Channel estimation and training sequence design for one-way and two-way relay networks

Wireless relay networking is a highly active research field. Several relay standards have been or are being specified for next-generation mobile broadband communication systems. Channel estimates are required by wireless nodes to perform essential tasks such as precoding, beamforming and data detection. Thus this thesis focuses on channel estimation for amplify-and-forward (AF) one-way relay networks (OWRNs) and two-way relay networks (TWRNs).;For orthogonal frequency-division multiplexing (OFDM) based TWRNs, joint carrier frequency offset (CFO) and channel estimation is investigated. Two new zero-padding (ZP) and cyclic-prefix (CP) transmission protocols are proposed. Both protocols enable an estimator based on the nulling-based least square (NLS) algorithm and perform identically when the block length is large. A detailed performance analysis is given by proving the unbiasedness of the estimator at high signal-to-noise ratio (SNR) and by deriving the closed-form expression of the mean-square error (MSE). Since the two protocols and corresponding NLS algorithm can only estimate the convoluted channel parameters, a superimposed training strategy is proposed to estimate all the individual channel parameters. Specifically, three different algorithms that require different lengths of trainings are designed for the initial parameter estimation and an iterative algorithm is developed to refine the initial estimation results.;For TWRNs operating over time-varying fading environments, channel estimation and training sequence design are investigated. A new complex exponential basis expansion model (CE-BEM) is proposed to represent the mobile-to-mobile time-varying channel. To estimate the parameters of this model, a novel pilot symbol-aided transmission scheme is developed such that a linear approach can estimate the convoluted channels. More essentially, two algorithms are designed to extract the BEM coefficients of the individual channels. The optimal training parameters are derived by minimizing the estimation MSE.;For OWRNs operating over doubly-selective channels, estimation algorithms and training sequence design are investigated. The CE-BEM is utilized to approximate the doubly-selective channel. Since direct estimation of the CE-BEM coefficients requires large pilot overhead, an efficient estimator is developed that targets only useful channel parameters that could guarantee effective data detection. The training sequence design that can minimize the estimation MSE is also.