Performance analysis of algorithmic noise-tolerance techniques

Author

Shim, Byonghyo ; Shambhag, N.R.

Author_Institution

Dept. of Electr. & Comput. Eng., Illinois Univ., Urbana, IL, USA

Volume

4

fYear

2003

fDate

25-28 May 2003

Abstract

In this paper, we present performance analysis of algorithmic noise-tolerance (ANT) techniques. First, we analyze the predictor and RPR based ANT schemes. Next, we present a hybrid ANT scheme which is resilient to burst errors usually occurring in a high soft-error rate (P_cr) region. For a frequency selective FIR filtering, it is shown that simulation results match well with the analytic bounds while providing about 40 dB improvement in the mean square error over a conventional DSP system at a P_cr = 10^-4. It is also shown that the proposed hybrid ANT scheme maintains its robustness in noise mitigation even in the high soft-error rate region of up to P_cr = 10^-2.

Keywords

FIR filters; burst noise; digital signal processing chips; integrated circuit noise; mean square error methods; prediction theory; ANT; RPR based schemes; algorithmic noise-tolerance techniques; analytic bounds; burst errors; frequency selective FIR filtering; mean square error; noise mitigation; performance analysis; predictor based schemes; soft-error rate region; Analytical models; Digital signal processing; Error analysis; Finite impulse response filter; Frequency; Noise robustness; Performance analysis; Power system reliability; Redundancy; Signal to noise ratio;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 2003. ISCAS '03. Proceedings of the 2003 International Symposium on

Print_ISBN

0-7803-7761-3

Type

conf

DOI

10.1109/ISCAS.2003.1205786

Filename

1205786