James Kimery 5G: From Concept to Reality Director of Marketing, Wireless Research @jameskimery ni.com | NI CONFIDENTIAL 1
Hyper Connected Everything Data rate Power Consumption Security Capacity Coexistence Monitoring ni.com | NI CONFIDENTIAL 5
Wireless Requires Massive Platform Expansion > 10 Gbps peak rates 5G 5G 5G > 100K < 1 ms connections latency per cell ni.com | NI CONFIDENTIAL 6
Prototyping Is Critical for Algorithm Research “Experience shows that the real world often breaks some of the assumptions made in theoretical research, so testbeds are an important tool for evaluation under very realistic operating conditions ” “…development of a testbed that is able to test radical ideas in a complete, working system is crucial ” 1 NSF Workshop on Future Wireless Communication Research Platforms Enabling Advanced Wireless Research (PAWR) ni.com | NI CONFIDENTIAL 7
Approaching The Pivotal Moment Technical ideas from a huge number of sources Multiple Ideas 1982-85 GSM 5G We are here Pivotal Moment of Standardization The Pivotal Year Birth of GSM 1987 Broad Based Adoption 1988-91 Implementation by a large number of suppliers & operators Billions of Devices Source: www.gsmhistory.com/who_created-gsm ni.com 8
Proposed 3GPP Release 15 / 5G Timeline WI Rel 15 NSA “New Radio” Phase 1 R15 SA R15 RAN#79 RAN#79 RAN#75 RAN#69 RAN#72 RAN#75 Mar 18 Jun 18 Dec 17 Sep 15 Jun 16 Mar 17 RAN1 Scope / Req’s Channel Modeling RAN1 SI Evaluation of Solutions RAN1 WG Specification of Solutions ni.com | NI CONFIDENTIAL 9
RF Communications Lead User Program • Established in 2010 • Goals: Further wireless research through prototyping • Research Institutions • Academic • Industry • Over 100 research papers published ni.com | NI CONFIDENTIAL 10
5G Vectors Advanced Multi-RAT Massive MIMO mmWave Wireless Networks Enhanced PHY Dramatically increased Consistent connectivity Improve bandwidth Utilize potential of number of antenna meeting the 1000x traffic utilization through evolving extremely wide bandwidths elements on base station. demand for 5G PHY Level and flexible at frequency ranges once numerology thought impractical for • Densification commercial wireless. • SDN • NFV • CRAN ni.com | NI CONFIDENTIAL 12
NI 5G Research Initiatives Massive MIMO Wireless Networks Multi-RAT mmWave Personal Computers PXI Systems USRP RIO ni.com | NI CONFIDENTIAL 13
5G Vectors Advanced Multi-RAT Massive MIMO mmWave Wireless Networks Enhanced PHY Dramatically increased Consistent connectivity Improve bandwidth Utilize potential of number of antenna meeting the 1000x traffic utilization through evolving extremely wide bandwidths elements on base station. demand for 5G PHY Level and flexible at frequency ranges once numerology thought impractical for • Densification commercial wireless. • SDN • NFV • CRAN ni.com | NI CONFIDENTIAL 14
NI and Lund University Collaborate on Massive MIMO Goal : Build a massive MIMO prototype 100 antenna system with real time processing capabilities Challenges : System complexity 100 Synchronized Tx / Rx chains Data throughput for processing Aggregation of multiple channels Heterogeneous computation Prof Ove Edfors Prof Fredrik Tufvesson ni.com | NI CONFIDENTIAL 16
Massive MIMO Successes Facebook announces ARIES Bristol Smashes Spectral Efficiency Testbed Record 71 bits/Second/Hz >146.4 bits/Second/Hz 96 Antennas, 24 Users 128 Antennas, 22 Users (USRP RIO, LV, Custom LV Code Base) (USRP RIO, LV, NI LU Code Base) April 2016 May 2016 ni.com | NI CONFIDENTIAL 17
Hardware Architecture: 128 Base Station Antennas x 12 UE Antennas @ 20 MHz NI PCIe Switch Boxes for Data NI FlexRIOs or Atrox for NI Octoclocks for Clocking MIMO Co-processing Laptops+USRP RIO As UEs … ni.com | NI CONFIDENTIAL 19
5G Vectors Advanced Multi-RAT Massive MIMO mmWave Wireless Networks Enhanced PHY Dramatically increased Consistent connectivity Improve bandwidth Utilize potential of number of antenna meeting the 1000x traffic utilization through evolving extremely wide bandwidths elements on base station. demand for 5G PHY Level and flexible at frequency ranges once numerology thought impractical for • Densification commercial wireless. • SDN • NFV • CRAN ni.com | NI CONFIDENTIAL 21
