Controlling the critical path in time adaptive QR/sup -1/ recursions

In signal processing, it is customary to reduce critical computation paths by introducing delays in the signal flow graph. Sometimes this is easy. More often, it is not that easy. In this paper it is shown-by example-that critical paths can be controlled systematically in the algorithmic domain, that is prior to the derivation of the flow graph. The example presented here is the adaptive inverse QR algorithm. The method, though, is not restricted to this example. It applies to all algorithms where associativity and commutativity properties can be exploited.

[1]  Marc Moonen,et al.  On the formal derivation of a systolic array for recursive least squares estimation , 1994, Proceedings of ICASSP '94. IEEE International Conference on Acoustics, Speech and Signal Processing.

[2]  Hylke W. van Dijk,et al.  Systematic exploration of the space of Jacobi algorithms , 1994, Optics & Photonics.