Studies in Statistical Signal Processing.
Abstract
Several new results in the modeling, analysis and prediction of nonstationary second order processes have been developed. They include (1) the derivation of constant-parameter lattice filters for general nonstationary processes; (2) the development of fast algorithms for adaptive filtering with fixed-order transversal filters, including modifications necessary to ensure stability and convergence with finite precision computations; (3) the invention of a new adaptive beam-forming array for separating coherent (completely correlated) signals and interference. Two Ph.D theses have been completed during this period, H. Lev-Ari, 'Nonstationary Lattice-Filter Modeling', December 1983 and J. M. Cioffi, 'Fast Transversal Filters for Communications Applications,' March 1984.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 28, 1984
- Accession Number
- ADA147130
Entities
People
- Thomas Kailath
Organizations
- Stanford University