A lagrangian relaxation based decoding algorithm for LTE turbo codes

Maximum likelihood decoding (MLD) of LTE turbo codes can be formulated as an integer programming (IP) problem. Based on a novel IP formulation we propose a two step decoding algorithm. Under certain assumptions it can be guaranteed that our two step algorithm outputs the ML codeword. We also show that on LTE turbo codes with short block length, e.g., 132, 228, our algorithm outperforms linear programming decoding (LPD). For the (132,40) LTE turbo code it is also shown that the proposed algorithm has better error correcting performance than standard iterative turbo decoding.