Report

2013 National Instruments Week Smart Grid Communications Prof. Brian L. Evans Dept. of Electrical & Computer Engineering Wireless Networking & Communications Group The University of Texas at Austin 7 August 2013 In collaboration with Ms. Jing Lin, Mr. Yousof Mortazavi, Mr. Marcel Nassar & Mr. Karl Nieman at UT Mr. Mike Dow & Dr. Khurram Waheed at Freescale Semiconductor (Austin) Dr. Anuj Batra, Dr. Anand Dabak & Dr. Il Han Kim at Texas Instruments (Dallas) Dr. Doug Kim, Mr. James Kimery, Mr. Mike Trimborn and Dr. Ian Wong (NI) http://users.ece.utexas.edu/~bevans/projects/plc/index.html Austin, Texas USA ISTOCKPHOTO.COM/© SIGAL SUHLER MORAN Outline • Smart power grids • Powerline noise Types Modeling • Receiver design • Testbeds • Conclusion IEEE Signal Processing Magazine Special Issue on Signal Processing Techniques for the Smart Grid, September 2012. 1 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Smart Grid Wind farm HV-MV Transformer Central power plant Grid status monitoring Utility control center Smart meters Integrating distributed energy resources Houses Offices Device-specific billing Automated control for smart appliances Medium Voltage (MV) 1 kV – 33 kV three phase Industrial plant High Voltage (HV) 33 kV – 765 kV three phase 2 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Smart Grid Goals • Improve asset utilization and operating efficiencies Reduce peak load (generation cost 30x vs. average load) Reduce excess power generation (12% margin in US) Accommodate all energy sources (renewable, storage) Scale grid voltage with energy demand • Smart meter communications Communicate grid load snapshots to utility for analysis Enable reduction of peak demand (e.g. duty cycling AC and scaling billing rate) Monitor power quality 75M smart meters sold in 2011 Disconnect/reconnect remotely EU goal of 80% smart meter Notify outage/restoration event deployments by 2020 Enable informed customer participation Source: Jerry Melcher, IEEE Smart Grid Short Course, 22 Oct. 2011, Austin TX USA 3 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Smart Meter Communications Communication backhaul carries traffic between concentrator and utility on wired or wireless links Local utility Data concentrator Low voltage (LV) under 1 kV single phase Smart meters MV-LV transformer Smart meter communications between smart meters and data concentrator via powerline or wireless links Home area data networks connect appliances, EV charger and smart meter via powerline or wireless links 4 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Wireless Smart Meter Communications Category Band Bit Rates Coverage Enables Standards Meter to customer 2.4 GHz Up to 250 kbps 100m Customer participation • 802.11b/g • 802.15.4 (ZigBee) Meter to concentrator 900 MHz Up to 250 kbps 1000m Smart meter • 802.11ah (draft) communication • 802.15.4g Concentrator to utility 900 MHz Up to 800 kbps 1000m Smart meter • 802.11ah (draft) communication • 802.15.4g • Use orthogonal frequency division multiplexing (OFDM) • Communication challenges Channel distortion IEEE 802.15.4g will likely initially use frequency shift keying (FSK) Non-Gaussian noise/interference in unlicensed bands 5 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Powerline Communications (PLC) for Smart Meters Category Band Bit Rates Coverage Enables Standards Broadband 1.8-250 MHz Up to 200 Mbps <1500 m Narrowband 3-500 kHz Up to 800 kbps • PRIME, G3 MultiSmart meter • ITU-T G.990x kilometer communication • IEEE P1901.2 • HomePlug Home area • ITU-T G.996x data networks • IEEE P1901 • Use orthogonal frequency division multiplexing (OFDM) • Communication challenges Channel distortion Non-Gaussian noise/interference 6 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Narrowband PLC Transceiver Design • Periodic bursty transmission of customer load profile Once every 15 minutes is common