Project Details
Projekt Print View

Construction and decoding of convolutional codes over the erasure channel

Subject Area Mathematics
Term from 2017 to 2020
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 392752124
 
Final Report Year 2020

Final Report Abstract

During the first year of my research fellowship, I managed to estimate the necessary field size and the probability for MDP and complete MDP convolutional codes. Moreover, I introduced j-complete MDP convolutional codes, developed a low delay decoding algorithm for convolutional codes and showed that j-complete MDP convolutional codes are optimal for this low delay decoding. Besides, we obtained the minimal possible field size and corresponding constructions for j-complete MDP convolutional codes with certain parameters and were able to classify all complete MDP convolutional codes over a field of minimal size for these parameters. Furthermore, we provided a construction for MDP convolutional codes using superregular matrices and developed decoding algorithms for 2-dimensional convolutional codes and for convolutional codes over integer residue rings. In the second year of my project, we developed a decoding algorithm for convolutional codes using the linear-systems representation and provided conditions on the corresponding matrices (A,B,C,D) in order to get a good performance with this algorithm. Moreover, we derived a matrix theoretic result on the left primeness of certain polynomial matrices and applied this result to simplify the most commonly used criterion to check whether a convolutional code is MDP. Finally, we were working on the construction of LDPC convolutional codes using certain combinatorial objects called difference triangle sets.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung