File Exchange

image thumbnail

Block Levinson solver

version (1.7 KB) by Tobin Fricke
Efficiently solve symmetric block Toeplitz matrix equations using Block Levinson recursion


Updated 01 Apr 2011

View License

BLOCK_LEVINSON(Y, L) solves the matrix equation T * x = y, where T is a symmetric matrix with block Toeplitz structure, and returns the solution vector x. The matrix T is never stored in full (because it is large and mostly redundant), so the input parameter L is actually the leftmost "block column" of T (the leftmost d columns where d is the block dimension).

By Keenan Pepper; uploaded with permission.

Cite As

Tobin Fricke (2020). Block Levinson solver (, MATLAB Central File Exchange. Retrieved .

Comments and Ratings (4)


how can I solve the matrix equation T * x = y when T is a block Toeplitz matrix and asymmetric?

liu hu

möp möp

Hi, I'm currently working on a matlab routine where now I came to the problem of solving a matrix equatin A*x=b where A is a matrix with block-Toeplitz structure, but is not symmetric. I wondered if the code block_levinson can also be used to solve my problem. Or does anyone know in which way it has to be modified?
Thanks for your advice in advance, it will be very helpful.



Hi. I have modified your code to work when y has dimension nxm (ie, to solve several systems at once) in accordance with the usual b\A solver. I can send you the code to upload if you would like.

- Robert Rosenbaum


Added attribution to Keenan Pepper in the description.

MATLAB Release Compatibility
Created with R2007b
Compatible with any release
Platform Compatibility
Windows macOS Linux
Tags Add Tags