Optimal filtering versus regularization methods in the Fourier precompensation based proximity neurocorrection in Electron Beam Lithography
نویسنده
چکیده
In this paper proximity effects correction in Electron Beam Lithography by means of an artificial neural network is presented. Supporting approximations to cope with negative doses inherent in Gibbs oscillations which occur from step-like function representation in the Fourier space are introduced. Miller regularization theory as better alternative to Tikhonov one is presented. Optimal filtering with prolate spheriodal wave functions ( PSWFs ) is exploited and as an ultimate solution for neurocorrection is proposed. A quasi-autoassociative network with Optimal Brain Surgeon ( OBS ) pruning method is suggested. The error analysis for the generalisation process as a function of the pruning process will be discussed. The possible hardware implementations of the neurocrrector will be proposed.
منابع مشابه
Computer Simulation from Electron Beam Lithography to Optical Lithography
Simulation of electron beam lithography and optical lithography has been combined to investigate the influence of a distorted photomask feature on final photoresist image. Unlike the previous optical lithography simulation which was based on ideal mask design, the combined simulation has shown that mask distortion due to electron proximity effect play an important role in worsening the optical ...
متن کاملOptimal Design and Fabrication of Fine Diffractive Optical Elements Using Proximity Correction with Electron-beam Lithography
متن کامل
Binary mask optimization for inverse lithography with partially coherent illumination.
Recently, a set of generalized gradient-based optical proximity correction optimization methods have been developed to solve for the inverse lithography problem under coherent illumination. Most practical lithography systems, however, operate under partially coherent illumination. This paper focuses on developing gradient-based binary mask optimization methods that account for the inherent nonl...
متن کاملA Quantitative Comparison between Helium Ion and Electron Beam Lithography on PMMA Resist
Helium ion beam lithography (HIBL), an emerging technique that uses a sub-nanometre focused beam of helium ions to expose resist, has introduced an alternative to electron beam lithography (EBL) to extend beyond existing minimum feature sizes. HIBL has several advantages over EBL, including a higher patterning resolution due to a smaller spot size [1] and a reduced proximity effect due to low i...
متن کاملA New Similarity Measure Based on Item Proximity and Closeness for Collaborative Filtering Recommendation
Recommender systems utilize information retrieval and machine learning techniques for filtering information and can predict whether a user would like an unseen item. User similarity measurement plays an important role in collaborative filtering based recommender systems. In order to improve accuracy of traditional user based collaborative filtering techniques under new user cold-start problem a...
متن کامل