Trellis-Based Iterative Adaptive Blind Sequence Estimation for Uncoded/Coded Systems with Differential Precoding
- PDF / 862,608 Bytes
- 16 Pages / 600 x 792 pts Page_size
- 98 Downloads / 170 Views
Trellis-Based Iterative Adaptive Blind Sequence Estimation for Uncoded/Coded Systems with Differential Precoding Xiao-Ming Chen Information and Coding Theory Lab, Faculty of Engineering, University of Kiel, 24143 Kiel, Germany Email: [email protected]
Peter A. Hoeher Information and Coding Theory Lab, Faculty of Engineering, University of Kiel, 24143 Kiel, Germany Email: [email protected] Received 1 October 2003; Revised 23 April 2004 We propose iterative, adaptive trellis-based blind sequence estimators, which can be interpreted as reduced-complexity receivers derived from the joint ML data/channel estimation problem. The number of states in the trellis is considered as a design parameter, providing a trade-off between performance and complexity. For symmetrical signal constellations, differential encoding or generalizations thereof are necessary to combat the phase ambiguity. At the receiver, the structure of the super-trellis (representing differential encoding and intersymbol interference) is explicitly exploited rather than doing differential decoding just for resolving the problem of phase ambiguity. In uncoded systems, it is shown that the data sequence can only be determined up to an unknown shift index. This shift ambiguity can be resolved by taking an outer channel encoder into account. The average magnitude of the soft outputs from the corresponding channel decoder is exploited to identify the shift index. For frequency-hopping systems over fading channels, a double serially concatenated scheme is proposed, where the inner code is applied to combat the shift ambiguity and the outer code provides time diversity in conjunction with an interburst interleaver. Keywords and phrases: joint data/channel estimation, blind sequence estimation, iterative processing, turbo equalization.
1.
INTRODUCTION
In most digital communication systems, a training sequence is inserted in each data burst for the purpose of channel estimation or for the adjustment of the taps of linear or decisionfeedback equalizers. For an efficient usage of bandwidth, however, blind equalization techniques attract considerable attentions [1, 2]. Furthermore, blind detection schemes may be embedded in existing systems as an add-on in order to improve the system performance in difficult environments. Blind linear and nonlinear equalization techniques have been investigated since the pioneering work of Sato [3]. Conventionally, blind linear equalizers exploit the higher-order statistical relationship between the data signal and the equalizer output signal. On-line adaptive algorithms based on the zero-forcing principle have been proposed in [3, 4, 5], for example. For burst-wise transmission, an iterative batch implementation of these algorithms is also possible [6], that is, the equalizer coefficients obtained at the end of one iteration are employed as the initial values in the next iteration. Based on the minimum mean-square error (MMSE) crite-
rion, algorithms for blind identification and blind equalization have been proposed in [7, 8] for multipath
Data Loading...
