Andre Marguinaud

Nonrecursive synthesis of FIR filters for approximate processing

In approximate and real time processing, one encounters the problem of making efficient use of the limited processing power available. Furthermore it can be desirable to enable a system to react dynamically to a change in requirements. For digital filters this implies the possibility of calculating filter coefficients on the fly with low complexity algorithms. Such an algorithm is presented for the design of linear phase FIR lowpass filters. It has the additional property that subsets of coefficients of one filter constitute by themselves filters of reduced stop band attenuation and/or lower bandwidth reduction.

Méthode de calcul rapide de convolution entre deux séquences dont l’une est binaire

RésuméCet article présente une méthode de calcul rapide de produits de convolution apériodique entre deux séquences de valeurs numériques dans le cas où l’une des séquences est composée d’éléments binaires de valeurs +1 et -1. La méthode utilise une décomposition de la séquence non binaire en segments élémentaires et une combinaison des produits polynomiaux issus de cette décomposition. Dans ces conditions les seules opérations élémentaires sont des additions et soustractions, et la méthode présentée réduit le nombre de ces opérations par un facteur 4 à 10 par rapport au calcul direct, pour des longueurs de séquences comprises entre 100 et 10000 éléments.AbstractThis article presents a method for the fast calculation of the convolution between two numerical sequences when one of the sequence is made up of binary elements with values +1 and -1. The method is based on the decomposition of the non binary sequence into segments and the combination of the polynomial products resulting from this decomposition. Under these conditions, the only operations needed are additions and substructions. Compared to the direct calculation, this method involves a reduced number of such operations and results in the order of 4 to 10-fold reductions in calculation load for sequence lengths ranging from 100 to 1 000.

Maximum Likelihood Processing of Antenna Array Signals

The antenna array considered here is composed of m receiving elements whose output signals are analytically digitized. Covariance matrix estimator and the correlation indices of its estimated entries are derived from the Maximum Likelihood Principle (ML). Practical methods for direction finding and Jammer cancellation can be designed to approach the optimum deduced from ML. Implementation complexity is reduced by suitable choice of dimensioning of data representation during processing.