Download presentation

Presentation is loading. Please wait.

Published byTimothy Bates Modified over 2 years ago

1
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 1 Clear Channel Assessment Energy Detection (CCA-ED) in y 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 IEEEs name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEEs 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 TAG 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 Chair 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 TAG. If you have questions, contact the IEEE Patent Committee Administrator at.http:// Date: Authors:

2
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 2 Abstract This is an evaluation of clear channel assessment energy detection (CCA-ED) in y The other system being considered with y is h Similar energy detection is being considered in 16h for its listen-before-talk (LBT) protocol

3
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 3 Channel Bandwidths and Sensing Times There are three possible bandwidths for y and for each bandwidth there is a different sensing time BandwidthSensing Time 20 MHz 4 s 10 MHz 8 s 5 MHz 16 s

4
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 4 Energy Detection Test Statistic The energy detection CCA mechanism estimates the power of the signal observed over a sensing time and compares the estimate to a threshold This estimate of the power is the test statistic From [1] we have the formula for the test statistic B is the signal bandwidth M is the number of samples

5
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 5 Number of Samples in Estimate The number of samples used in the estimate is the sensing time the sample rate (also the bandwidth) BandwidthSensing TimeM 20 MHz 4 s MHz 8 s 80 5 MHz 16 s 80

6
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 6 Density Function of Test Statistic This is a binary hypothesis test Two hypotheses –Noise only –Signal plus noise x(t) signal w(t) noise

7
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 7 Parameters BBandwidth NNoise Power Spectral Density N= NF N= = -164 dBm/Hz PSignal Power MNumber of samples Assume a conservative 10 dB receiver noise figure

8
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 8 Probability Density Function of T For large M the central limit theorem says the density function of T is approximately Gaussian PDF of T under H 0 PDF of T under H 1

9
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 9 Detector Threshold In y there is a different threshold for each bandwidth The draft [2] only specifies a maximum CCA-ED threshold BandwidthMAX Threshold 20 MHz-72 dBm 10 MHz-75 dBm 5 MHz-78 dBm

10
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 10 Detector Threshold BandwidthThreshold 20 MHz-72 dBm 10 MHz-75 dBm 5 MHz-78 dBm We will use typical values from [3] These are the same as the maximum threshold values

11
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 11 Probability of Detection

12
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 12 Preliminary Observations The threshold is set so that there is a very low probability of false alarm Due to the use of M samples the transition band is small, about 2 dB. –So we can approximate CCA-ED with an idealized detector that detects above the threshold and does not detect below the threshold –The simulations will use the actual probability of detection curve, but the results are not that much different that those of an ideal detector

13
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 13 Significant Interference To simplify the analysis we will specify a level of interference that we consider significant This is clearly an approximation It is used to allow for a simplified analysis to gain insight into the operation of the energy detector for CCA We define significant interference if the interference is 10 dB greater than the noise floor For a 20 MHz system this means,

14
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 14 Simple Coexistence Scenario An y network and an h network –Each network only two stations Both systems 20 MHz bandwidth Co-channel operation Path loss model No shadow fading model y performing CCA-ED –Due to symmetry the same conclusions should apply to LBT in h

15
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 15 Simple Coexistence Scenario Simple two-node networks y yTXTransmitter yRXReceiver h hTXTransmitter hRXReceiver

16
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 16 Definition of Interference Events Event LabelDescription Iy2hSignificant interference from y at h Ih2ySignificant interference from h at y

17
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 17 Goal of CCA If Iy2h=Ih2y=False then CCA should be True –We would like CCA to be True if a resulting transmission would not cause in interference and would also be successfully received at its destination If Iy2h=True or Ih2y=True then CCA should be False –We would like CCA to be False if a resulting transmission would result in interference or if the message would not be properly received at its destination

18
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 18 Enumeration of Error Events Error 1 –Iy2h = False and Ih2y= False and CCA = False –In this case 11y could have completed a transmission without jamming 16h but did not since the channel appeared busy –This results in lower throughput for 11y –This is the exposed node problem [4] Error 2 –Iy2h = True and Ih2y = False and CCA = True –11y transmits and jams 16h –This results in lower 16h throughput –This is the famous hidden node problem [5]

19
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 19 Error Events Error 3 –Iy2h = False and Ih2y = True and CCA = True –11y transmits but due to interference it is unsuccessful –Does not harm 16h transmission –This is not a big deal. If CCA were False 11y would not have transmitted anyway Error 4 –Iy2h = True and Ih2y = True and CCA = True –This is a combination of 2 and 3. The one that counts is the jamming of 16h

20
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 20 Error Events of Interest Exposed Node Event –Iy2h = False and Ih2y= False and CCA = False –CCA believes the channel is busy but if the STA did transmit it would not cause interference and would have been successfully received Hidden Node Event –Iy2h = True and CCA = True –CCA believes the channel is not busy so the station transmits and jams 16h We want to evaluate under what conditions these two events occur

21
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 21 Successful CCA-ED y CCA- ED detects busy channel If CCA-ED had not detected 16h then 11y would have jammed 16h

22
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 22 Exposed Node Event y CCA- ED detects busy channel However if 11y had transmitted no error would have occurred Based on current ED value this is an unlikely event

23
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 23 Hidden Node Event y CCA- ED does not detect a busy channel 11y transmits and jams 16h

