Abstract
A l? suboptimal deconvolution filter is proposed for multi-input multi-output (MIMO) signal transmission systems with unknown but persistently bounded noises. The l? optimal deconvolution filter is to reconstruct the transmitter signal such that the peak of reconstruction error due to the noises is as small as possible. The l? optimal deconvolution filter design problem is transformed to suboptimal deconvolution filter problem by minimizing the upper bound of l? norm with some linear matrix inequality (LMI) constraints. In this situation, the very complicate l? optimal deconvolution filter design problem in MIMO systems can be easily solved by LMI method via convex optimization technique. In order to improve the transient response of l? deconvolution filter, the suboptimal l? filter with desired pole placement is also discussed. Further the mixed l?/H? deconvolution filter is considered for performance robustness.