Resource efficient parallel architectures for linear matrix algebra in real time adaptive control algorithms on reconfigurable logic

Parallel and systolic structures for matrix algebra algorithms have been around for quite a long time. Various implementations of different numerical techniques exist. With the advent of reconfigurable logic, especially FPGAs, a need has arisen to revisit these architectures and produce resource efficient versions of these algorithms. We have produced resource efficient parallel architectures for LU decomposition and triangular matrix inversion, keeping in view data computational rate requirements for real time control systems. These architectures decrease memory logic resources considerably and also maintain excellent clock period results. They also have the capability to be mapped over each other thereby further reducing resource usage and also providing us with the additional facility of matrix multiplication.

[1]  F. A. Seiler,et al.  Numerical Recipes in C: The Art of Scientific Computing , 1989 .

[2]  Shawki Areibi,et al.  An FPGA Implementation of the LMS Adaptive Filter for Audio Processing , 2006, 2006 IEEE International Conference on Reconfigurable Computing and FPGA's (ReConFig 2006).

[3]  Keshab K. Parhi,et al.  VLSI digital signal processing systems , 1999 .

[4]  Daniel Massicotte,et al.  A systolic architecture for Kalman-filter-based signal reconstruction using the wave pipeline method , 1998, 1998 IEEE International Conference on Electronics, Circuits and Systems. Surfing the Waves of Science and Technology (Cat. No.98EX196).

[5]  Venkatesh Krishnan,et al.  Implementation of an LMS adaptive filter on an FPGA employing multiplexed multiplier architecture , 2003, The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003.

[6]  Kai Hwang,et al.  Computer architecture and parallel processing , 1984, McGraw-Hill Series in computer organization and architecture.

[7]  George W. Irwin,et al.  Systolic Kalman filtering: an overview , 1990 .

[8]  K.R. Santha,et al.  Parallel-Pipelined Architecture for the Kalman Based Adaptive Equalizer , 2007, 2007 International Conference on Signal Processing, Communications and Networking.

[9]  DSP Co-Processing in FPGAs : Embedding High-Performance , Low-Cost DSP Functions , 2004 .

[10]  Zoran A. Salcic,et al.  A New Pipelined Systolic Array-Based Architecture for Matrix Inversion in FPGAs with Kalman Filter Case Study , 2006, EURASIP J. Adv. Signal Process..

[11]  Jenq-Neng Hwang,et al.  Systolic array designs for Kalman filtering , 1991, IEEE Trans. Signal Process..

[12]  Zoran Salcic,et al.  A fully-hardware-type maximum-parallel architecture for Kalman tracking filter in FPGAs , 1997, Proceedings of ICICS, 1997 International Conference on Information, Communications and Signal Processing. Theme: Trends in Information Systems Engineering and Wireless Multimedia Communications (Cat..

[13]  Eduardo Mario Nebot,et al.  Parallel and sequential block Kalman filtering and their implementations using systolic arrays , 1991, IEEE Trans. Signal Process..

[14]  Ali M. Reza,et al.  FPGA implementation of adaptive temporal Kalman filter for real time video filtering , 1999, 1999 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings. ICASSP99 (Cat. No.99CH36258).

[15]  Lan-Da Van,et al.  An efficient systolic architecture for the DLMS adaptive filter and its applications , 2001 .

[16]  Tze-Yun Sung,et al.  Parallel VLSI Implementation of the Kalrman Filter , 1987, IEEE Transactions on Aerospace and Electronic Systems.