Characterization of an Unintentional Wi-Fi Interference Device - PowerPoint PPT Presentation

About This Presentation
Title:

Characterization of an Unintentional Wi-Fi Interference Device

Description:

ISM band is crowded. ... is power scattered over entire ISM band. The wideband power is ... Transient Signals with a wide bandwidth to span the entire ISM band. ... – PowerPoint PPT presentation

Number of Views:28
Avg rating:3.0/5.0
Slides: 24
Provided by: ayhama
Learn more at: http://www.ece.iit.edu
Category:

less

Transcript and Presenter's Notes

Title: Characterization of an Unintentional Wi-Fi Interference Device


1
Characterization of an Unintentional Wi-Fi
Interference Device The Residential Microwave
Oven
  • Tanim M. Taher
  • Ayham Z. Al-Banna
  • Joseph L. LoCicero
  • Donald R. Ucci

Presented by Tanim M. Taher
2
Outline
  • Motivation
  • Experimental Analysis of Microwave Oven (MWO)
    signal
  • Frequency Shifting part
  • Transients
  • AM modulated Envelope of the signal
  • Model Developed
  • Simulation Results
  • Interference Mitigation
  • Conclusions
  • Ongoing and Future Work

3
Motivation (1)
4
Motivation (2)
  • ISM band is crowded.
  • MWOs were designed many years ago and they
    radiate energy in the 2.4 GHz ISM band.
  • EM waves radiated by MWO devices interfere with
    ISM devices unintentionally (no intelligence
    signal in the MWO signal).
  • Understanding the interference signal helps
    mitigation
  • Spectral signatures of MWO signals need to be
    identified.

5
Experimental Analysis of MWO Signal (1)
  • The Residential MWO signal is synchronized with
    the 60 Hz AC line cycle, and it radiates in the
    positive half cycle.
  • The MWO signal has the main characteristics
  • An AM-FM signal that is radiated for about 5-6
    ms.
  • Transient signals before and after FM signal
    (1ms).

On cycle of MWO
6
Experimental Analysis of MWO Signal (2)
  • Spectrogram shows FM nature of MWO signal.
  • The frequency sweeping is roughly sinusoidal in
    nature.

AM-FM Signal
Figure Courtesy of Stevens Institute of
Technology
7
Experimental Analysis of MWO Signal (3)
  • The AM-FM signal has a bandwidth of between 15-20
    MHz (depends on MWO).

AM-FM Signal
8
Experimental Analysis of MWO Signal (4)
  • The MWO emits wideband Transient Signals before
    and after the FM signal. Transient durations are
    around 1 ms each.
  • Figure shows wideband nature of transients.
  • Observe the high transient energy concentrated in
    frequencies near FM signal.

Transients
9
Experimental Analysis of MWO Signal (5)
  • Zero Span Measurements show Transient Signal
    durations. Observe, they exist for only about 15
    of the time during the time period of 16.67 ms.
  • Zero-span measurements at 2.46 GHz and 2.44 GHz
    over two 60 Hz cycles.
  • Transients classified as Turn-on and Turn-off

10
Experimental Analysis of MWO Signal (6)
  • The amplitude of the FM signal is not constant,
    it varies sinusoidaly!
  • The Zero-Span Measurement indicates this. So the
    FM signal is further AM modulated (AM-FM signal
    obtained).
  • Zero-span measurement at 2.455 GHz. Note the
    changing amplitude in the middle.
  • Transients are also observable before and after
    the AM-FM signal.

11
Experimental Analysis of MWO Signal (7)
  • Power Spectral Density of MWO Signal
  • Most power is concentrated over the narrow
    frequency range (15 MHz) swept by AM-FM signal.
  • There is power scattered over entire ISM band.
  • The wideband power is due to transients.

12
Model for the MWO Signal (1)
  • A model for the time-domain MWO signal was
    developed featuring its main characteristics.
  • Model Composition
  • An FM signal with instantaneous frequency
    proportional to AC line voltage.
  • The FM signal is further AM modulated forming an
    AM-FM signal. The AM amplitude, again, is
    proportional to the 60 Hz AC cycle.
  • Transient Signals with a wide bandwidth to span
    the entire ISM band.
  • Transient Signals with a narrow bandwidth with
    power concentrated in the AM-FM-swept frequency
    band.

13
Model for the MWO Signal (2)
  • Qualitative representation of the MWO signal
    model.

14
Model for the MWO Signal (3)
  • Mathematical Representation of model MWO signal.

, where T 1/fac and fac 60 Hz.
15
Model for the MWO Signal (4)
  • Mathematical Representation of model MWO signal
    (contd.).

16
Simulation Results (1)
  • The model developed was simulated. Simulated
    Spectrograms and Power Spectral Density plots
    were obtained and compared to the experimental
    plots.
  • The simulations were performed at frequencies
    much lower than the ISM band for computational
    convenience. The model is scalable to higher
    frequencies, and the spectral signatures are
    preserved.

17
Simulation Results (2)
Power Spectral Densities
Experimental PSD
Simulated at 100 KHz carrier frequency
Simulated at 1 MHz carrier frequency (parameters
different from first one)
18
Simulation Results (3)
Spectrograms
Experimental Spectrogram
Simulated at 100 KHz carrier frequency
Simulated at 1 MHz carrier frequency (parameters
different from first one)
19
Interference Mitigation (1)
  • The transients of the MWO signal interfere with
    all Wi-Fi channels, however, only for 15 of
    time. Since the transients are synchronized to
    the 60 Hz line cycle, we can predict the
    transient times and avoid interference by not
    transmitting at those times.
  • The AM-FM signal is narrowband and interferes
    with only some IEEE 802.11 channels (like channel
    11). In such a case, the Wi-Fi channel can be
    changed to another channel outside the AM-FM
    signals frequency band (like channel 1).

20
Interference Mitigation (2)
  • Data transmission using 802.11 channel 1 (shaded
    areas are transient locations)

21
Conclusions
  • MWO signal was thoroughly studied and
    characterized.
  • A novel model for the MWO signal was developed.
  • Simulation and experimental results supported the
    theoretical model.
  • Interference mitigation techniques were proposed.

22
Ongoing Future Work
  • Investigating random variations in the MWO signal
    signature
  • Refining the model to include the random aspects
    of MWO signal behavior
  • Further research on the proposed interference
    mitigation techniques

23
Thank you!Questions?
Write a Comment
User Comments (0)
About PowerShow.com