A platform to test Turbo Codes with different kind of Interleavers. This is the software for my Thesis in Turbo Codes. It is now open sourced for anyone to use.
Abstract: In the recent years the field of communications proposed a huge amount of advanced services that gave the users the ability to use very high bandwidth applications and high speed connection rates. These advancements and others were based on the early discoveries of new codes and algorithms with the support of development chip technologies. One of the main ones is the turbo code first produced in 1992, since then an era of new coding techniques proved to hop the performance to a degree very close to shannons limit, a limit that was the research of lots of engineers. The main component of the turbo codes structure is the interleaver, an interleaver is the part of the encoder in which the block of data are rearranged in a specific manner to be given to other parts to be processed. The interleaver is very important in turbo codes as it directly affect the performance graphs. In this thesis a new type of interleaver is introduced, the new design is based on an algorithm that separates the samples as far as possible in an attempt to reach the optimum design. It was combined with a constraint that lead both parts of the encoders to the zero state, a technique known as Similed Interleaving Constraint. It was found that a similed interleaver with the new design performs better than the similed interleavers with other types of interleavers.
In this thesis the complete framework environment for the analysis of turbo codes was programmed using the object oriented approach in C++. The implemented "GenData" library provides comprehensive tools for ready to use classes that can be implemented for different sets of block diagram algorithms.
Ammar Falih Al-Sudani