《高頻電子線路》實驗教學(xué)大綱

1、高頻電子線路實驗教學(xué)大綱高頻電子線路實驗教學(xué)大綱2OutlineIntroductionSingle carrier single antenna systemsRadio frequency interference modelingEstimation of interference model parametersFiltering/detectionMulti-input multi-output (MIMO) single carrier systemsCo-channel interference modelingConclusionsFuture work2高頻電子線路實驗教學(xué)大

2、綱3Impulsive NoiseAlmost instantaneous (impulse-like) phenomenon Unwanted “clicks” and “pops” in an audio recordingNon-Gaussian statisticsModels electromagnetic interference through conduction as well as induction (via radiation)Example sourcesStopping and starting of electrical devicesClocks and har

3、monicsCommunication transmissionTalk focuses on modeling asynchronous non-periodic noise高頻電子線路實驗教學(xué)大綱4Radio Frequency InterferenceElectromagnetic interference from radiationLimits wireless communication performanceApplications of RFI modelingSense and mitigate strategies for coexistenceof wireless ne

4、tworks and servicesSense and avoid strategies for cognitive radioWe focus on sense and mitigate strategies for wireless receivers embedded in notebooksPlatform noise from users computer subsystemsCo-channel interference from other in-band wireless networks and services高頻電子線路實驗教學(xué)大綱5Computational Plat

5、form Noise5ObjectivesDevelop offline methods to improve communication performance in presence of computer platform RFIDevelop adaptive online algorithms for these methodsApproachStatistical modeling of RFIFiltering/detection based on estimated model parametersWithin computing platforms, wireless tra

6、nsceivers experience radio frequency interference from clocks and bussesWe will use noise and interference interchangeablyBackup高頻電子線路實驗教學(xué)大綱6Impact of RFI6Impact of LCD noise on throughput for an IEEE 802.11g embedded wireless receiver Shi, Bettner, Chinn, Slattery & Dong, 2006BackupBackup高頻電子線路實驗教學(xué)

7、大綱7Statistical Modeling of RFI7Radio frequency interferenceSum of independent radiation eventsPredominantly non-Gaussian impulsive statisticsKey statistical-physical modelsMiddleton Class A, B, C modelsIndependent of physical conditions (canonical)Sum of independent Gaussian and Poisson interference

8、Symmetric Alpha Stable modelsApproximation of Middleton Class B modelBackupBackup高頻電子線路實驗教學(xué)大綱8Assumptions for RFI Modeling8 Key assumptions for Middleton and Alpha Stable modelsMiddleton, 1977Furutsu & Ishida, 1961 Infinitely many potential interfering sources with same effective radiation power Pow

9、er law propagation loss Poisson field of interferers with uniform intensity lPr(number of interferers = M |area R) Poisson(M; lR) Uniformly distributed emission times Temporally independent (at each sample time) Limitations Alpha Stable models do not include thermal noise Temporal dependence may exi

10、st高頻電子線路實驗教學(xué)大綱9Our Contributions9Computer Platform Noise ModellingEvaluate fit of measured RFI data to noise models Middleton Class A model Symmetric Alpha StableParameter EstimationEvaluate estimation accuracy vs complexity tradeoffsFiltering / DetectionEvaluate communication performance vs complex

11、ity tradeoffs Middleton Class A: Correlation receiver, Wiener filtering, and Bayesian detector Symmetric Alpha Stable: Myriad filtering, hole punching, and Bayesian detector Mitigation of computational platform noise in single carrier, single antenna systems Nassar, Gulati, DeYoung, Evans & Tinsley,

12、 ICASSP 2008, JSPS 2009高頻電子線路實驗教學(xué)大綱10Middleton Class A model10Probability Density Function12!)(220222AmwhereemAezfmzmmmAZm-10-5051000.10.20.30.40.50.60.7Noise amplitudeProbability density functionPDF for A, = 0.8 AParameterDescriptionRangeOverlap Index. Product of average number of emissions per sec

13、ond and mean duration of typical emissionA 10-2, 1Gaussian Factor. Ratio of second-order moment of Gaussian component to that of non-Gaussian component 10-6, 1高頻電子線路實驗教學(xué)大綱11Symmetric Alpha Stable Model11Characteristic Function Closed-form PDF expression only for = 1 (Cauchy), = 2 (Gaussian), = 1/2 (

