HRR profile estimation using SLIM. Issue 4 (1st April 2019)

• Record Type: Journal Article
• Authors: Addabbo, Pia; Aubry, Augusto; De Maio, Antonio; Pallotta, Luca; Ullo, Silvia Liberata
• Is Part Of: IET radar, sonar & navigation. Volume 13:Issue 4(2019)
• Page Start: 512
• Page End: 521
• Publication Date: 2019-04-01
• Subjects: learning (artificial intelligence) -- Bayes methods -- minimisation -- least squares approximations -- iterative methods -- maximum likelihood estimation -- radar resolution; receiver -- radar high‐range‐resolution profile reconstruction -- transmitter -- continuous feedback -- coordinated feedback -- lq‐norm constraint -- optimised frequency hopping patterns -- precise HRR reconstruction -- least‐squares approach -- BIC -- Bayesian information criterion -- actual active scatterers -- range cell -- design stage -- interference power level -- regularised maximum‐likelihood estimation paradigm -- regularised minimisation approach -- iterative minimisation paradigm -- sparse learning -- narrow instantaneous bandwidth -- stepped‐frequency waveforms -- cognitive paradigm -- target range profile estimation capabilities -- iterative adaptive approach -- HRR profile recovery -- SLIM‐based procedure; Signal processing -- Periodicals; Radar -- Periodicals; Sonar -- Periodicals; Electronics in navigation -- Periodicals; Navigation -- Periodicals; 621.3848
• DOI: 10.1049/iet-rsn.2018.5102
• Languages: English
• ISSNs: 1751-8784

Abstract

Abstract : In this study, authors address high‐range‐resolution (HRR) profile reconstruction, when stepped‐frequency waveforms are eventually used to maintain a narrow instantaneous bandwidth, resorting to the sparse learning via iterative minimisation (SLIM) paradigm, a regularised minimisation approach with an l q ‐norm constraint (for 0 < q ≤ 1 ), providing a variant to the original method. Particularly, the proposed method resorts to the regularised maximum‐likelihood estimation paradigm including a term promoting the sparsity of the profile and related to the l q ‐norm of the vector containing the scatterers' reflectivities. A priori information on the interference power level is also accounted for, at the design stage, and, assuming that each range cell under test contains at most one scatterer, the actual active scatterers composing the target are determined by exploiting the Bayesian information criterion (BIC). BIC is also used to automatically select the optimised q, so as to make the procedure adaptive with respect to q . Once the location of the active scatterers has been determined, a least‐squares approach is also used to obtain even more precise HRR reconstruction. Furthermore, an efficient algorithm to define optimised frequency hopping patterns, in the presence of a continuous and coordinated feedback between the transmitter and receiver, is presented and assessed. The carried out analysis shows that the SLIM‐based procedure presents higher accuracy in the HRR profile recovery than other widely used techniques, i.e. the iterative adaptive approach (IAA). Moreover, results demonstrate that the target range profile estimation capabilities are enhanced, both for SLIM and IAA, when the cognitive paradigm is employed.