Abstract
Constant-modulus algorithms try to separate linear mixtures of sources with modulus 1. We study the identifiability of this problem: how many samples are needed to ensure that in the noiseless case we have a unique solution? For finite-alphabet (L-PSK) sources, finite sample identifiability can hold only with a probability close to but not equal to 1. In a previous paper, we provided a sub-exponentialy decaying upper bound on the probability of non-identifiability. Here, we provide an improved exponentialy decaying upper bound, based on Chernoff bounds. We show that under practical assumptions, this upper bound is much tighter than previously known bounds.
| Original language | English |
|---|---|
| Pages (from-to) | 329-332 |
| Number of pages | 4 |
| Journal | Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing |
| Volume | 4 |
| State | Published - 2003 |
| Externally published | Yes |
| Event | 2003 IEEE International Conference on Accoustics, Speech, and Signal Processing - Hong Kong, Hong Kong Duration: 6 Apr 2003 → 10 Apr 2003 |