Massive MIMO a overview. Chandrasekaran CEWiT

Transcription

1 Massive MIMO a overview Chandrasekaran CEWiT

2 Outline Introduction Ways to Achieve higher spectral efficiency Massive MIMO basics Challenges and expectations from Massive MIMO Network MIMO features Summary

3 Wireless traffic growth Growth in Wireless Traffic One Million times in last 45 years Martin Cooper s law The number of simultaneous voice/data connections has doubled every 2.5 years (+32% per year) since the beginning of wireless Src: Martin Cooper

4 Spectrum and Limitations Radio spectrum is a very scarce resource Limited availability for cellular communications Spectral efficiency needs to be improved Spectral efficiency of point-to-point transmission Shannon s capacity limit log SINR Cannot do much: 4 bps/hz -> 8 bps/hz costs 17 times more power

5 Spectral Efficiency Frequency reuse Dense deployments using smaller cells Inter-cell interference Diminishing returns with increase in number of smaller cells MIMO Capacity increasing linearly with factor of min{nt, Nr} SM for SU-MIMO Limited by number of antennas at the user in cellular frequency SDMA for MU-MIMO Larger number of antennas from distributed user terminals

6 Conventional MU-MIMO Performance depends on scheduler and link adaptation CSI feedback from UE report (FDD mode) Precoder design using uplink channel (Reciprocity principle for TDD)

7 MU-MIMO in LTE MU-MIMO MU-MIMO (TM5) Dual layer DMRS based SM (TM8) 8 layer DMRS based SM (TM9) Code book based scheme Maximum 2 user pairing Single layer transmission to each UE Non Code book based scheme Adaptive SU/MU MIMO More user pairing Higher rank transmission to each UE

8 FD-MIMO in LTE Release 13 Vertical sectorization Creation of vertical sectors Just like having multiple sectors in the horizontal direction Beam formed CSIRS based scheme Virtual sectorization using beam selection Kronecker based precoding Vertical and horizontal precoder reporting from UE Forming the final precoder using KP SRS based scheme (TDD) Precoder selection using uplink channel Reciprocity property

9 Massive MIMO Hundreds of BS antennas Tens of active users Higher order improved MU-MIMO MU-MIMO Scheduler challenges User pairing algorithm MU-CQI prediction assuming co-user interference

10 Active Antenna Array (AAS) Antenna element Antenna element Src : TR AAS Structure Sub array TXRU model Full connection TXRU model

11 Massive MIMO deployment Linear Rectangular array Cylindrical

12 Virtual sectorization Conventional beamforming in horizontal direction 3D beamforming for single UE 3D beamforming for multiple UEs

13 Virtual beamforming Active Antenna System (AAS) Elevation beamforming Horizontal beamforming Base station Beamforming using user location

14 Challenges of Massive MIMO UE specific beamforming Cell wide coverage for broadcast/control channels To achieve cell wide coverage for broadcast /control channels Narrow beam for data channels CQI estimation for the UE specific beam

15 Further challenges Feedback and codebook design in FDD How to reduce the feedback overhead? How FDD reciprocity can be used? Uplink sounding in TDD How to accurately estimate a large number of channels? Channel estimation complexity Precoding, Scheduling & Link adaptation CQI prediction and scheduling Beam identification With and without UE location information Antenna grouping High mobility scenarios Channel ageing effects Need for better diversity scheme Pilot contamination It limits MU-MIMO performance Coordination between BSs is needed

16 Network MIMO Backhaul network Coordinated transmission from multiple base stations a.k.a. CoMP Fast backhaul is a challenge Improves area spectral efficiency, system capacity

17 Summary Massive/ FD MIMO is a promising technology to significantly improve cellular capacity Pilot design, Channel estimation, Precoder estimation are the main challenges Wider coverage for control and broadcast channels using larger MIMO Design of better Diversity schemes for high mobility and low SINR users

18 Thank you