Co-prime sampling-based third-order cumulants for UWB channel order determination

This paper proposes an algorithm deriving higher-order statistics (HOS) from co-prime sampled sequences and implements it in the order determination of the parametric approach for modeling ultra-wide bandwidth indoor channel. We extend the co-prime sampling to pairwise co-prime sequences, and apply singular value decomposition in the matrix formed by third-order cumulants for the order determination of autoregressive model. We also analyze the variance of third-order cumulants and use their basic characteristic to determine the order of moving average model. In simulations, the proposed algorithm reduces the computational complexity to 17% of the existing HOS algorithms with negligible performance loss in high noise-to-signal ratio (SNR) environment. In low SNR scenario, the proposed algorithm is more sensitive to order mismatch in terms of the variances of expectations, which makes it a more reliable indicator to confirm the correct order.