Iterative Nonlinear Equalizer with Serial Concatenation Convolutional Code SC3 over Selective Mobile Channel

The paper focuses exclusively on a new proposed model, which is joining equalizer with serial turbo code over a mobile frequency selective channel. The separation between equalizer and decoder is presented in the system for comparison, where there is no feedback connection between them. The proposed model depends on a Log-MAP algorithm. Therefore some modifications of Soft In Soft Out SISO output in the SC³ decoder with the use of puncturing and depuncturing enhancement by using new connection buffer. The proposed model uses hard decision bit for LLR code output bit and feedback to the non-linear equalizer. The gain earned at different frame length and different number of iterations is around 0.25 dB. This gain is earned form joining the priori code bits LLR with the decoder to be a feedback to the non-linear equalizer. These bits enhance the decision for non-linear equalizer leading to canceling the ISI iteratively.