Skip to main content

Multiuser Channel Estimation from Higher-Order Statistical Matrix Pencil

Abstract

This paper presents a new statistical approach to the blind estimation of linear multiple-input multiple-output (MIMO) channels with finite impulse response. A matrix pencil is constructed from a set of fourth-order cumulant matrices of the channel output signals. The MIMO channel impulse responses can then be efficiently estimated from the generalized eigendecomposition of this cumulant matrix pencil. Random weighting is applied in the matrix pencil construction to improve the reliability of the algorithm. The proposed new method requires a relaxed channel identifiability condition and is robust in the sense that it does not require the exact knowledge of the MIMO channel order.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Jing Liang.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Liang, J., Ding, Z. Multiuser Channel Estimation from Higher-Order Statistical Matrix Pencil. EURASIP J. Adv. Signal Process. 2002, 175389 (2003). https://doi.org/10.1155/S1110865702209117

Download citation

Keywords

  • blind system identification
  • higher order statistics
  • multiple-input multiple-output linear systems
  • matrix pencil
  • identifiability condition