Pattern synthesis of sparse linear array by off‐grid Bayesian compressive sampling. Issue 25 (1st December 2015)

• Record Type: Journal Article
• Authors: Lin, Jincheng; Ma, Xiaochuan; Yan, Shefeng; Jiang, Li
• Is Part Of: Electronics letters. Volume 51:Issue 25(2015)
• Page Start: 2141
• Page End: 2143
• Publication Date: 2015-12-01
• Subjects: compressed sensing -- array signal processing -- Bayes methods -- mean square error methods -- learning (artificial intelligence); sparse linear array pattern synthesis -- off‐grid Bayesian compressive sampling -- OG pattern synthesis algorithm -- computational efficiency -- BCS‐based algorithm -- relevance vector machine -- normalised mean square error; Electronics -- Periodicals; 621.381
• DOI: 10.1049/el.2015.2455
• Languages: English
• ISSNs: 0013-5194

Abstract

Abstract : An off‐grid (OG) pattern synthesis algorithm for sparse non‐uniform linear arrays is presented. It is based on Bayesian compressive sampling (BCS), and the design of maximally sparse linear arrays for the given reference patterns can be obtained. The proposed algorithm novelly introduces the OG model into the pattern synthesis problem, and it makes the synthesis more accurate than the conventional BCS algorithm. Moreover, the proposed algorithm has the advantage of high computational efficiency, since the BCS‐based algorithms can be realised by the fast relevance vector machine. Numerical experiments show that the proposed algorithm has improved accuracy in terms of normalised mean square error.