14、Levy), = 0 (not very useful) Approximate PDF using inverse transform of power series expansion Second-order moments do not exist for do not exist|)(jePDF for = 1.5, = 0, = 10-5005000.010.020.030.040.050.060.07Noise amplitudeProbability density functionParameterDescriptionRangeCharacteristic Exponent

15、. Amount of impulsivenessLocalization. Analogous to meanDispersion. Analogous to variance2,0),(),0(BackupBackup高頻電子線路實驗教學(xué)大綱12Example Power Spectral DensitiesMiddleton Class ASymmetric Alpha Stable00.10.20.30.40.50.60.70.80.91-10-8-6-4-20246810FrequencyPower Spectrum Magnitude (dB)Power Spectal Densi

16、ty of Class A noise, A = 0.15, = 0.100.10.20.30.40.50.60.70.80.91-10-8-6-4-20246810FrequencyPower Spectrum Magnitude (dB)Power Spectal Density of S S noise, = 1.5, = 10, = 0Overlap Index (AGaussian Factor () Characteristic Exponent ()Localization () = 0Dispersion () = 10Simulated Densities高頻電子線路實驗教學(xué)

17、大綱13Estimation of Noise Model Parameters13Middleton Class A modelBased on Expectation Maximization Zabin & Poor, 1991Find roots of second and fourth order polynomials at each iterationAdvantage: Small sample size is required (1000 samples)Disadvantage: Iterative algorithm, computationally intensiveS

18、ymmetric Alpha Stable ModelBased on Extreme Order Statistics Tsihrintzis & Nikias, 1996Parameter estimators require computations similar to mean and standard deviation computationsAdvantage:Fast / computationally efficient (non-iterative)Disadvantage: Requires large set of data samples (10000 sample

19、s)BackupBackup高頻電子線路實驗教學(xué)大綱1414Wireless Networking and Communications GroupResults on Measured RFI Data1425 radiated computer platform RFI data sets from Intel50,000 samples taken at 100 MSPSEstimated Parameters for Data Set #18Symmetric Alpha Stable ModelLocalization ()0.0065KL Divergence 0.0308Char

20、acteristic exp. ()1.4329Dispersion ()0.2701Middleton Class A ModelOverlap Index (A)0.0854KL Divergence0.0494Gaussian Factor ()0.6231Gaussian ModelMean ()0KL Divergence0.1577Variance (2) 1KL Divergence: Kullback-Leibler divergence-6-4-2024600.10.20.30.40.50.60.70.80.9Noise amplitudeProbability Densit

21、y Function Measured PDFEst. -Stable PDFEst. Class A PDFEst. Gaussian PDFMeasured PDF Gaussian PDF Middleton Class A PDF Alpha Stable PDF 高頻電子線路實驗教學(xué)大綱15051015202500.050.10.150.20.25Measurement Set NumberKullback-Leibler (KL) Divergence Estimated Alpha Stable modelEstimated Class A modelEstimated Gaus

22、sian model15Results on Measured RFI DataBest fit for 25 data sets under different platform RFI conditionsKL divergence plotted for three candidate distributions vs. data set numberSmaller KL value means closer fitGaussianClass AAlpha Stable高頻電子線路實驗教學(xué)大綱16Video over Impulsive ChannelsVideo demonstrati

23、on for MPEG II video stream 10.2 MB compressed stream from camera (142 MB uncompressed) Compressed over additive impulsive noise channel Binary phase shift keyingRaised cosine pulse10 samples/symbol10 symbols/pulse length Composite of transmitted and received MPEG II video streams Shows degradation

24、of video quality over impulsive channels with standard receivers (based on Gaussian noise assumption)Wireless Networking and Communications Group16Additive Class A Noise ValueOverlap index (A)0.35Gaussian factor ()0.001SNR19 dB高頻電子線路實驗教學(xué)大綱17Filtering and Detection17Pulse ShapingPre-FilteringMatched

25、FilterDetection RuleImpulsive NoiseMiddleton Class A noiseSymmetric Alpha Stable noiseFilteringWiener Filtering (Linear)DetectionCorrelation Receiver (Linear)Bayesian DetectorSpaulding & Middleton, 1977Small Signal Approximation to Bayesian detectorSpaulding & Middleton, 1977FilteringMyriad Filterin