24
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 24 Effect of Varying the CCA-ED Threshold If the CCA-ED is reduced to a lower value (more sensitive CCA-ED) then, –The false alarm rate will increase –The exposed node probability will increase –The hidden node probability will decrease So there is a natural tradeoff between –False alarm rate & exposed node probability And –Hidden node probability

25
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 25 Simulation Parameters y –Fixed TX Power = 40 dBm –Portable TX Power = 20 dBm –0 dBi Antenna –Noise Figure = 10 dB –Receiver Sensitivity = -82 dBm –CCA-ED Threshold = -72 dBm h –Fixed TX Power = 40 dBm –Portable TX Power = 20 dBm –0 dBi Antenna –Noise Figure = 10 dB –Receiver Sensitivity = -80 dBm

26
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 26 Simulation Parameters Path Loss Model –Single break point at 100 m as in [6] –Free space out to 100 m –Exponent of 3.5 beyond 100m

27
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 27 Simulation Procedure Place an y and an base station –The separation between these base station is varied Randomly place a single y client station within the coverage area –Coverage area depends on transmit power of client station (which is less than or equal to power of base station) and path loss model Randomly place a single h client station within the coverage area –Coverage area depends on transmit power of client station (which is less than or equal to power of base station) and path loss model Ran 10 5 trials for each 11y/16h base station separation

28
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 28 Simulation Scenarios Four Possible Scenarios y BS performs CCA-ED while 16h BS is TX y BS performs CCA-ED while 16h STA is TX y STA performs CCA-ED while 16h BS is TX y STA performs CCA-ED while 16h STA is TX Evaluate exposed node probability and hidden node probability in each of these four cases

29
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 29 Statement About Probabilities These probabilities are dependent on spatial randomness and not temporal randomness If the stations do not move the situation will not change You cannot necessarily solve any problems by waiting for the situation to change because it is a spatial process and not a temporal process

30
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 30 Two Sets of Simulations First set of simulations –Fixed stations for both 11y and 16h –High power fixed base stations –High power fixed client stations Second set of simulations –Portable client stations for both 11y and 16h –High power fixed base stations –Low power portable client stations

31
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 31 Fixed – Scenario 1

32
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 32 Fixed – Scenario 2

33
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 33 Fixed – Scenario 3

34
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 34 Fixed – Scenario 4

35
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 35 Observations The exposed node probability is low in all four cases –This is due to the rather high CCA-ED threshold of -72 dBm The hidden node probability could get quite high –It depended on which of the four cases we are considering Once the base stations are separated enough, as one would expect, the hidden node probability drops to zero

36
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 36 Portable – Scenario 1

37
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 37 Portable – Scenario 2

38
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 38 Portable – Scenario 3

39
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 39 Portable – Scenario 4

40
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 40 Observations Exposed node probabilities are still very low The hidden node problem is much worse in the case of the base station performing CCA-ED –The Base station CCA-ED region is a subset of the interference region –The Portable station CCA-ED region is larger than the interference region For some values of base station separations the hidden node probability is one!

41
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 41 Another Set of Simulations Modified the CCA-ED threshold to -82 dBm which makes the CCA-ED much more sensitive P FA is still quite low The sensing region is comparable to the interference region –A set of simulations for fixed high power clients –A set of simulations for portable low power clients

42
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 42 Fixed – Scenario 1

43
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 43 Fixed – Scenario 2

44
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 44 Fixed – Scenario 3

45
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 45 Fixed – Scenario 4

46
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 46 Portable – Scenario 1

47
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 47 Portable – Scenario 2

48
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 48 Portable – Scenario 3

49
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 49 Portable – Scenario 4

50
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 50 Observations Increased CCA-ED sensitivity resulted in –Increase in exposed node probability –Decrease in hidden node probability The results were better in the fixed (equal power) case In the portable (unequal power) case the was still a significant problem when the high-power 11y base station was performing CCA-ED and the low-power 16h client was transmitting

51
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 51 Conclusions A method for evaluating the CCA-ED was introduced With the current CCA-ED threshold the exposed node probability is quite near-zero With the current CCA-ED threshold the hidden node probability for CCA-ED can be quite high This hidden node cannot be addressed by using RTS- CTS since h cannot send an y packet This high hidden node probability can result in significant interference By symmetry LBT in h will have similar issues

52
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 52 Future Work If the 11y/16h/19 team decides on different simulation parameter the simulation can be updated Decide what is the right value for Significant Interference This work could be integrated into a system level simulation

53
doc.: IEEE /0010r0 Submission May 2007 Steve Shellhammer, QualcommSlide 53 References 1.Steve Shellhammer, Performance of the Power Detector, IEEE /75r0, May Draft IEEE y, D2.0, March Paul Piggin, Parameters for simulation of Wireless Coexistence in the US and Canada 3.65GHz band, IEEE /11r0, April http://en.wikipedia.org/wiki/Exposed_terminal_problemhttp://en.wikipedia.org/wiki/Exposed_terminal_problem 5.http://en.wikipedia.org/wiki/Hidden_terminal_problemhttp://en.wikipedia.org/wiki/Hidden_terminal_problem 6.V. Erceg, et. al., Channel Models for Fixed Wireless Applications, IEEE c-01/29, January 2001

Similar presentations

© 2016 SlidePlayer.com Inc.

All rights reserved.

Ads by Google