Performance of convolutional interleavers with different spacing parameters in turbo codes

This paper considers application of a convolutional interleaver and the issue of influence of the space parameter in the turbo code performance. Similarly to previously considered convolutional interleavers, the new interleavers are designed as block interleavers and their performance in different code structures is compared with the interleaver having higher periods and space value of 1. In each comparison, the number of inserted stuff bits at the end of each data block is considered to be of similar order. Finally, suitable modification to the new interleavers is proposed improving performance for the codes with lower number of stuff bits.