26、gOptimal Myriad Gonzalez & Arce, 2001Selection MyriadHole Punching Ambike et al., 1994DetectionCorrelation Receiver (Linear)MAP approximationKuruoglu, 1998BackupBackupBackupBackupBackupAssumptionMultiple samples of the received signal are available N Path Diversity Miller, 1972 Oversampling by N Mid

27、dleton, 1977Backup高頻電子線路實驗教學(xué)大綱18Results: Class A Detection18Pulse shapeRaised cosine10 samples per symbol10 symbols per pulseChannelA = 0.35 = 0.5 10-3MemorylessMethodComp. ComplexityDetection Perform.Correl.LowLowWienerMediumLowBayesian S.S. Approx.MediumHighBayesianHighHigh-35-30-25-20-15-10-50510

28、1510-510-410-310-210-1100SNRBit Error Rate (BER) Correlation ReceiverWiener FilteringBayesian DetectionSmall Signal ApproximationCommunication PerformanceBinary Phase Shift KeyingBackupBackupBackup高頻電子線路實驗教學(xué)大綱19Results: Alpha Stable Detection19Use dispersion parameter in place of noise variance to g

29、eneralize SNRMethodComp. ComplexityDetection Perform.Hole PunchingLowMediumSelection MyriadLowMediumMAP Approx.MediumHighOptimal MyriadHigh MediumBackupBackup-10-50510152010-210-1100Generalized SNR (in dB)Bit Error Rate (BER) Matched FilterHole PunchingMAPMyriadCommunication PerformanceSame transmit

30、ter settings as previous slideBackupcBackupBackupBackup高頻電子線路實驗教學(xué)大綱20Video over Impulsive Channels #2Video demonstration for MPEG II video stream revisited 5.9 MB compressed stream from camera (124 MB uncompressed) Compressed over additive impulsive noise channel Binary phase shift keyingRaised cosi

31、ne pulse10 samples/symbol10 symbols/pulse length Composite of transmitted video stream, video stream from a correlation receiver based on Gaussian noise assumption, and video stream for a Bayesian receiver tuned to impulsive noiseWireless Networking and Communications Group20Additive Class A Noise V

32、alueOverlap index (A)0.35Gaussian factor ()0.001SNR19 dB高頻電子線路實驗教學(xué)大綱21Video over Impulsive Channels #2Structural similarity measure Wang, Bovik, Sheikh & Simoncelli, 2004Score is 0,1 where higher means better video qualityFrame numberBit error rates for 50 million bits sent:6 x 10-6 for correlation

33、receiver0 for RFI mitigating receiver (Bayesian)高頻電子線路實驗教學(xué)大綱22Extensions to MIMO systems22Radio Frequency Interference Modeling and Receiver Design for MIMO systemsRFI ModelSpatial Corr.Physical ModelCommentsMiddleton Class ANoYesUni-variate modelAssume independent or uncorrelated noise for multiple

34、 antennasReceiver design:Gao & Tepedelenlioglu, 2007 Space-Time CodingLi, Wang & Zhou, 2004 Performance degradation in receiversWeighted Mixture of Gaussian DensitiesYesNoNot derived based on physical principlesReceiver design:Blum et al., 1997 Adaptive Receiver DesignBivariate Middleton Class AMcDo

35、nald & Blum, 1997YesYesExtensions of Class A model to two-antenna systemsBackup高頻電子線路實驗教學(xué)大綱23Our Contributions23RFI Modeling Evaluated fit of measured RFI data to the bivariate Middleton Class A model McDonald & Blum, 1997 Includes noise correlation between two antennas Parameter Estimation Derived

36、parameter estimation algorithm based on the method of moments (sixth order moments)Performance Analysis Demonstrated communication performance degradation of conventional receivers in presence of RFI Bounds on communication performanceChopra , Gulati, Evans, Tinsley, and Sreerama, ICASSP 2009Receive

37、r Design Derived Maximum Likelihood (ML) receiver Derived two sub-optimal ML receivers with reduced complexity2 x 2 MIMO receiver design in the presence of RFIGulati, Chopra, Heath, Evans, Tinsley & Lin, Globecom 2008BackupBackupBackup高頻電子線路實驗教學(xué)大綱24-10-50510152010-310-210-1SNR in dBVector Symbol Err

