users – Adapt Rate also select which user(s) in a TTI • Multi-cell – Co-operation with adjacent cells for Interference Management – Resource partition with small cells Sendil Kumar, TCOE-IIT Madras 16
  17. 17. Scheduler : Other metrics • Diversity in the users’ channel • Queue Size of data for the user • Quality of service – Mean delay at upper layers – Reliability of the data • Maintain the system capacity by efficiently sharing resources Sendil Kumar, TCOE-IIT Madras 17
  18. 18. Scheduling : Intra Cell User 4 User 3 User 2 User 1 Frequency Time Sendil Kumar, TCOE-IIT Madras Time Unit 18
  19. 19. Scheduling : Inter-cell Interference Management (1/2) Sendil Kumar, TCOE-IIT Madras 19
  20. 20. Scheduling : Inter-cell Interference Management (2/2) In-cell Cell-edge UE Fractional Frequency Reuse Sendil Kumar, TCOE-IIT Madras 20
  21. 21. • More than as a 4G standard • Flexibility to vendors, operators KEY 4G TECHNIQUES Sendil Kumar, TCOE-IIT Madras 21
  22. 22. 4G systems (1/3) Packet only network (always on NET) MIMO capabilities Higher order modulation 64QAM Efficient Uplink transmission Sendil Kumar, TCOE-IIT Madras Using wide bandwidth ~ 20MHz Scalable bandwidth 22
  23. 23. 4G systems (2/3) Spectral Efficiency in reuse-1 deployment HetNET deployments Inter-technology handoff Cell Edge interference management Pico cells UMTS/HSDPA Improve mean cell-throughput Portable basestations (Femtocell, Relays) EVDO (xHRPD) Co-operative communication Remote Radio Heads 1xCDMA voice Sendil Kumar, TCOE-IIT Madras Proprietary techniques : Resource Partitioning, Handoff between Macro/HETNET, inter-RAT mobility 23
  24. 24. 4G systems : Addons (3/3) Green technology M2M communication Network offloading WIFI n/w Sendil Kumar, TCOE-IIT Madras Self-Organizing Network (SON) 24
  25. 25. • 4G timings with 3G, CDMA2000 • Multimode devices NUMERICAL COMPATIBILITY Sendil Kumar, TCOE-IIT Madras 25
  26. 26. Numerical Compatibility Channel Bandwidth [MHz] 1.4 3 5 10 15 20 73 181 301 601 901 1201 128 256 512 1024 1536 2048 Sampling Rate [MHz] 1.92 ½ 3.84 3.84 7.68 2x3.84 15.36 4x3.84 23.04 6x3.84 30.72 8x3.84 No. of Resource Blocks (NRB) 6 15 25 50 75 100 No. of Occupied Subcarriers including DC (NSC) FFT Size (N) EVDO HSDPA 3G - HSDPA : 5MHz : 2ms Subframe 2560*3 /2= 3840 chips /1ms= 240 symbols /ms LTE 5MHz : 600*14 =8400 symbols in 1ms Subframe Sendil Kumar, TCOE-IIT Madras 26
  27. 27. • Symbols • Slots, Subframe • Frame RESOURCE PLANE Sendil Kumar, TCOE-IIT Madras 27
  28. 28. Timing definition Shorter TTI => Fast user switching Sendil Kumar, TCOE-IIT Madras Slide 13 28
  29. 29. Time-Frequency view (DL) Sendil Kumar, TCOE-IIT Madras 29
  30. 30. Physical Resources Tslot The grouping of resources allows reduce the address size when scheduling data, when indicating control information Resource element Resource block (180 KHz x 0.5 ms) 12x7 (subcarrier x symbols) Resource Block Group frequency time Sendil Kumar, TCOE-IIT Madras 30
  31. 31. • Align reception with BS • Determine Cell Info DETECTING & LATCHING TO LTE Sendil Kumar, TCOE-IIT Madras 31
  32. 32. UE procedures (1/2) • some signature !! • Periodic correlation • Energy detection Device Reads N/w parameter (Public Read) • Sync time & frequency • slot alignment • System frame sequence Device Discovers N/w Sendil Kumar, TCOE-IIT Madras • Random Access • Collision • N/w detects • Reliable channel to listen Try to push request to “Associate” 32