Advanced Wireless Prototypes Texas A&M CROWD Advanced MAC SDN ABSF Network Research System NS-3 NI Lead Intel User Group CRAN-Massive Open LTEWiFi MIMO Coexistence Testbed ni.com | NI CONFIDENTIAL 22
5G Advanced Wireless Networks Research • Hyper dense networks • Software defined networking (SDN) • Cloud radio access network (cRAN) • Cellular/802.11 coexistence and co-ordination • Next generation 802.11 stack ni.com | NI CONFIDENTIAL 23
Architecture for Protocol Stack Explorations 802.11 LTE MTC IoT Open Source Upper Layer Stack (e.g. ns-3) 802.11 Ref Design LTE Ref Design PHY/MAC Stack in LabVIEW NI Hardware ni.com | NI CONFIDENTIAL 24
Real LTE OTA Experiments with NS-3 Network Simulator SGW/PGW SGW/PGW APP APP IP IP IP IP PDCP GTP PDCP GTP PDCP GTP PDCP GTP RLC RLC UDP UDP RLC RLC UDP UDP MAC MAC IP IP PHY PHY MAC MAC IP IP Emu Emu L1-L2 API L1-L2 API UE eNB DL L1-L2 API L1-L2 API PHY PHY DA/AD+RF DA/AD+RF UL UE eNB ni.com | NI CONFIDENTIAL 25
5G Vectors Advanced Multi-RAT Massive MIMO mmWave Wireless Networks Enhanced PHY Dramatically increased Consistent connectivity Improve bandwidth Utilize potential of number of antenna meeting the 1000x traffic utilization through evolving extremely wide bandwidths elements on base station. demand for 5G PHY Level and flexible at frequency ranges once numerology thought impractical for • Densification commercial wireless. • SDN • NFV • CRAN ni.com | NI CONFIDENTIAL 26
The Question of Frequency: WRC-15 Outcome • The ITU released a list of globally viable frequencies for 5G mmWave technologies ni.com | NI CONFIDENTIAL 27
FCC Embraces mmWave Further Notice of Proposed Rulemaking (FNPRM) Upper microwave flexible use Unlicensed 31.8 – 33.4 37 - 38.6 38.6 - 40 81 -86 27.5 - 28.35 64 - 71 71 -76 42 - 42.5 24.75 – 25.25 47.2 – 50.2 24.25 – 24.45 50.4 – 52.6 37 to 38.6 3 7 t o 3 8 . 6 ni.com | NI CONFIDENTIAL 28
Verizon “5G” Specification • Adaption of “LTE” OFDM technology to cmWave/mmWave frequencies • Fc = 28, 37, 39 GHz • Adds control signaling for beam forming • Applications • Fixed, broadband wireless • Mobility? • Verizon contributing to 3GPP but this is NOT the 5G specification ∆ fsc Scaling BW Slot dur. Sampling Rate TX SRC factor TX SRC Symb dur CP dur Symbols [kHz] factor [MHz] [ms] [MSps] (fs / 3072 MSps) factorization [us] [us] per slot 0,9 75 5 100 0,100 153,60 20,00 5*2*2 13,33 7 0,5 150 10 200 0,100 307,20 10,00 5*2 6,67 14 300 20 400 0,100 614,40 5,00 5 3,33 0,2 28 ni.com | NI CONFIDENTIAL 29
mmWave Prototypes and Channel Sounding NYU Wireless NTT Docomo mmWave 73 GHz Channel Sounding BW = 1 GHz Prof. Ted Rappaport AT&T Nokia Channel Sounding Multiple Prototypes ni.com | NI CONFIDENTIAL 32
Timeline w/ NI Platform Brooklyn 5G NIWeek 2015 MWC 2016 Summit 2014 Frequency 73 GHz 73 GHz 73 GHz Bandwidth 1 GHz 2 GHz 2 GHz 1x1 2x2 2x2 Streams 16 QAM 16 QAM 64 QAM Modulation 2.3 Gbps >10 Gbps >14.5 Gbps Peak rate ni.com 36
Q2 2016 NI mmWave Transceiver System The world’s first 2 GHz bandwidth mmWave transceiver covering the 71-76 GHz frequency band for channel sounding and real-time 2-way MIMO communications with coding. Applications • Cellular • Wi-Fi • Channel sounding Capabilities • 71 – 76 GHz • 2 GHz Bandwidth • MIMO • Real-time 2-way communication ni.com 38
Software Defined Radio Architecture Multi-Processor Subsystem RF Front End • General Purpose Real-time signal processor Host processor • Physical Layer (PHY) • Medium Access Control (MAC) – RF • ex FPGA, DSP • Dual LOs Rx/Txcontrol PLL • Contiguous • ex. Host GPP , multi-core CPU VCO Frequency Range D/A 0 90 D/A CPU FPGA GPP DSP A/D 0 A/D 90 PLL Host Connection Baseband VCO Determines Streaming Converters Bandwidth Ex. Gigabit E-net, PCIe ni.com 40
www.ni.com/5g www.ni.com/sdr ni.com | NI CONFIDENTIAL
Recommend
More recommend
