Error analysis methods for the fixed-point implementation of linear systems

In this paper we propose to perform a complete error analysis of a fixed-point implementation of any linear system described by data-flow graph. The system is translated to a matrix-based internal representation that is used to determine the analytical errors-to-output relationship. The error induced by the finite precision arithmetic (for each sum-of-product) of the implementation propagates through the system and perturbs the output. The output error is then analysed with three different point of view: classical statistical approach (errors modeled as noises), worst-case approach (errors modeled as intervals) and probability density function. These three approaches allow determining the output error due to the finite precision with respect to its probability to occur and give the designer a complete output error analysis. Finally, our methodology is illustrated with numerical examples.

[1]  Joseph Cavanagh,et al.  Computer Arithmetic and Verilog HDL Fundamentals , 2009 .

[2]  Jianwen Zhu,et al.  An analytical approach for dynamic range estimation , 2004, Proceedings. 41st Design Automation Conference, 2004..

[3]  Nicholas Hale,et al.  An Algorithm for the Convolution of Legendre Series , 2014, SIAM J. Sci. Comput..

[4]  Z. Zhao,et al.  On the generalized DFIIt structure and its state-space realization in digital filter implementation , 2004, IEEE Transactions on Circuits and Systems I: Regular Papers.

[5]  James F. Whidborne,et al.  Digital Controller Implementation and Fragility , 2001 .

[6]  Philippe Chevrel,et al.  A Unifying Framework for Finite Wordlength Realizations , 2007, IEEE Transactions on Circuits and Systems I: Regular Papers.

[7]  Lajos Gazsi,et al.  Explicit formulas for lattice wave digital filters , 1985 .

[8]  Daniel Ménard,et al.  A case study of the stochastic modeling approach for range estimation , 2010, 2010 Conference on Design and Architectures for Signal and Image Processing (DASIP).

[9]  James Demmel,et al.  IEEE Standard for Floating-Point Arithmetic , 2008 .

[10]  Olivier Sentieys,et al.  A methodology for evaluating the precision of fixed-point systems , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[11]  A. Fettweis.  Wave digital filters: Theory and practice , 1986, Proceedings of the IEEE.

[12]  Octavio Nieto-Taladriz,et al.  Improved Interval-Based Characterization of Fixed-Point LTI Systems With Feedback Loops , 2007, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[13]  John G. Proakis,et al.  Digital Signal Processing: Principles, Algorithms, and Applications , 1992 .

[14]  Clifford T. Mullis,et al.  Synthesis of minimum roundoff noise fixed point digital filters , 1976 .

[15]  S. Boyd,et al.  On computing the worst-case peak gain of linear systems , 1992 .

[16]  Andrei Banciu,et al.  A Stochastic Approach For The Range Evaluation , 2012 .

[17]  John von Neumann,et al.  First draft of a report on the EDVAC , 1993, IEEE Annals of the History of Computing.

[18]  M. Gevers,et al.  Parametrizations in control problems , 1993 .

[19]  Bernard Widrow,et al.  Quantization Noise: Roundoff Error in Digital Computation, Signal Processing, Control, and Communications , 2008 .

[20]  Siegfried M. Rump,et al.  New Results on Verified Inclusions , 1985, Accurate Scientific Computations.

[21]  Jong-Seong Kang,et al.  GENERALIZED CONVOLUTION OF UNIFORM DISTRIBUTIONS , 2010 .

[22]  Thibault Hilaire,et al.  Formatting Bits to Better Implement Signal Processing Algorithms , 2014, PECCS.

[23]  F. Killmann,et al.  A Note on the Convolution of the Uniform and Related Distributions and Their Use in Quality Control , 2001 .

[24]  Anastasia Volkova,et al.  Reliable Fixed-Point Implementation of Linear Data-Flows , 2016, 2016 IEEE International Workshop on Signal Processing Systems (SiPS).

[25]  J. Aplevich.  Implicit Linear Systems , 1991 .

[26]  David M. Bradley,et al.  On the Distribution of the Sum of n Non-Identically Distributed Uniform Random Variables , 2002, math/0411298.

[27]  Christoph Quirin Lauter,et al.  Determining fixed-point formats for a digital filter implementation using the worst-case peak gain measure , 2015, 2015 49th Asilomar Conference on Signals, Systems and Computers.

[28]  William Feller,et al.  An Introduction to Probability Theory and Its Applications , 1951 .

[29]  Christoph Quirin Lauter,et al.  Reliable Evaluation of the Worst-Case Peak Gain Matrix in Multiple Precision , 2015, 2015 IEEE 22nd Symposium on Computer Arithmetic.