UBC Theses and Dissertations
Rehashing the bit-interleaved coded modulation Anhari , Alireabout:za Kenarsari
Bit-interleaved coded modulation (BICM) is a pragmatic yet powerful approach for spectrally efficient coded transmission. BICM was originally designed as a superior alternative to the conventional trellis coded modulation in fading channels. However, its flexibility and ease of implementation also make BICM an attractive scheme for transmission over unfaded channels. In fact, a noticeable advantage of BICM is its simplicity and flexibility. Notably, most of today’s communication systems that achieve high spectral efficiency such as ADSL, Wireless LANs, and WiMax feature BICM. Perceptibly, the design of efficient BICM-based transmission strategies relies on the existence of a general analytical framework for evaluating its performance. Therefore, alongside its vast popularity and deployment, performance evaluation of BICM has attracted considerable attention. Developing such a performance evaluation framework is one of the main contributions of this thesis. In addition to conventional additive white Gaussian noise model, the practically important case of transmission over fading channels impaired by Gaussian mixture noise has also been studied. Different from previously proposed methods, our scheme results in closed-form expressions and is valid for arbitrary mapping rules and fading distributions. Furthermore, making use of the newly developed framework, we propose two novel transmission strategies. First, we consider the problem of optimal power allocation for a BICM system employing orthogonal frequency division multiplexing. In particular, we show that this problem translates into a linear program in the high signal-to-noise ratio regime. This reformulation extends the applicability and delivers considerable complexity reduction in comparison to existing algorithms. Finally, we propose novel detector architectures for a BICM system employing iterative decoding using hard-decision feedback at the receiver. We show that, taking the feedback error into account results in considerable performance improvement while retains decoding complexity.
Item Citations and Data
Attribution-NonCommercial-NoDerivatives 4.0 International