On gradient adaptation with unit-norm constraints

In this correspondence, we describe gradient-based adaptive algorithms with in parameter spaces that are specified by \\w\\ = 1, where \\ . \\ is any v ector norm. We provide several algorithm forms and relate them to true grad ient procedures via their geometric structures. We also give algorithms tha t mitigate an inherent numerical instability for L-2-norm-constrained optim ization tasks. Simulations showing the performance of the techniques for in dependent component analysis are provided.