2017, Robert W. W. Heath Jr. Jr. Millimeter wave communication: From Origins to Disruptive Applications Professor Robert W. Heath Jr. Situation Aware Vehicular Engineering Systems Wireless Networking and Communications Group Department of Electrical and Computer Engineering The University of Texas at Austin Thanks to sponsors including the U.S. Department of Transportation through the Data-Supported Transportation Operations and Planning (D-STOP) Tier 1 University Transportation Center, the Texas Department of Transportation under Project 0-6877 entitled Communications and Radar- Supported Transportation Operations and Planning (CAR- STOP), AT&T, Huawei, Toyota ITC, Honda, NXP, Qualcomm, and the National Science Foundation
Introduction 2
Cellular networks are connecting everyone (wirelessly) TABLETS WEARABLE NETWORKS PETS! SMART PHONES 3
Future networks will connect things beyond people DRONES OFFICES HOMES VEHICLES FACTORIES CITIES 4
Wireless communication Transmitter TX Propagation channel Receiver RX Wireless systems send information using radio frequency signals 5
Frequency and wavelength Wavelength frequency c wavelength Low frequency High frequency High Frequency (HF) Amateur radio Ultra High Frequency (UHF) Cellular signals Millimeter Wave mmwave WiFi 100 m 10 cm 1 mm Wavelength Wireless communication occurs and different frequencies 6
Information rides on fluctuations of the carrier Carrier and bandwidth Fluctuations carry different frequencies Power BANDWIDTH carrier frequency More rapid fluctuations consume more bandwidth Less fluctuations, lower bandwidth More fluctuations, higher bandwidth Bandwidth is the basic resource in a communication system 7
Data rate high bandwidth low bandwidth few bits per second many bits per second 11101100100 001101000101 100110111101 low bandwidth 11101100100 001101000101 100110111101 time high bandwidth 01000101101 100011101100 011010100110 10101000100 101111000101 001110010110 time The higher the bandwidth, the higher the data rate the system can achieve 8
Wireless systems can also exploit multiple antennas TX MIMO: Multiple Input Multiple Output RX streams 11101100100 001101000101 100110111101 01000101100 111100011101 001101010001 11101100100 001101000101 100110111101 Multiple antennas enable transmission of several parallel data streams using the same frequency resources 10001011010 0100101101100 11101100100 1101110110101 001101000101 100110111101 11101100100 001101000101 100110111101 time MIMO spatial multiplexing makes better use of bandwidth 9
What influences the rate experienced by a user? MIMO spatial multiplexing gain depends on the number of antennas in the system Claude Shannon Inventor of Information Theory rate per user = (bits per second) bandwidth X MIMO # of users X spectral efficiency Depends on signal power, noise power and interference power, improves with interference cancellation Bandwidth is the easiest leverage for higher data rates
Millimeter wave spectrum Spectra below 3 GHz is packed and $$/Hz of bandwidth is huge Amateur radio AM radio Aviation commun. FM, TV Blutooth, WiFi, TV, Cellular,... DSRC, satellite, WiFi, Cellular,... Unlicensed Expected to be unlicensed LF MF HF VHF UHF SHF below 3 GHz 57 GHz 64 GHz 71 GHz Millimeter Wave 86 GHz Lots of potential spectrum available at mmwave for consumer applications currently used for backhaul or legacy systems 11
First millimeter wave experiments Jagadis Chandra Bose Transmitter antennas * Pictures from D. T. Emerson, The work of Jagadis Chandra Bose: 100 years of millimeter-wave research, IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 45, NO. 12, DECEMBER 1997 Radiation receiver First mmwave experiments were undertaken more than 100 years ago! 12
Millimeter wave band uses MmWave has a long history in sensing and communication WiFI IEEE 802.11ad 2001 Mars Odyssey Security Backhaul screening Satellite Satellite-based links remote sensing Atacama millimeter wave array (Chile) SCR-584 Radio astronomy WirelessHD* Automotive radar 2014 2009 Fire control radar 1970 Bose experiments 1895 1944 *http://www.electronicdesign.com/communications/qa-exploring-millimeter-waves-new-breed-devices MmWave has just now reached consumer applications 13
Consumer challenge #1: device size and cost Fig. 20. Simplified schematic of the 60-GHz amplifier. 0.87 x 0.70mm! Fig. 21. Micrograph of the 60-GHz amplifier. The chip area is 0.87 mm 0.70 mm. OKI 35V11 millimeter wave klystron1,2 [1] http://www.oki.com/en/130column/07.html [2] R. True, The Evolution of Microwave and Millimeter Wave Tubes, 2012 60 GHz amplifier, 20083 B. 60-GHz Amplifier Design and Measurement Results Fig. 22. solid lin S-param The amplifier is a three-stage single-ended design. A simpli[3] M.fied Varonen, Mi. and Ka rkka inen, M. Kantanen, and K. A.are I. Halonen, schematic a micrograph of the amplifier shown in Millimeter-Wave Integrated Circuits in 65-nm CMOS, IEEE Fig. 20 and Fig. 21, respectively. The chip area is 0.87 Transactions mm on Solid 0.70State mm Circuits, including2008 pads. Series CPW-lines and short-circuited and be CPW-shunt stubs are used for input, output and interstage matching. At the output, an open-ended shunt stub is needed to complete the output match to 50- impedance. As with the 40-GHz design, multiple finger capacitors are used in parallel for the on-chip short circuit and DC-blocking capacitors. The low frequency stability is ensured by resistor-capacitor networks. The preliminary measured transistor data was used in Until recently, mmwave devices were expensive, bulky, or made with expensive semiconductor processes charac eter m shown pressio hibits sured p viousl 14