today and up to once every minute in future Uses carrier sense multiple access (CSMA) to see if medium is available • OFDM transmission Most of transmission band unusable Pilot tones, and null tones on band edges and unused tones • Channel modeling [Nassar12mag] Transfer functions – include effect of MV-LV transformer for US and Brazil Additive noise/interference – impulsive noise up to 40 dB higher than thermal • Global synchronization to AC main frequency (50 or 60 Hz) • PLC modem should consume less power than small light bulb (30W) Low power consumption should enable large-scale deployments Largest power consumption in power amplifier for transmission (e.g. 10V / 1.5A) 7 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Types of Powerline Noise Cyclostationary Impulsive Noise Background Noise Asynchronous Impulsive Noise -50 -100 -150 0 100 200 300 Frequency (kHz) 400 time 500 Spectrally shaped noise with 1/f spectral decay Periodic: Synchronous and asynchronous to half AC cycle Random impulsive bursts micro to milliseconds long Superposition of low intensity noise sources Switching power supplies and rectifiers Circuit transient noise and uncoordinated interference Present in all PLC Dominant in Narrowband PLC Dominant in Broadband PLC 8 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Periodic Noise from DC-DC Buck Converter • Spectrum has first peak at twice main AC frequency • Harmonics at multiples of MOSFET switching frequency (16.9 kHz) Buck converter Resulting noise has periodicities in time domain at 120 Hz &16.9 kHz 9 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Periodic Noise from DC-DC Buck Converter Time-domain voltage output ripple varies periodically at 120 Hz Impulsive noise at switching transients Note: DC value has been filtered out Zoom in to see 16.9 kHz component 10 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Cyclostationary Impulsive Noise Medium Voltage Site Segment: 1 2 3 Low Voltage Site Segment: 1 2 3 Period is one half of AC cycle Field measurements collected jointly with Aclara and Texas Instruments near St. Louis, Missouri USA 11 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Cyclostationary Impulsive Noise Modeling Measurement data from UT/TI field trial Cyclostationary Gaussian Model Proposed model uses three filters [Katayama06] [Nassar12] Demux Period is one half of an AC cycle s[k] is zero-mean Gaussian noise Adopted by IEEE P1901.2 narrowband PLC standard 12 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Asynchronous Impulsive Noise Modeling • Additive interference from multiple sources Interference from source i Assume source emissions are modeled by Poisson distribution Attenuation g(d) = exp(-a(f) d) where d is distance Homogeneous network li = l, mi = m Ex. Semi-urban areas, apartment complexes Middleton class A Ex. Dense urban and commercial settings Gaussian mixture model General (heterogeneous) network li, mi Middleton Class A is a special case of Gaussian mixture model 13 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Asynchronous Noise Model Fitting Homogeneous PLC Network General PLC Network Tail probabilities (which direct relate to communication performance) Models also work for additive uncoordinated wireless interference Middleton Class A for a Wi-Fi receiver in a Wi-Fi hotspot Gaussian mixture model for a Wi-Fi receiver in a cluster of Wi-Fi hotspots 14 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion OFDM Systems in Impulsive Noise • FFT in receiver spreads impulsive energy over all tones Signal-to-noise ratio (SNR) in each subchannel decreases • Narrowband PLC systems operate over -5 dB to 5 dB in SNR Data subchannels carry same number of bits (1-4) in current standards Each 3 dB increase in SNR on data subchannels could give extra bit 15 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Mitigating Impulsive Noise in OFDM Systems • A linear system with Gaussian disturbance v y = Fe FHF * x Fn = Fe v, g v ~ CN (x, 2 I ) Estimate the impulsive noise and remove it from the received signal yˆ = y Feˆ x g Then apply standard OFDM decoder as if only AWGN were present 16 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Proposed Non-Parametric Receiver Methods • Exploit sparsity of impulsive noise in time domain time Build statistical model each OFDM symbol using sparse Bayesian learning (SBL) At receiver, null tones contain only additive noise (Gaussian + impulsive) • SNR gain vs. conventional OFDM systems at bit error rate 10-4 Complex OFDM, 128-point FFT, QPSK, data tones 33-104, rate ½ conv. code Test SBL algorithms using additive three-term Gaussian mixture model (GMM) noise and Middleton Class A (MCA) noise with A = 0.1 and = 0.01 System Uncoded Coded Noise SBL w/ null tones SBL w/ all tones SBL w/ decision feedback GMM 8 dB 10 dB - MCA 6 dB 7 dB - GMM 2 dB 7 dB 9 dB MCA 1.75 dB 6.75 dB 8.75 dB Every SNR gain of 3 dB could mean +1 bit/tone 17 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Time Domain Interleaving Bursts span consecutive OFDM symbols Coded performance in cyclostationary noise Interleave Bursts spread over many OFDM symbols PLC standards use frequency-domain interleaving Complex OFDM, 128-point FFT, QPSK, data tones 33-104, rate ½ conv. Code Burst duty cycle of 30% 18 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Testbed #1: Built on Previous DSL Testbed • Adaptive signal processing algorithms for bit loading and interference mitigation Hardware Software • NI x86 controllers stream data • Transceiver algorithms in C on x86 • NI cards generates/receives analog signals • Desktop LabVIEW configures system • TI front end couples to power line and visualizes results 1x1 Testbed 19 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Testbed #2: Noise Playback/Analysis • G3 link using two Freescale G3 PLC modems • Freescale software tools allow frame-by-frame analysis • Test setup allows synchronous noise injection into power line Freescale PLC G3-OFDM Modem • One modem to sample Freescale PLC Testbed powerline noise in field • Collected 16k 16-bit 400 kS/s at each location 20 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Testbed #2: Cyclic Power Line Noise 10 subcarrier 50 0 -10 40 -20 -30 30 -40 10 20 30 40 OFDM Symbol 50 subcarrier 50 noise power vs avg (dB) • Analyzed cyclic properties of PLC noise measurements • Developed cyclic bit loading method for transmitter D8PSK DQPSK DBPSK 40 1. Receiver measures noise power over half AC cycle 2. Feedback modulation map to transmitter 3. Allocate more bits in higher SNR subchannels ROBO 30 NONE 10 20 30 40 OFDM Symbol 50 2x increase in bit rate Won ISPLC 2013 Best Paper Award 21 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Testbed #3: FPGA Implementation • Built NI/LabVIEW testbed with real-time link (G3 PLC settings) • Redesigned parametric impulsive noise mitigation algorithm Based on approximate message passing (AMP) framework Converted matrix operations to distributed calculations on scalars • Mapped transceiver to fixed-point data/arithmetic using Matlab • Synthesized NI LabVIEW DSP Diagram onto Xilinx Vertex 5 FPGAs SNR gain of up to 8 dB Received QPSK constellation at equalizer output conventional receiver with AMP Utilization FPGA Trans. Rec. AMP+Eq 1 2 3 total slices 32.6% 64.0% 94.2% slice reg. 15.8% 39.3% 59.0% slice LUTs 17.6% 42.4% 71.4% DSP48s 2.0% 7.3% 27.3% blockRAMs 7.8% 18.4% 29.1% 22 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Conclusion • Powerline communication systems are interference limited • Statistical models powerline interference Cyclostationary model is synchronous with zero crossings of AC cycle Gaussian mixture model is for asynchronous impulsive noise • Interference mitigation algorithms give up to 10 dB of SNR gain Non-parametric