Time varying individual channel estimation for one-way relay networks with UKF and URTSS

Shun Zhang,Feifei Gao,Hongyan Li

DOI: https://doi.org/10.1109/GLOCOM.2014.7037361

2014-01-01

Abstract:In this paper, we examine the pilot-based channel estimation in an amplify-and-forward (AF) one-way relay network (OWRN) under time selective flat fading scenario. Different from most existing works, our target is to estimate and track the individual channel of each relay hop instead of the composite channel. To reduce the number of the channel parameters to be estimated, we apply the polynomial basis-expansion-model (P-BEM) and convert the problem to estimating the channel coefficient-vector (called in-BEM-CV) of each relay hop. With the aid of the autoregressive (AR) model, we formulate the dynamic state space for the pilot-based in-BEM-CV estimation. Specifically, we adopt the unscented Kaiman filter (UKF) to track the in-BEM-CV dynamic variations in an online manner, and utilize the unscented Rauch-Tung-Striebel smoother (URTSS) to smooth the UKF's estimations in an offline manner. Finally, numerical results are provided to corroborate the proposed studies.