sublinearfourier sampling off the grid · sublinearfourier sampling off the grid petros...
TRANSCRIPT
![Page 1: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/1.jpg)
What's the Frequency, Kenneth?: Sublinear Fourier Sampling Off the Grid
Petros Boufounos (MERL), Volkan Cevher (EPFL),Anna Gilbert (UMich), Yi Li (UMich), Martin Strauss (UMich)
with application to bearing estimation
![Page 2: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/2.jpg)
Intuitive problem description
Input: signal = linear combination of k ``frequencies’’
Sample signal at roughly k positions in time
Output: k frequencies + coefficients in time comparable to # samples
Discrete setting: k frequencies in some finite group, usually
![Page 3: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/3.jpg)
History
Cooley‐Tukey[1960s]: Fast Fourier Transform
Sublinear Fourier algorithmsKushilevitz‐Mansour[1993] (Boolean cube; poly( , log ))Mansour[1995] ( , prime; poly( , log ))Gilbert‐Guha‐Indyk‐Muthukrishnan‐Strauss[2002]Gilbert‐Muthukrishnan‐Strauss[2005] ( log log )Iwen[2010] (deterministic, log )Akavia[2010] (deterministic)
Hassenieh‐Indyk‐Katabi‐Price[2012] ( log log )
![Page 4: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/4.jpg)
More precisely,
Hassenieh‐Indyk‐Katabi‐Price[2012]
Return coeffs and frequencies such that
and
in time log log
Randomized algorithm: succeed with constant probability overchoice of samples
![Page 5: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/5.jpg)
Great success but…
Sparse Fourier sampling algorithms tremendously successful but address a certain discrete model problem
Very specific specialized set‐up
Not such a good approximation for analog (real) world
Discrete approximations can degrade the sparsity for sparse analog signals
![Page 6: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/6.jpg)
Examples of world signals
AM/FM radio signals
Musical instruments
Doppler radar
Analog signal generator
![Page 7: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/7.jpg)
New model
Find s.t. coefficients are significant:
Input signal = exponential polynomial + noisefrequencies are contained in , ∪ ,
minimum frequency resolution
![Page 8: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/8.jpg)
Nyquist‐Shannon Sampling
Sample interval has to be small to distinguish two high frequencies
Sample duration should be large to distinguish a low frequency from 0
≃1⋅ ≃
1
![Page 9: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/9.jpg)
Main result
Theorem: there is a distribution on points (in time) s.t. w.h.p. for each input signal, return list
For each sign. coefficient , and
# samples, running time =
Sample duration/extent =
![Page 10: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/10.jpg)
Application: bearing of sources
Determine angles of sources transmitting sound
Minimum angular resolution
aperture
![Page 11: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/11.jpg)
Application: bearing of sources
Receivers on the x‐axis: ⋅ cos
Find sources with angles in ,For in this range, cos has a minimum separation of Θ
Rotate the receiver array
Ambiguities
![Page 12: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/12.jpg)
Algorithm
IdentifyIsolate frequencies by hashing => multiply samples by filter weightsRead off bits (up to desired resolution) by dilation + hashingGenerate list of candidate frequencies
EstimateMedian of values in hashed buckets for specific freqs. in list
1. Need extremely good filter for hashing2. Non‐iterative
Similar to [HIKP 2012a], but with simple bit‐testing
![Page 13: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/13.jpg)
1. Continuous on 2. Fourier transform has finite support on 3. Approximates , well for parameter
small transition region
A good filter/hash
pass regionsmall outside pass + transition region,
Example: Dolph‐Chebyshev convolved with ,
log1
![Page 14: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/14.jpg)
Groups and dual groups
Sample at equidistant discrete points ↔ Frequencies in
Dilation and translation
![Page 15: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/15.jpg)
Distribution = random uniform spacing + bit testers
Sample duration
Random uniform spacing ∈ ,
d/2 d/4
Total # samples = log ⋅ log ⋅ log
log1
kernel’s spectrum size
# bits # repetitions
![Page 16: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/16.jpg)
Whither iterative algorithm(s)?
[GMS 2005]: lousy filter but same # buckets in each iteration, # iterations depends on dynamic range of signal, improve est. each iteration
[GLPS 2010, HIKP 2012b]: iterative, # buckets decreases in each iteration
Wider buckets => lose resolution
Can’t subtract recovered frequencies (easily)
Would need k * longer duration
![Page 17: SublinearFourier Sampling Off the Grid · SublinearFourier Sampling Off the Grid Petros Boufounos(MERL), Volkan Cevher ... iterations depends on dynamic range of signal, improve est](https://reader036.vdocuments.net/reader036/viewer/2022062607/6021b806b7e0fc653431a953/html5/thumbnails/17.jpg)
Open problems
Lower boundsSample duration (aperture size)Number of samples
Iterative vs. non‐iterative algorithm
Simple discretization
Error Metric(s)
Thank you!