  33. 33. UE procedures(2/2) • All UE UL data arrives at same time at eNodeB • UL Power Control Listen to Control/Paging • Decode Data if scheduled • Feedback of UL data Timing Alignment • Quality of Channel (CQI) • Supported Data rate • ACK/NACK of DL data Report Measurement Sendil Kumar, TCOE-IIT Madras 33
  34. 34. Downlink Transmission Signals • Pilots • Synchronization (Primary & Secondary) Control Channels • System Information • Scheduling • Power control, ACK/NACK Sendil Kumar, TCOE-IIT Madras Shared Data Channels • Data Traffic • Multicast Traffic 34
  35. 35. Synchronization / Cell Search • Frequency stability – Within regulatory & 3GPP requirements – Handover • Timing Synchronization – UL and DL associated transmission Sendil Kumar, TCOE-IIT Madras 35
  36. 36. Cell Sync & Read Info • Scalable Bandwidth operation – Common Resource Blocks 72 carrier at centre – Correlation – Energy Maximation • Slot Boundary & Frame boundary • Read Broad Cast channel “Proprietary” way • Determine 1/504 cell ID of doing this Esp. in multi-mode devices Sendil Kumar, TCOE-IIT Madras 36
  37. 37. Primary & Secondary Sync Signals Twice in a Frame duration - Centre of the Bandwidth - Centre of the Time (Frame & Subframe) (5 ms interval) to look for other RAT Sendil Kumar, TCOE-IIT Madras 37
  38. 38. Sync & Broadcast channels Immediate Symbol after SYNC - provides necessary system information (like MIB, overall DL BW, Sys Frame Number) Step. A Detect 1 of 3 Primary Code Step. B Detect 1 of 168 using Sec Sync 12*6 = 72 subcarriers 1.08MHz Broadcast Info Primary SYNC BS :Deployment can help quick cell search UE: ability to demap BCH faster Secondary SYNC Sendil Kumar, TCOE-IIT Madras Slide 24 TV80-W1738-1 Rev B 38 MAY CONTAIN U.S. AND INTERNATIONAL EXPORT CONTRO
  39. 39. • Update link Quality • Measurement Reporting CHANNEL ADAPTATION Sendil Kumar, TCOE-IIT Madras 39
  40. 40. Flexible Channel Adaption Feedback – Wideband CQI – Band based CQI – UE selected Band CQI – Periodic Reporting, aperiodic reporting – Multiple code-block transmission Scheduling – Dynamic – Semi persistent – Localized, Distributed Sendil Kumar, TCOE-IIT Madras 40
  41. 41. E.g. CASES – When & What is used Wideband CQI Best Band CQI • Maps to the best MCS supported for distributed mapping • Kind of fall-back metric if eNodeB cant grant the best to the UE • Group RBs, report CQI for each/best Group • Frequency selective channels eNodeB triggered bands only • e.g restricted bands available • FFR zones GSM BW : 200KHz, 1 slot = 0.58ms (4.615ms/8) vs 1 RB-pair : 180Khz for 1ms subframe Sendil Kumar, TCOE-IIT Madras 41
  42. 42. HARQ Depending on the PRB locations ,eB can decide what to do Accumulated energy Transmitted bits CRC Insertion, turbo coding; Rate 1/3 E 2E 3E 4E Input to decoder Net code rate IR 1 1/2 1/3 1/3 CC Net code rate 1 1 Sendil Kumar, TCOE-IIT Madras 1 1 42
  43. 43. Pilot/Reference Signals (RS) • Measure Channel Response – Along time & frequency – Sufficient density for interpolation • Distinguish multiple cells – Orthogonal sequence across the 504 Smart Planning of cells with Reusing cells sequences !! – High Auto-correlation &searchcross-correlation For quick Zero • 168*3 = 504 sequences Sendil Kumar, TCOE-IIT Madras 43
  44. 44. Pilot Symbol physical locations 1 Tx Antenna 2 Tx Antenna What about adjacent sector -Distance between two RS -Freq : Urban / Mutlipath Env. Pilots & More antennas ?? -Maximum speed -Pilots Orthogonal by locations -No Pilot to Pilot interferenceKumar, TCOE-IITcell Sendil in same Madras 44
