Channel Coding & Modulation – Turbo DeCodulation
Turbo DeCodulation
Since the discovery of the Turbo principle by Berrou et al., the field of channel coding has experienced a substantial progress and it still results in a steady approach of the Shannon limit. Especially in the emerging wireless communication systems the data throughput has to be continuously increased in order to comply with the customers' demands and to ensure a constant quality of service, even in adverse conditions, such as fading channels.
Various transmission systems based on iterative decoding have been developed, aiming at the enhancement of different issues in the encoding and decoding chain. All are founded on the same basic concept, the exchange of extrinsic information between at least two receiver components.
One example of such a transmission system is the bit-interleaved coded modulation with iterative decoding (BICM-ID), consisting of an iterative interconnection of the demodulator and the channel decoder. This system is the iterative extension of the widely used noniterative transmission scheme bit-interleaved coded modulation (BICM) that was originally designed for higher order modulation transmission over fading channels.
A further approach, namely, iterative source-channel decoding (ISCD) denotes the iterative concatenation of a channel decoder and a soft decision source decoder (SDSD). It aims at the exploitation of the residual redundancy of source codec parameters that remains after source encoding.
The so-called Turbo DeCodulation (TDeC) scheme, which is depicted in the figure to the left, is obtained by combining the basic ideas of BICM-ID and ISCD to a single system, where all three mentioned receiver components, i.e., the demodulator, the channel decoder and the SDSD, act jointly in order to increase the system performance.
EXIT chart analysis
An important tool needed for the analysis of iterative receivers is the so-called EXIT chart. It describes the mutual information at the extrinsic output of a receiver component of an iterative system if a certain a priori information is available. The advantage of EXIT charts is that only one characteristic has to measured for each component of the receiver. The overall performance of the receiver can then be evaluated using a joint representation of the characteristics.
If a system with two iterative loops is used at the receiver, i.e., three decoder components are evaluated in the iterative part of the receiver, the two-dimensional EXIT charts have to be extended to three dimensions. An example of a three-dimensional EXIT chart is given in the picture on the right side. Each surface describes a receiver component (red: demodulator, blue: channel decoder, green: source decoder). The "tunnel" between the surfaces as well as the intersection point of all surfaces are an indicator for the overall performance of the system. During system design, one goal is to move the intersection point to the upper right corner of the graph.