Sparse Signal Recovery from Quadratic Measurements via Convex Programming
نویسندگان
چکیده
In this paper we consider a system of quadratic equations |〈zj ,x〉|2 = bj , j = 1, ...,m, where x ∈ R is unknown while normal random vectors zj ∈ R and quadratic measurements bj ∈ R are known. The system is assumed to be underdetermined, i.e., m < n. We prove that if there exists a sparse solution x i.e., at most k components of x are non-zero, then by solving a convex optimization program, we can solve for x up to a multiplicative constant with high probability, provided that k ≤ O( √ m logn ). On the other hand, we prove that k ≤ O(log n√m) is necessary for a class of naive convex relaxations to be exact.
منابع مشابه
Performance Analysis of Joint-Sparse Recovery from Multiple Measurements and Prior Information via Convex Optimization
We address the problem of compressed sensing with multiple measurement vectors associated with prior information in order to better reconstruct an original sparse matrix signal. l2,1− l2,1 minimization is used to emphasize co-sparsity property and similarity between matrix signal and prior information. We then derive the necessary and sufficient condition of successfully reconstructing the orig...
متن کاملCS168: The Modern Algorithmic Toolbox Lecture #14: Linear and Convex Programming, with Applications to Sparse Recovery
Recall the setup in compressive sensing. There is an unknown signal z ∈ R, and we can only glean information about z through linear measurements. We choose m linear measurements a1, . . . , am ∈ R. “Nature” then chooses a signal z, and we receive the results b1 = 〈a1, z〉, . . . , bm = 〈am, z〉 of our measurements, when applied to z. The goal is then to recover z from b. Last lecture culminated i...
متن کاملCS168: The Modern Algorithmic Toolbox Lecture #18: Linear and Convex Programming, with Applications to Sparse Recovery
Recall the setup in compressive sensing. There is an unknown signal z ∈ R, and we can only glean information about z through linear measurements. We choose m linear measurements a1, . . . , am ∈ R. “Nature” then chooses a signal z, and we receive the results b1 = 〈a1, z〉, . . . , bm = 〈am, z〉 of our measurements, when applied to z. The goal is then to recover z from b. Last lecture culminated i...
متن کاملPerformance Limits of Sparse Support Recovery Algorithms
Compressed Sensing (CS) is a signal acquisition approach aiming to reduce the number of measurements required to capture a sparse (or, more generally, compressible) signal. Several works have shown significant performance advantages over conventional sampling techniques, through both theoretical analyses and experimental results, and have established CS as an efficient way to acquire and recons...
متن کاملSignal recovery from incomplete and inaccurate measurements via ROMP
We demonstrate a simple greedy algorithm that can reliably recover a vector v ∈ R from incomplete and inaccurate measurements x = Φv + e. Here Φ is a N × d measurement matrix with N ≪ d, and e is an error vector. Our algorithm, Regularized Orthogonal Matching Pursuit (ROMP), seeks to close the gap between two major approaches to sparse recovery. It combines the speed and ease of implementation ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید
ثبت ناماگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید
ورودعنوان ژورنال:
- SIAM J. Math. Analysis
دوره 45 شماره
صفحات -
تاریخ انتشار 2013