  45. 45. Inter-antenna Pilot Interference Management Transmit Antennas With Multiple Receive Antenna – Estimate Correlated Channel Response !! Receive Antennas UE Sendil Kumar, TCOE-IIT Madras 45
  46. 46. Adjacent Sector Pilot Interference Management Sector 0 Sector 1 Sector 2 Deployment Planning Sendil Kumar, TCOE-IIT Madras 46
  47. 47. Informing the UE on PDCCH • Information : PRBs and MCS to each UE • UE monitors the PDCCH(s) to find possible allocation • No. of allocation & size of allocation • Control Channel Address by CCE : 9*(4 RES) • Power control, uplink grant, etc.. 3 to 4 OFDM symbols fn of BW 4/14 = 28% Sendil Kumar, TCOE-IIT Madras 47
  48. 48. For reliability For capacity MIMO TECHNIQUES Sendil Kumar, TCOE-IIT Madras 48
  49. 49. MIMO Abilities – Improve SNR Diversity for improved system performance Beamforming for improved coverage Rx : Equalization , Tx : Open Loop (Alamouti like) Low mobility or High Mobility users Transmit “rotates” the signal for better signal reception at the UE Codebook based, non-codebook based Sendil Kumar, TCOE-IIT Madras 49
  50. 50. MIMO Abilities – Increase Capacity Beamforming for improved coverage SDMA for improved capacity Capacity !! : a) Smart Pick two users who can tolerate/cancel interference on the same Resource b) Artifical channel for the interfering data Sendil Kumar, TCOE-IIT Madras Multi-layer transmission for High data-rate Capacity !! : Assign two Streams to the same user 50
  51. 51. MIMO Data Mapping Turbo Coded Block : CW1 Turbo Coded Block : CW2 K = 1,2 Code Words M = 1,2,3,4 Layers N = 1,2, or 4 Antennas Sendil Kumar, TCOE-IIT Madras 51
  52. 52. Y1=H1X1 + H2X2 + H3X3 + H1X4 4 Tx antenna UE Receive Antennas Sendil Kumar, TCOE-IIT Madras 52
  53. 53. Single Antenna Transmission Y1=H1X1 + H2X2 + H3X3 + H1X4 4 Tx antenna UE Receive Antennas Sendil Kumar, TCOE-IIT Madras 53
  54. 54. Open Loop Transmit Diversity X1 A1,A2=fn(X1) Y1=H1A1 + H2A2 + H3X3 + H1X4 4 Tx antenna UE Receive Antennas Sendil Kumar, TCOE-IIT Madras 54
  55. 55. BEAMFORMING A1=P1X1 A2=P2X1 X1 Y1=H1A1 + H2A2 + H3X3 + H1X4 4 Tx antenna Receive Antennas Feedback of P1, P2 Sendil Kumar, TCOE-IIT Madras UE 55
  56. 56. Closed Loop , Two Streams (Spatial MUX) X2 X1 Y1=H1A1 + H2A2 + H3X3 + H1X4 Y2=G1A1 + G2A2 + G3X3 + G1X4 4 Tx antenna Feedback of P by Hypothesis Sendil Kumar, TCOE-IIT Madras UE Receive Antennas 56
  57. 57. Transmission Modes Open Loop • Mode 1: Single antenna port • Mode 2: Tx Diversity (Alamoutilike ) • Mode 3: Spatial multiplexing • Transmissions using spatial multiplexing (2 CW). (RI , - ) • Exploit Cyclic Delay Diversity (CDD) transmissions Closed Loop • Mode 4: Spatial multiplexing (SU- MIMO) : (RI, PMI), 2CW, 4Layers, 4Tx • Mode 5: Multi-user MIMO • CL MU-MIMO / (SDMA) • Mode 6: ( ~ Mode4 : (-,-) ) Rank=1 precoding • Mode 7: Single antenna port; port 5 ( ~Mode1 : DRS) Sendil Kumar, TCOE-IIT Madras 57
  58. 58. Physical Resource Planes Data Layer X1,X2, X3.. Phy Ant 1 : RP Precoding based on CodeBook (Looktable ) Closed Loop Or Open Loop Phy Ant 0 : RP Sendil Kumar, TCOE-IIT Madras 58
  59. 59. MIMO Implementation A very flexible and Vendor specific Implementation , - Cell Capacity - Multiple UE serving - Slow mobile terminals, - Rural / Urban channels Sendil Kumar, TCOE-IIT Madras 59
  60. 60. BACKUP SLIDES Sendil Kumar, TCOE-IIT Madras 60

×