Abstract :
A new solution to the numerical instability of fast recursive least squares adaptive filter algorithms is presented. Using interval arithmetic, an error analysis is effectively performed in parallel with the computations of the algorithm. If necessary the algorithm with be rescued, preventing divergence. This is shown in software simulations and in a hardware implementation of an adaptive equaliser.
Keywords :
adaptive filters; digital arithmetic; digital filters; equalisers; error analysis; filtering and prediction theory; least squares approximations; signal processing; DSP implementation; adaptive equaliser; adaptive filters; error analysis; fast-recursive algorithms; hardware implementation; interval arithmetic; numerical stability; numerically robust; software simulations;