doc.: IEEE /0034r1 May 2007 Slide 1Submission Huawei Technologies Simulation Results for Spectral Correlation Sensing with Real DTV Signals IEEE P Wireless RANs Date: Authors: Notice: This document has been prepared to assist IEEE It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any modifications thereof, in the creation of an IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEE’s sole discretion to permit others to reproduce in whole or in part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this contribution may be made public by IEEE Patent Policy and Procedures: The contributor is familiar with the IEEE 802 Patent Policy and Procedures including the statement "IEEE standards may include the known use of patent(s), including patent applications, provided the IEEE receives assurance from the patent holder or applicant with respect to patents essential for compliance with both mandatory and optional portions of the standard." Early disclosure to the Working Group of patent information that might be relevant to the standard is essential to reduce the possibility for delays in the development process and increase the likelihood that the draft publication will be approved for publication. Please notify the Chairhttp://standards.ieee.org/guides/bylaws/sb-bylaws.pdf Carl R. StevensonCarl R. Stevenson as early as possible, in written or electronic form, if patented technology (or technology under patent application) might be incorporated into a draft standard being developed within the IEEE Working Group. If you have questions, contact the IEEE Patent Committee Administrator at >

doc.: IEEE /0034r1 May 2007 Slide 2Submission Huawei Technologies INTRODUCTION Spectral correlation sensing schemes have been proposed ( ). –Incumbent signals can be identified/sensed/detected calculating and using correlation values between spectral components of the received signals and spectral information of target incumbent signals pre-stored at the receiver. Simulation results are provided in this presentation. –For 12 real DTV signals provided, probabilities of misdetection are plotted for various SNR values for three probabilities of false alarm of 10 %, 1 %, and 0.1 %. –The real DTV processing procedure and simulation models suggested by Sensing tiger team are applied for these simulations ( and ).

doc.: IEEE /0034r1 May 2007 Slide 3Submission Huawei Technologies BRIEF DESCRIPTION OF THE SCHEME (1) TV band signal sensing for one channel band –Use only spectral components – not time domain components Less sensitive on other parameters used to design TV band tuners – for example, phase noise, etc. –Take FFT transform of received real DTV signals at the receiver for only one TV band –One example BW=F=6 MHz for one band case Sampling interval T=1/B=1/6 us, sampling rate=BW=6 MHz Frequency resolution (or frequency separation) F 0 =3 KHz Time period T 0 =1/F 0 =1/3 ms Number of samples needed N 0 =T 0 /T= 2 KHz Needs 2K point FFTs

doc.: IEEE /0034r1 May 2007 Slide 4Submission Huawei Technologies BRIEF DESCRIPTION OF THE SCHEME (2) Sensing procedure for DTV signals used –If more components are taken in a band, better performance can be achieved.  only 50, 100, and 200 frequency components taken in a 6 MHz band  only 50, 100, 200 multiplications needed for processing –Compare these values calculating correlations: compare the shape of spectrum of a received signal Calculate correlations between spectral components of the received signal and pre-stored spectral component values for DTV signals If this correlation value is larger than a predetermined value, the threshold, the judgment is that DTV signal exists. –Average frequency component values for several symbol periods can be used to have better sensing results.  for simulations here the procedure is applied for 1/3 ms (one symbol duration), 2 ms, and 10 ms

doc.: IEEE /0034r1 May 2007 Slide 5Submission Huawei Technologies SPECTRA OF TV CHANNELS Analyzing the Signal Quality of NTSC and ATSC Television RF Signals.htm, Glen Kropuenske, Sencore NTSC signal spectrumDTV signal spectrum

doc.: IEEE /0034r1 May 2007 Slide 6Submission Huawei Technologies ONE ILLUSTRATED EXAMPLE 8 measured spectral components Using 8 measured components, a correlation is calculated between frequency spectral components of a received signal and pre-stored spectral information of an incumbent signal. Spectral information of an incumbent signal

