scholarly journals Superresolution without separation

2017 ◽  
Vol 7 (1) ◽  
pp. 1-30 ◽  
Author(s):  
Geoffrey Schiebinger ◽  
Elina Robeva ◽  
Benjamin Recht
Keyword(s):  
Polynomial Interpolation ◽  
Point Sources ◽  
Finite Collection ◽  
Finite Rate ◽  
Weighted Version ◽  
Close Point ◽  
Root Finding ◽  
Proof Techniques ◽  
Finite Rate Of Innovation ◽  
Polynomial Root Finding

Abstract This article provides a theoretical analysis of diffraction-limited superresolution, demonstrating that arbitrarily close point sources can be resolved in ideal situations. Precisely, we assume that the incoming signal is a linear combination of $M$ shifted copies of a known waveform with unknown shifts and amplitudes, and one only observes a finite collection of evaluations of this signal. We characterize properties of the base waveform such that the exact translations and amplitudes can be recovered from $2M+1$ observations. This recovery can be achieved by solving a weighted version of basis pursuit over a continuous dictionary. Our analysis shows that $\ell_1$-based methods enjoy the same separation-free recovery guarantees as polynomial root finding techniques, such as de Prony’s method or Vetterli’s method for signals of finite rate of innovation. Our proof techniques combine classical polynomial interpolation techniques with contemporary tools from compressed sensing.

scholarly journals Computing one billion roots using the tangent Graeffe method

2020 ◽  
Vol 54 (3) ◽  
pp. 65-85
Author(s):  
Joris van der Hoeven ◽  
Michael Monagan
Keyword(s):  
Finite Field ◽  
Finite Fields ◽  
Polynomial Interpolation ◽  
Polynomial Algorithms ◽  
Polynomial Factorization ◽  
Root Finding ◽  
Multiplication Algorithm ◽  
Sparse Polynomial ◽  
Polynomial Root Finding ◽  
Intel Xeon

Let p be a prime of the form p = σ2 k + 1 with σ small and let F p denote the finite field with p elements. Let P ( z ) be a polynomial of degree d in F p [ z ] with d distinct roots in F p . For p =5 · 2 55 + 1 we can compute the roots of such polynomials of degree 10 9 . We believe we are the first to factor such polynomials of size one billion. We used a multi-core computer with two 10 core Intel Xeon E5 2680 v2 CPUs and 128 gigabytes of RAM. The factorization takes just under 4,000 seconds on 10 cores and uses 121 gigabytes of RAM. We used the tangent Graeffe root finding algorithm from [27, 19] which is a factor of O (log d ) faster than the Cantor-Zassenhaus algorithm. We implemented the tangent Graeffe algorithm in C using our own library of 64 bit integer FFT based in-place polynomial algorithms then parallelized the FFT and main steps using Cilk C. In this article we discuss the steps of the tangent Graeffe algorithm, the sub-algorithms that we used, how we parallelized them, and how we organized the memory so we could factor a polynomial of degree 10 9 . We give both a theoretical and practical comparison of the tangent Graeffe algorithm with the Cantor-Zassenhaus algorithm for root finding. We improve the complexity of the tangent Graeffe algorithm by a factor of 2. We present a new in-place product tree multiplication algorithm that is fully parallelizable. We present some timings comparing our software with Magma's polynomial factorization command. Polynomial root finding over smooth finite fields is a key ingredient for algorithms for sparse polynomial interpolation that are based on geometric sequences. This application was also one of our main motivations for the present work.

scholarly journals DoA Estimation Using Neural Tangent Kernel under Electromagnetic Mutual Coupling

Electronics ◽  
2021 ◽  
Vol 10 (9) ◽  
pp. 1057
Author(s):  
Qifeng Wang ◽  
Xiaolin Hu ◽  
Xiaobao Deng ◽  
Nicholas E. Buris
Keyword(s):  
Mutual Coupling ◽  
Deep Neural Network ◽  
Doa Estimation ◽  
Antenna Element ◽  
Root Finding ◽  
Multiple Signal ◽  
Estimation Problems ◽  
Limiting Case ◽  
Signal Doa Estimation ◽  
Polynomial Root Finding

Antenna element mutual coupling degrades the performance of Direction of Arrival (DoA) estimation significantly. In this paper, a novel machine learning-based method via Neural Tangent Kernel (NTK) is employed to address the DoA estimation problem under the effect of electromagnetic mutual coupling. NTK originates from Deep Neural Network (DNN) considerations, based on the limiting case of an infinite number of neurons in each layer, which ultimately leads to very efficient estimators. With the help of the Polynomial Root Finding (PRF) technique, an advanced method, NTK-PRF, is proposed. The method adapts well to multiple-signal scenarios when sources are far apart. Numerical simulations are carried out to demonstrate that this NTK-PRF approach can handle, accurately and very efficiently, multiple-signal DoA estimation problems with realistic mutual coupling.

Sign in / Sign up

Export Citation Format

Share Document