A novel specification based test pattern generation using genetic algorithm and wavelets

Abstract

For analog and mixed signal circuits, the traditional tests are time consuming and also most expensive in terms of both test development and test implementation costs .So for testing these ICs new testing strategies are investigated. In this paper, a novel test generation methodology for the detection of faults in analog cores is proposed. A transient signal is taken as input stimulus for the circuit under test and the output response is analysed using wavelets. Specifications of the circuit are mapped to the performance space and the specifications are checked implicitly. The proposed fault oriented test generator also computes the optimal test patterns based on genetic algorithm.

Evolutionary Computation in pre-silicon verification of complex microprocessors

“…Kalpana and Gunavathi [3] used a GA to evolve test patterns for detecting faults in analog circuits. Faults were present when the response of the CUT differed by some amount from that of a fault-free circuit.…”

Selecting input current waveforms using a hardware testing implementation incorporated in FPGAs

“…Among others, the root-mean-square (RMS) value of the supply current and the magnitude and phase components of its spectrum (Fourier transform) have been used [11]. Another approach is based on the use of the wavelet transform, which resolves a signal in both time and frequency simultaneously [12][13][14]. It gives a better approximation of a transient current waveform than the Fourier transform for a certain frequency of the signal.…”

Input stimulus comparison using an adaptive FPGA-based testing system

“…Another approach is based on the use of the wavelet transform, which resolves a signal in both time and frequency simultaneously [10][11][12]. It gives a better approximation of a transient current waveform than the Fourier transform for a certain frequency of the signal.…”