doc.: IEEE /0034r1 May 2007 Slide 7Submission Huawei Technologies TARGET REAL DTV SIGNALS USED Site NameChannelType of CaptureField Strength (dBuV/m) WAS-047/48/0148in-home10 WAS-311/48/0148Outdoor-30 feet13 WAS-311/35/0135Outdoor-30 feet16 WAS-311/36/0136Outdoor-30 feet17 WAS-086/48/0148Outdoor-30 feet35 WAS-006/34/0134Outdoor-30 feet41 WAS-003/27/0127Outdoor-30 feet43 WAS-051/35/0135Outdoor-30 feet44 WAS-049/39/0139in-home44.8 WAS-032/48/0148in-home46.3 WAS-068/36/0136Outdoor-30 feet48 WAS-049/34/0134in-home48.2

doc.: IEEE /0034r1 May 2007 Slide 8Submission Huawei Technologies SPECTRAL PATTERN USED: AFTER SMOOTHING The spectrum signature is obtained by averaging 50 DTV signals power spectra.

doc.: IEEE /0034r1 May 2007 Slide 9Submission Huawei Technologies THRESHOLD SETTING (1) The threshold is set by 10,000 Monte Carlo experiments. First calculate a correlation between spectrum mask and noise power spectrum. Then according to the correlation value obtained in the first step we can set different probability of false alarm corresponding to threshold. We make 10,000 Monte Carlo experiments and adjust threshold to satisfy different probability of false alarm precisely.

doc.: IEEE /0034r1 May 2007 Slide 10Submission Huawei Technologies THRESHOLD SETTING (2) The threshold is set by using 10,000 correlation calculations for 10,000 randomly generated noise spectra with pre-stored spectral information. Calculation procedure –Generate a random noise whose average amplitude is 1 and get its spectrum. Calculate a correlation value with the pre-stored spectral information. Do this procedure for each of 10,000 randomly generated noise signals. Keep 10,000 correlation values from 10,000 calculations. –Choose a correlation value, T, corresponding to a probability of false alarm such that the probability that a correlation value is larger than T is equal to the probability of false alarm. This T is the threshold for the probability of false alarm. Apply this procedure for each probability of false alarm.

doc.: IEEE /0034r1 May 2007 Slide 11Submission Huawei Technologies FILTER SPECIFICATIONS Filter the signal using a passband filter with a 6 MHz bandwidth with a center frequency of f= MHz. The filter adopts a “ brick wall ” filter. Generate white noise sampled at MHz and filter it through the same filter.

doc.: IEEE /0034r1 May 2007 Slide 12Submission Huawei Technologies DETAILED SIMULATION PARAMETERS Sensing time: 1/3, 2, and 10 ms 50, 100, and 200 points are used for correlation calculation. –If more points are used, higher computational complexity, but better performance. SNR= [-25:5] dB Simulation results for each signal file are suggested.

doc.: IEEE /0034r1 May 2007 Slide 13Submission Huawei Technologies SIMULATION RESULTS 1/3 ms, 50 POINTS (1) Probabilities of misdetection vs SNRs under false detection probability=10%

doc.: IEEE /0034r1 May 2007 Slide 14Submission Huawei Technologies SIMULATION RESULTS 2 ms, 100 POINTS (1) Probabilities of misdetection vs SNRs under false detection probability=10%

doc.: IEEE /0034r1 May 2007 Slide 15Submission Huawei Technologies SIMULATION RESULTS 10 ms, 200 POINTS (1) Probabilities of misdetection vs SNRs under false detection probability=10%

doc.: IEEE /0034r1 May 2007 Slide 16Submission Huawei Technologies SIMULATION RESULTS 1/3 ms, 50 POINTS (2) Probabilities of misdetection vs SNRs under false detection probability=1%

doc.: IEEE /0034r1 May 2007 Slide 17Submission Huawei Technologies SIMULATION RESULTS 2 ms, 100 POINTS (2) Probabilities of misdetection vs SNRs under false detection probability=1%

doc.: IEEE /0034r1 May 2007 Slide 18Submission Huawei Technologies SIMULATION RESULTS 10 ms, 200 POINTS (2) Probabilities of misdetection vs SNRs under false detection probability=1%