Consumer challenge #2: propagation effects Penetration loss Scattering More scatterers but fewer paths TX RX Direct path blocked Diffraction not significant Propagation has not been well understood by systems engineers 15
Consumer challenge #3: antennas become too small aperture at mmwave TX aperture at a conventional cellular frequency RX Small antennas do not capture as much of the impinging wave 16
Making mmwave viable for consumers 17
Idea 1: An antenna array at the receiver fixes shrinkage TX RX highly directive reception leads to array gain Large antenna array captures the same amount of energy avoiding the misconception that losses increase with frequencies 18
Idea 2: An antenna array at the transmitter focuses energy TX highly directive transmission RX highly directive reception Beamforming at the transmitter adds additional array gain and reduces caused interference 19
The antenna arrays are small at mmwave antennas are about 10 mm (the large objects are antenna connectors, used only for prototyping) Samsung Galaxy S7* Mockup of a Galaxy with mmwave** Base station may have 64 to 512 antennas Mobile station may have 4 to 32 antennas [1] From https://www.ifixit.com/teardown/samsung+galaxy+s7+teardown/56686 [2] W. Roh et al. "Millimeter-wave beamforming as an enabling technology for 5G cellular communications: theoretical feasibility and prototype results," in Communications Magazine, IEEE, vol.52, no.2, pp.106-113, February 2014 20
Idea 3: Analog processing beamformer combiner RFain DAC RF Chain RF Chain RFain ADC Analog precoding Phase shifters DAC RF Chain RF Chain ADC Conventional MIMO DAC RF Chain RF Chain ADC DAC RF Chain RF Chain ADC Forming beams using analog components reduces the amount of RF hardware and subsequent baseband processing required 21
Idea 4: Beam training TX TX sends training sectors on each RX sector Antennas need to be adaptively pointed RX Beam training finds the best beam pair over the air 22
Commercial mmwave applications Sony wearable HDTV * Talon Multi-Band Wi-Fi Router Standard Bandwidth Rates Approval WirelessHD 2.16 GHz 3.807 Gbps Jan. 2008 WirelessHD 1.1 2.16 GHz 4 x 7.138 Gbps Jan. 2010 IEEE 802.11ad 2.16 GHz 6.76 Gbps Dec. 2012 Zyxel AeroBeam HDTV kit * Epson projector * Dell Laptop * * http://www.wirelesshd.org/consumers/product-listing/ Current standards for personal networks and WiFi support arrays and beam training 23
Bringing mmwave to 5G and beyond 24
Taking advantage of MIMO processing. + RF Chain 1-bit ADC ADC. + RF Chain 1-bit ADC ADC. + RF Chain 1-bit ADC ADC Analog precoding Hybrid precoding Hybrid precoding enables multi stream transmission with low power, but requires changes in conventional MIMO algorithms 25
Reducing resolution in data converters RF Chain 1-bit ADC 1-bit ADC Original sine wave Quantized sine wave, 4 bits Baseband Processing Baseband Precoding Original sine wave Quantized sine wave, 2 bits RF Chain 1-bit ADC 1-bit ADC Original sine wave Quantized sine wave, 1 bit Higher levels of quantization dramatically reduce power consumption, but require new algorithms that can deal with extra distortion 26
Overcoming different types of blockage Secondary reflectors, multiple paths, and active reflectors Dense deployment of infrastructure and fast switching Handset antenna diversity, warning signs, and electric shock 27
Alternatives to conventional beam training Simultaneously sampling from multiple spatial directions Exploit the fact that there are a few good paths via compressive sensing 28
Adaptive reconfiguration in high mobility Leverage out-of-band information, multi-band communication, position, sensors, and machine learning to reduce overheads during beam reconfiguration 29
Disruptive applications 30
5G cellular networks will exploit mmwave Fixed wireless access Air-to-X Vehicle-to-X Robots Wireless backhaul UE 5G @ mmwave will provide high data rate connectivity for different types of applications Backhaul 31
Vehicle-to-everything (V2X) communication Full automated driving Cloud assisted driving Intelligent navigation Traffic efficiency Cooperative collision avoidance Information society on the road High data rate connectivity is relevant for the automotive industry Low latency and Gbps data rates are not supported 10 32
Communications for aerial vehicles Aerial node Sensor fusion in disaster areas Team of UAVs Panoramic VR streaming of live video Communications relay Ambulance drone Mobile cellular infrastructure Safe navigation for package delivery Factory inspection, pipeline monitoring High data rate networking between manned and unmanned aerial Current vehicles solutions enables for A2X revolutionary do not support applications most applications 33
People (going beyond smart phones) Virtual reality: high-resolution multi-view video in real-time Wearable networks: multiple communicating devices in and around the body (>5 according to market trends) Augmented reality: real-time overlay of information High data rates are required for virtual and augmented reality and wearable networks 34
Connected robots Video cameras 3D image sensor IR camera Pressure sensor Inertial motion sensor Laser scanner Sonar Tactile sensor Central unit Radar 35
MmWave communication prototyping Sensors and communication equipment for V2X Quanergy M8 Lidar High precision GPS Cohda Wireless DSRC Questions? www.profheath.org www.utsaves.org www.wncg.org www.ece.utexas.edu State of the art research platform Area Scan Cameras RSU Phased array platforms Delphi Radar Three different type vehicles 36