A Maximum-Length Sequence (MLS) is a periodic two-level signal of length P = 2 N – 1, where N is an integer and P is the periodicity, which yields the impulse response of a linear system under circular convolution. The impulse response is extracted by the deconvolution of the system's output when excited with an MLS signal. This chapter presents the underlying theory and how the Fast Hadamard Transform can provide a very efficient means of analysing an MLS sequence. A MATLAB implementation of… CONTINUE READING

Add a zero to the first element and reorder the averaged data according to P S . 5. Perform a Fast Walsh-Hadamard transform

Add a zero to the first element and reorder the…

Reorder the transformed data according to P L and divide by P+1

Reorder the transformed data according to P L and…

See the user guide in Appendix 1 and the Matlab code in Appendix 3 for further information about how this algorithm is implemented

See the user guide in Appendix 1 and the Matlab…

Stimulate the system with at least two successive MLS bursts and record the output. The SNR is improved by 3dB for a doubling in the number of repetitions

Stimulate the system with at least two successive…

Take the mean of all but the first burst

Take the mean of all but the first burst

The importance of approximating circular convolution is often overlooked and can yield highly inaccurate results if it is ignored

The importance of approximating circular…

The result is an estimation of a system's impulse response of length P+1. Perform an FFT to find the frequency-domain representation

The result is an estimation of a system's impulse…