sparse Bayesian learning algorithms do not map well to FPGAs Parametric distributed approximate message algorithms map well to FPGAs • Future research for smart meter communications Use diversity of powerline and wireless links to data concentrator Maintain minimum quality link under extreme conditions Reduce power consumption in transmitter front end Project Web site: http://users.ece.utexas.edu/~bevans/projects/plc/index.html 23 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion References • • • • • • • • • • [Caire08] G. Caire, T.Y. Al-Naffouri, and A.K. Narayanan. Impulse noise cancellation in OFDM: an application of compressed sensing. Proc. IEEE Int. Symp. Information Theory, pages 1293–1297, 2008. [Cho04] J. H. Cho. Joint transmitter and receiver optimization in additive cyclostationary noise. IEEE Trans. on Information Theory, vol. 50, no. 12, 2004. [Garcia07] R. Garcia, L. Diez, J.A. Cortes, and F.J. Canete. Mitigation of cyclic short-time noise in indoor power-line channels. Proc. IEEE Int. Symp. Power Line Comm. and Its Applications, pp. 396–400, 2007. [Haring02] J. Haring. Error Tolerant Communication over the Compound Channel. Aachen, 2002. [Haring03] J. Haring and A. J. H. Vinck. Iterative decoding of codes over complex numbers for impulsive noise channels. IEEE Trans. on Information Theory, 49(5):1251–1260, 2003. [Katayama06] M. Katayama, T. Yamazato, and H. Okada. A mathematical model of noise in narrowband power line communication systems. IEEE J. Sel. Areas in Commun., vol. 24, no 7, pp. 1267-1276, 2006. [Lampe11] L. Lampe. Bursty impulse noise detection by compressed sensing. Proc. IEEE Int. Symp. Power Line Commun. and Appl., pages 29–34, 2011 [Liano11] A. Liano, A. Sendin, A. Arzuaga, and S. Santos. Quasi-synchronous noise interference cancellation techniques applied in low voltage PLC. Proc. IEEE Int. Symp. Power Line Comm. and Its Applications, 2011. [Lin11] J. Lin, M. Nassar, and B. L. Evans, “Non-Parametric Impulsive Noise Mitigation in OFDM Systems Using Sparse Bayesian Learning”, Proc. IEEE Int. Global Comm. Conf., 2011. [Lin12] J. Lin and B. L. Evans, “Cyclostationary Noise Mitigation in Narrowband Powerline Communications”, Proc. APSIPA Annual Summit and Conf., 2012. 24 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion References • • • • • • • • • [Nassar09] M. Nassar, K. Gulati, M. DeYoung, B.L. Evans, and K. Tinsley. Mitigating near-field interference in laptop embedded wireless transceivers. Journal of Signal Proc. Systems, pp. 1–12, 2009. [Nassar11] M. Nassar and B.L. Evans. Low Complexity EM-based Decoding for OFDM Systems with Impulsive Noise. In Proc. Asilomar Conf. on Sig., Systems, and Computers, 2011. . [Nassar12] M. Nassar, A. Dabak, I.H. Kim, T. Pande, and B.L. Evans. Cyclostationary noise modeling in narrowband powerline communication for smart grid applications. Proc. IEEE Int. Conf. on Acoustics, Speech and Sig. Proc., pages 3089–3092, 2012. [Nassar12mag] M.Nassar, J.Lin, Y. Mortazavi, A.Dabak, I.H.Kim and B.L.Evans, “Local Utility Powerline Communications in the 3-500 kHz Band: Channel Impairments, Noise, and Standards”, IEEE Signal Processing Magazine, vol. 29, no. 5, pp. 116-127, Sep. 2012. [Nieman13] K. Nieman, J. Lin, M. Nassar, K. Waheed and B. L. Evans, “Cyclic Spectral Analysis of Power Line Noise in the 3-200 kHz Band”, Proc. IEEE Int. Sym. on Power Line Communications and Its Applications, Mar. 24-27, 2013. [Pauli06] V. Pauli, L. Lampe, and R. Schober. ”turbo dpsk” using soft multiple-symbol differential sphere decoding. IEEE Trans. on Information Theory, 52(4):1385–1398, 2006. [Raphaeli96] D. Raphaeli. Noncoherent coded modulation. IEEE Trans. on Comm., vol. 44, no. 2, pp. 172–183, 1996. [Tipping01] M.E. Tipping. Sparse Bayesian learning and the relevance vector machine. Journal of Machine Learning Research, vol. 1, pp. 211–244, 2001. [Umehara01] D. Umehara, M. Kawai, and Y. Morihiro. Performance analysis of noncoherent coded modulation for power line communications. Proc. Int. Symp. Power Line Commun. and Its Appl., pp. 291–298, 2001. 