38、or Rate Optimal ML Receiver (for Gaussian noise)Optimal ML Receiver (for Middleton Class A)Sub-Optimal ML Receiver (Four-Piece)Sub-Optimal ML Receiver (Two-Piece)Results: RFI Mitigation in 2 x 2 MIMO 24ANoise CharacteristicImprove-ment0.01Highly Impulsive15 dB0.1Moderately Impulsive8 dB1Nearly Gauss

39、ian0.5 dBImprovement in communication performance over conventional Gaussian ML receiver at symbol error rate of 10-2Communication Performance (A = 0.1, 1= 0.01, 2= 0.1, k = 0.4)高頻電子線路實驗教學(xué)大綱25Results: RFI Mitigation in 2 x 2 MIMO 25Complexity AnalysisReceiverQuadratic FormsExponentialComparisonsGaus

40、sian MLM200Optimal ML2M22M20Sub-optimal ML (Four-Piece)2M202M2Sub-optimal ML (Two-Piece)2M20M2Complexity Analysis for decoding M-level QAM modulated signalCommunication Performance (A = 0.1, 1= 0.01, 2= 0.1, k = 0.4)-10-50510152010-310-210-1SNR in dBVector Symbol Error Rate Optimal ML Receiver (for

41、Gaussian noise)Optimal ML Receiver (for Middleton Class A)Sub-Optimal ML Receiver (Four-Piece)Sub-Optimal ML Receiver (Two-Piece)高頻電子線路實驗教學(xué)大綱2626Co-Channel Interference ModelingWireless Networking and Communications Group26Region of interferer locations determines interference model Gulati, Chopra,

42、Evans & Tinsley, Globecom 2009Symmetric Alpha StableMiddleton Class A高頻電子線路實驗教學(xué)大綱2727Co-Channel Interference ModelingPropose unified framework to derive narrowband interference models for ad-hoc and cellular network environments Key result: tail probabilities (one minus cumulative distribution funct

43、ion)Wireless Networking and Communications Group27Case 1: Ad-hoc networkCase 3-a: Cellular network (mobile user)00.10.20.30.40.50.60.70.80.9110-410-310-210-1100Interference threshold (a)P ( Interfernce amplitude a ) Simulated Symmetric Alpha Stable00.10.20.30.40.50.60.70.80.9110-1010-5100Interferenc

44、e threshold (a)P ( Interference amplitude a) SimulatedSymmetric Alpha StableGaussianMiddleton Class A高頻電子線路實驗教學(xué)大綱28Conclusions28Radio Frequency Interference from computing platformAffects wireless data communication transceiversModels include Middleton and alpha stable distributionsRFI mitigation ca

45、n improve communication performanceSingle carrier, single antenna systemsLinear and non-linear filtering/detection methods exploredSingle carrier, multiple antenna systemsOptimal and sub-optimal receivers designed Bounds on communication performance in presence of RFIResults extend to co-channel int

46、erference modeling 高頻電子線路實驗教學(xué)大綱29RFI Mitigation Toolbox in MATLAB Provides a simulation environment forRFI generationParameter estimation algorithmsFiltering and detection methodsDemos for communication performance analysisLatest Toolbox ReleaseVersion 1.4 beta, Dec. 14th 200929Snapshot of a demo高頻電

47、子線路實驗教學(xué)大綱30Publications and Presentations30 Journal and conference papersM. Nassar, K. Gulati, M. R. DeYoung, B. L. Evans and K. R. Tinsley, “Mitigating Near-Field Interference in Laptop Embedded Wireless Transceivers”, J. of Signal Proc. Systems, Mar 2009, invited paper. M. Nassar, K. Gulati, A. K.

48、 Sujeeth, N. Aghasadeghi, B. L. Evans and K. R. Tinsley, “Mitigating Near-field Interference in Laptop Embedded Wireless Transceivers”, Proc. IEEE Int. Conf. on Acoustics, Speech, and Signal Proc., Mar. 30-Apr. 4, 2008, Las Vegas, NV USA.K. Gulati, A. Chopra, R. W. Heath Jr., B. L. Evans, K. R. Tins

49、ley, and X. E. Lin, “MIMO Receiver Design in the Presence of Radio Frequency Interference”, Proc. IEEE Int. Global Communications Conf., Nov. 30-Dec. 4th, 2008, New Orleans, LA USA.A. Chopra, K. Gulati, B. L. Evans, K. R. Tinsley, and C. Sreerama, “Performance Bounds of MIMO Receivers in the Presenc

