Coordinated cancellation and correction of non-stationary noise in OFDM

点击翻译按钮获取中文摘要

We report about the results of the second project phase addressing all the four work packages that were once envisaged in the application. From the treatment of LDPC codes in impulse noise and a Markov modeling of the impulse noise memory, an open question evolved, namely how the noise state estimation can be improved by a joint treatment of LDPC decoding and Markov state estimation. So far, impulsive states are determined by a simple, yet optimized threshold decision, which already offers a close to optimum performance. The small leftover gap to optimum performance may even be closed by interlinking state estimation and iterative decoding. We also report about the use of lattice space-time codes, where we leave one spatial dimension for impulse-noise detection and derive the detection threshold and the pairwise error probability. Furthermore, a simple repetition-based amplify-and-forward relaying situation is studied and the optimum combiner is derived. Additionally, we professionalized the parameter estimation based on impulse-noise measurements showing the nice fit of Middleton’s Class A model. In the first project phase, we already started some studies applying our impulse-noise treatment to impulses of aeronautical distance measurement equipment. Now, we managed to determine an optimum receiver which obtains a performance close to the disturbance-free case.