25 Backup Slides 26 Research Group • • • • Present: 9 PhD, 0 MS, 4 BS Alumni: 21 PhD, 9 MS, 142 BS 1376 alumni of real-time DSP course Communication systems Powerline communications (interference modeling & mitigation) Cellular, Wimax & Wi-Fi (interference modeling & mitigation) Mixed-signal IC design (mostly digital ADCs and synthesizers) • Image processing • Electronic design automation (EDA) tools/methods • Part of Wireless Networking & Communications Group 160 grad students, 20 faculty members, 13 affiliate companies Completed Projects 20 PhD and 9 MS alumni System SW release Prototype Funding equalization Matlab DSP/C Freescale, TI 2x2 testbed LabVIEW LabVIEW/PXI Oil&Gas Wimax/LTE resource alloc. LabVIEW DSP/C Freescale, TI Underwater comm. space-time comm. large rec. arrays Matlab Lake Travis testbed UT Applied Res. Labs Camera image acquisition Matlab DSP/C Intel, Ricoh Display image halftoning Matlab C HP, Xerox video halftoning Matlab C Qualcomm Matlab FPGA Intel, NI Linux/C++ Navy sonar Navy, NI ADSL Contribution Elec. design fixed point conv. automation distributed comp. DSP: Digital Signal Processor PXI: PCI Extensions for Inst. Current Projects 9 PhD students System Contributions Powerline interference reduction comm. testbeds Wi-Fi interference reduction SW release LabVIEW Matlab time-based analog-todigital converter Cellular (LTE) Matlab Handheld reducing rolling shutter camera artifacts Matlab reliability patterns Funding Freescale, TI Freescale, modems IBM, TI NI FPGA Intel, NI IBM 45nm TSMC 180nm cloud radio access net. baseband compression EDA Prototype Huawei Android TI NI Simulated Performance • Symbol error rate in different noise scenarios ~10dB ~6dB ~6dB ~8dB ~4dB Gaussian mixture model Middleton class A model • MMSE w/ (w/o) CSI: Parametric estimator assuming known (unknown) statistical parameters of noise • CS+LS: A compressed sensing and least squares based algorithm 30 A Smart Grid Communication to isolated area Power generation optimization Integrating alternative energy sources Load balancing Disturbance monitoring Smart metering Electric car charging & smart billing Source: ETSI 31 Power Lines • Built for unidirectional energy flow • Bidirectional information flow throughout smart grid will occur Low Voltage (LV) under 1 kV High Voltage (HV) 33 kV – 765 kV Medium Voltage (MV) 1 kV – 33 kV Transformer Source: ERDF 32 Today’s Power Grids in the United States • 7 large-scale power grids each managed by a regional utility company 700 GW generation capacity in total for long-haul high-voltage power transmission Synchronized independently, and exchange power via DC transfer • 130+ medium-scale power grids each managed by a local utility Local power distribution to residential, commercial and industrial customers • Heavy penalties in US for blackouts (2003 legislation) Utilities generate expected energy demand plus 12% Energy demand correlated with time of day Effect of plug-in electric vehicles (EVs) on energy demand uncertain Generation cost 30x higher during peak times vs. normal load • Traditional ways to increase capacity to meet peak demand increase Build generation plant $1B to $10B if new permit is issued Build transmission line at $0.6M/km which will take 5-10 years to complete Source: Jerry Melcher, IEEE Smart Grid Short Course, 22 Oct. 2011, Austin TX USA 33 Smart Power Meters at Customer Site • Enable local utilities to improve Operating efficiency System reliability Customer participation • Automatic metering infrastructure functions Interval reads (every 1/15/30/60 minutes) and on-demand reads and pings Transmit customer load profiles and system load snapshots Power quality monitoring Remote disconnect/reconnect and outage/restoration