50、e of Radio Frequency Interference”, Proc. IEEE Int. Conf. on Acoustics, Speech, and Signal Proc., Apr. 19-24, 2009, Taipei, Taiwan, accepted. K. Gulati, A. Chopra, B. L. Evans and K. R. Tinsley, “Statistical Modeling of Co-Channel Interference”, Proc. IEEE Int. Global Communications Conf., Nov. 30-D

51、ec. 4, 2009, Honolulu, HI USA, accepted.K. Gulati, A. Chopra, B. L. Evans and K. R. Tinsley, “Statistical Modeling of Co-Channel Interference in a Field of Poisson and Poisson Distributed Interferers”, Proc. IEEE Int. Conf. on Acoustics, Speech, and Signal Proc., Mar. 15-19, 2010, Dallas, TX USA. Pr

52、oject Web site高頻電子線路實驗教學(xué)大綱31Future Work31Extend RFI modeling forAdjacent channel interferenceMulti-antenna systemsTemporally correlated interferenceMulti-input multi-output (MIMO) single carrier systemsRFI modeling and receiver designMulticarrier communication systemsCoding schemes resilient to RFIS

53、ystem level techniques to reduce computational platform generated RFIBackup高頻電子線路實驗教學(xué)大綱3232Thank You.Questions ?高頻電子線路實驗教學(xué)大綱33References33RFI Modeling1 D. Middleton, “Non-Gaussian noise models in signal processing for telecommunications: New methods and results for Class A and Class B noise models”,

54、 IEEE Trans. Info. Theory, vol. 45, no. 4, pp. 1129-1149, May 1999.2 K.F. McDonald and R.S. Blum. “A physically-based impulsive noise model for array observations”, Proc. IEEE Asilomar Conference on Signals, Systems& Computers, vol 1, 2-5 Nov. 1997.3 K. Furutsu and T. Ishida, “On the theory of ampli

55、tude distributions of impulsive random noise,” J. Appl. Phys., vol. 32, no. 7, pp. 12061221, 1961.4 J. Ilow and D . Hatzinakos, “Analytic alpha-stable noise modeling in a Poisson field of interferers or scatterers”, IEEE transactions on signal processing, vol. 46, no. 6, pp. 1601-1611, 1998.Paramete

56、r Estimation5 S. M. Zabin and H. V. Poor, “Efficient estimation of Class A noise parameters via the EM Expectation-Maximization algorithms”, IEEE Trans. Info. Theory, vol. 37, no. 1, pp. 60-72, Jan. 19916 G. A. Tsihrintzis and C. L. Nikias, Fast estimation of the parameters of alpha-stable impulsive

57、 interference, IEEE Trans. Signal Proc., vol. 44, Issue 6, pp. 1492-1503, Jun. 1996RFI Measurements and Impact7 J. Shi, A. Bettner, G. Chinn, K. Slattery and X. Dong, A study of platform EMI from LCD panels - impact on wireless, root causes and mitigation methods,“ IEEE International Symposium on El

58、ectromagnetic Compatibility, vol.3, no., pp. 626-631, 14-18 Aug. 2006高頻電子線路實驗教學(xué)大綱34References (cont)34Filtering and Detection8 A. Spaulding and D. Middleton, “Optimum Reception in an Impulsive Interference Environment-Part I: Coherent Detection”, IEEE Trans. Comm., vol. 25, no. 9, Sep. 19779 A. Spau

59、lding and D. Middleton, “Optimum Reception in an Impulsive Interference Environment Part II: Incoherent Detection”, IEEE Trans. Comm., vol. 25, no. 9, Sep. 197710 J.G. Gonzalez and G.R. Arce, “Optimality of the Myriad Filter in Practical Impulsive-Noise Environments”, IEEE Trans. on Signal Processin

60、g, vol 49, no. 2, Feb 200111 S. Ambike, J. Ilow, and D. Hatzinakos, “Detection for binary transmission in a mixture of Gaussian noise and impulsive noise modelled as an alpha-stable process,” IEEE Signal Processing Letters, vol. 1, pp. 5557, Mar. 1994.12 J. G. Gonzalez and G. R. Arce, “Optimality of