Improved Blind Channel Estimation Performance by Nearby Channel Estimation

TL;DR

This paper proposes an improved blind channel estimation method that leverages nearby channel information as an initial estimate, enhancing accuracy especially when adjacent channels are highly correlated.

Contribution

The study introduces a novel approach combining adjacent channel estimates with subspace blind estimation to improve performance and reduce errors.

Findings

Higher correlation between channels improves estimation accuracy.

Adjusting the forgetting factor enhances performance based on channel correlation.

Using neighboring channel info as initial state reduces estimation errors.

Abstract

To obtain channel information for further data transmission, the blind channel estimation needs no pilot signal in advance is a considerable algorithm. To improve performance of the blind channel estimation schemes, we have employed the channel estimation with pilot signals by adjacent users; the acquired estimation channel information of the adjacent channel is applied as an initial state to the blind channel estimation of the target user. We have considered the subspace blind channel estimation scheme in this study. The estimated adjacent channel information is supported as an initial state of the auto-correlation matrix in the subspace estimation method, thereby providing information to reduce the channel estimation error. From our simulation study, we have found that, when the correlation between two nearby channels is high, the forgetting factor in the subspace estimation method can be increased, so that the weight of the initial state of the matrix is increased, and a more accurate channel state is estimated. If the correlation coefficient is lowered in two adjacent channels, the forgetting factor can be appropriately adjusted to rise the weight of the direct received signal from the objective channel to avoid errors increased.