event notification • Need low-delay highly-reliable communication link to local utility • 75M smart meters sold in 2011 (20% increase vs. 2010) Source: Jerry Melcher, IEEE Smart Grid Short Course, 22 Oct. 2011, Austin TX USA 34 Local Utility Powerline Communications (PLC) • PLC modems (PRIME, etc.) use carrier sensed multiple access to determine when the medium is available for transmission • MV router plays similar role as a Wi-Fi access point 35 Sources of Powerline Noise Uncoordinated transmission Power line disturbance Electronic devices Taken from a local utility point of view 36 PLC In Different Frequency Bands Category Band Ultra Narrowband Narrowband Broadband Bit Rate Applications Standards 0.3 – 3 kHz ~100 bps • Automatic meter reading • Outage detection • Load control N/A 3 – 500 kHz • Smart metering ~500 kbps • Real-time energy management 1.8 – 250 ~200 Mbps • Home area networks MHz • PRIME, G3 • ITU-T G.hnem • IEEE P1901.2 • HomePlug • ITU-T G.hn • IEEE P1901 All of the above standards are based on multicarrier communications using orthogonal frequency division multiplexing (OFDM). 37 Physical Layer Parameters for OFDM Narrowband PLC Standards CENELEC A band is from 3 to 95 kHz. FCC band is from 34.375 to 487.5 kHz. PRIME and G3 use real-valued baseband OFDM. Others are complex-valued. 38 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Comparison Between Wireless and PLC Systems Wireless Communications Narrowband PLC (3-500 kHz) Time-selective fading and Doppler shift (cellular) Periodic with period of half AC main freq. plus lognormal time-selective fading Power loss vs. distance d d –n/2 where n is propagation constant e – a(f) d plus additional attenuation when passing through transformers Propagation Dynamically changing Determinism from fixed grid topology Synchronization Varies AC main power frequency Additive noise/ interference Assumed stationary and Gaussian Gaussian plus non-Gaussian noise dominated by cyclostationary component Time selectivity Asynchronous interference MIMO Uncoordinated users in Due to power electronics and Wi-Fi bands; uncoordinated users using other standards Frequency reuse in cellular Standardized for Wi-Fi and cellular Number of wires minus 1; G.9964 standard for broadband PLC 39 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Cyclostationary Impulsive Noise • Linear periodically time-varying system model H1 vR H2 N n RN … HM Hi - Linear time invariant filter N - Period in samples o Period (half of the AC cycle) is partitioned into M segments o Noise within each segment is stationary Segment: 1 23 40 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Asynchronous Impulsive Noise Modeling Wireless Emissions Uncoordinated Meters (coexistence) Total interference at receiver: Interference from source i 41 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Two Asynchronous Impulsive Noise Models • Gaussian Mixture Model (isotropic, zero-centered) Amplitude distribution • Middleton Class A (without additive Gaussian component) Special case of the Gaussian Mixture Model • Also model for additive uncoordinated wireless interference Middleton Class A for a Wi-Fi receiver in a Wi-Fi hotspot Gaussian mixture model for a Wi-Fi receiver in a cluster of Wi-Fi hotspots 42 Non-Gaussian Noise: Challenge to PLC • Performance of conventional communication system degrades in non-AWGN environment • Statistical modeling of powerline noise • Noise mitigation exploiting the noise model or structure Listen to the environment Estimate noise model Use model or structure to mitigate noise 43 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Narrowband PLC Systems • Problem: Non-Gaussian impulsive noise is primary limitation to communication performance yet traditional communication system design assumes additive noise is Gaussian • Goal: Improve communication performance in impulsive noise • Approach: Statistical modeling of impulsive noise • Solution #1: Receiver design (standard compliant) Parametric Methods Nonparametric Methods Listen to environment No training necessary Find model parameters Learn statistical model from communication signal structure Use model to mitigate noise Exploit sparsity to mitigate noise • Solution #2: Joint transmitter-receiver design 44 Parametric vs. Nonparametric Noise Mitigation Parametric Nonparametric Must build a statistical model of the noise Yes No Requires training data to compute model parameters Yes No Degrades in performance due to model mismatch Yes No Has high complexity when receiving message data No Yes 45 Cyclostationary Noise Modeling in Narrowband PLC (3-500 kHz) 1. M. Nassar, A. Dabak, I. H. Kim, T. Pande and B. L. Evans, “Cyclostationary Noise Modeling In Narrowband Powerline Communication For Smart Grid Applications”, Proc. IEEE Int. Conf. on Acoustics, Speech, and Signal Proc., Mar. 25-30, 2012, Kyoto, Japan. 2. M. Nassar, J. Lin, Y. Mortazavi, A. Dabak, I. H. Kim and B. L. Evans, “Local Utility Powerline Communications in the 3-500 kHz Band: Channel Impairments, Noise, and Standards”, IEEE Signal Processing Magazine, Special Issue on Signal Processing Techniques for the Smart Grid, Sep. 2012, 14 pages. 46 Impulsive Noise in Broadband PLC: Modeling and Mitigation 3. M. Nassar, K. Gulati, Y. Mortazavi, and B. L. Evans, “Statistical Modeling of Asynchronous Impulsive Noise in Powerline Communication Networks”, Proc. IEEE Int. Global Communications Conf., Dec. 5-9, 2011, Houston, TX USA. 4. J. Lin, M. Nassar and B. L. Evans, “Non-Parametric Impulsive Noise Mitigation in OFDM Systems Using Sparse Bayesian Learning”, Proc. IEEE Int. Global Communications Conf., Dec. 5-9, 2011, Houston, TX USA. 47 Statistical-Physical Modeling • Interference from a single source Noise envelope k pulses in a window of duration T (k) (j) Tk Pulse emission duration (1) (2) τj Pulse arrival time t=0 Emission duration: geometrically distributed with mean μ Pulse arrivals: homogeneous Poisson point process with rate λ Assuming channel between interference source and receiver has flat fading 48 Parametric Vs. Non-Parametric Methods • Noise in different PLC networks has different statistical models • Mitigation algorithms need to be robust in different noise scenarios Parametric Methods Non-Parametric Methods Assume parameterized noise statistics Yes No Performance degradation due to model mismatch Yes No Training needed Yes No 49 Non-Parametric Mitigation Using Null Tones J : Index set of null tones FJ : DFT sub-matrix e: Impulsive noise in time domain g: AWGN with unknown variance • A compressed sensing problem Exploiting the sparse structure of the time-domain impulsive noise • Sparse Bayesian learning (SBL) Proposed initially by M. L. Tipping A Bayesian inference framework with sparsity promoting prior 50 Sparse Bayesian Learning • Bayesian inference Sparsity promoting prior: Likelihood: Posterior probability: e | g ~ CN (0, ), = diag (g ) yJ | g , 2 ~ CN (0, FF * 2 I ) e | yJ ; g , 2 ~ CN (m, e ) • Iterative algorithm Step 1: Maximum likelihood estimation of hyper-parameters (γ, σ2) Solved by expectation maximization (EM) algorithm (e is latent variable) Step 2: Estimate e from the mean of the posterior probability, go to Step 1 51 Non-Parametric Mitigation Using All Tones • Joint estimation of data and noise J : Index set of data tones z : Received signal in frequency domain Treat the received signal in data tones as additional hyper-parameters Estimate of zJ is sent to standard OFDM equalizer and symbol detector 52 Smart grids – Powerline noise – Receiver design – Testbeds – Conclusion Time-Domain Interleaving Coded performance in cyclostationary noise Burst duty cycle 10% Time-domain interleaving over an AC cycle Current PLC standards use frequency-domain interleaving (FDI) Burst duty cycle 30% 53