doc.: IEEE /0034r1 May 2007 Slide 19Submission Huawei Technologies SIMULATION RESULTS 1/3 ms, 50 POINTS (3) Probabilities of misdetection vs SNRs under false detection probability=0.1%

doc.: IEEE /0034r1 May 2007 Slide 20Submission Huawei Technologies SIMULATION RESULTS 2 ms, 100 POINTS (3) Probabilities of misdetection vs SNRs under false detection probability=0.1%

doc.: IEEE /0034r1 May 2007 Slide 21Submission Huawei Technologies SIMULATION RESULTS 10 ms, 200 POINTS (3) Probabilities of misdetection vs SNRs under false detection probability=0.1%

doc.: IEEE /0034r1 May 2007 Slide 22Submission Huawei Technologies COMPARISON OF SIMULATION RESULTS (1) Prob. of FA sensing time / number of points SNR (dB) Prob. MD = Prob. MD = % 1/3 ms, 50 points 2 ms, 100 points 10 ms, 200 points (ave -3.5) (ave -7) (ave -8) (ave -12) (ave -15.5) < -26 (ave < -26) 1% 1/3 ms, 50 points 2 ms, 100 points 10 ms, 200 points -6 – 4 (ave -1) -8 – 2 (ave -3) -9 – 2 (ave -3.5) (ave -7) (ave -8) (ave -11.5) 0.1 % 1/3 ms, 50 points 2 ms, 100 points 10 ms, 200 points -5 – 5 (ave 0) (ave -2) (ave -3) -10 – 0 (ave -5) (ave -6) (ave -9)

doc.: IEEE /0034r1 May 2007 Slide 23Submission Huawei Technologies COMPARISON OF SIMULATION RESULTS (2) Prob-MD = 10-2Prob-MD = 10-1

doc.: IEEE /0034r1 May 2007 Slide 24Submission Huawei Technologies PERFORMANCE DEGRADATION BY ADJACENT CHANNELS (1) With one adjacent channel and two adjacent channels Y = S +N + I, where I is adjacent interference. Based on the original sensing procedure, we consider adding adjacent interference in it. –The simulations can be done only by adding this interference part to the previous simulations. –Simulation parameters are the same as for the original sensing simulations. DTV signal file WAS_47_48_ _opt is used.

doc.: IEEE /0034r1 May 2007 Slide 25Submission Huawei Technologies PERFORMANCE DEGRADATION BY ADJACENT CHANNELS (2)

doc.: IEEE /0034r1 May 2007 Slide 26Submission Huawei Technologies SIMULATION RESULTS WITH NOISE UNCERTAINTY Noise uncertainty: +/- 1 dB We use robust statistics approach to model noise uncertainty. DTV signal file WAS_47_48_ _opt used Average PSD Range of PSD Upper limit of PSD to calculate Probability of False Alarm used Lower limit of PSD to calculate the Probability of Misdetection used

doc.: IEEE /0034r1 May 2007 Slide 27Submission Huawei Technologies SIMULATION RESULTS (1) Probabilities of misdetection vs SNRs under false detection probability=1%

doc.: IEEE /0034r1 May 2007 Slide 28Submission Huawei Technologies SIMULATION RESULTS (2) Probabilities of misdetection vs SNRs under false detection probability=10%

doc.: IEEE /0034r1 May 2007 Slide 29Submission Huawei Technologies SIMULATION RESULTS (3) Probabilities of misdetection vs SNRs under false detection probability=0.1%

doc.: IEEE /0034r1 May 2007 Slide 30Submission Huawei Technologies FOR DIFFERENT SAMPLING TIME AND NO. OF POINT For WAS_311_35_ _ref

doc.: IEEE /0034r1 May 2007 Slide 31Submission Huawei Technologies CONCLUSIONS Simulation results show that the spectral correlation sensing schemes have acceptable performance. This proposed scheme can overcome adjacent channel interference and noise uncertainty. The performance can be improved to meet any sensing requirements by: –Increasing sensing time: 1/3 ms  2 ms  10 ms : still this scheme needs shorter time comparing others –Increasing no. of samples: 50  100  200 Samples : still this scheme needs lower complexity comparing others –Modifying spectral information stored: smoothing and emphasizing