1: \begin{abstract}

2: In this work, we investigate an efficient numerical approach for solving

3: higher order statistical methods for blind and semi-blind signal

4: recovery from non-ideal channels. We develop numerical algorithms

5: based on convex optimization relaxation for minimization of higher

6: order statistical cost functions. The new formulation through convex

7: relaxation overcomes the local convergence problem of existing gradient

8: descent based algorithms and applies to several well-known cost functions

9: for effective blind signal recovery including blind equalization and

10: blind source separation in both single-input-single-output (SISO)

11: and multi-input-multi-output (MIMO) systems. We also propose a fourth order

12: pilot based cost function that benefits from this approach. The simulation results

13: demonstrate that our approach is suitable for short-length packet

14: data transmission using only a few pilot symbols. 

15: \end{abstract}