Bit-interleaved coded modulation with iterative decoding (BICM-ID), which possesses low complexity, strong agility, high spectrum efficiency and excellent BER performance, has been a key technique for next generation of wireless communication. The critical technology of BICM-ID is the feedback between the demodulator and the decoder, and also lies in the decoding between adjacent iterative. In the process of iterative decoding, the extrinsic information and the received systematic observations exert different influence on decision-making as the channel SNR changes. By multiplying a weighted coefficient for the priori information, this paper proposed a novel iterative decoding algorithm for BICM-ID. The improved scheme alters the proportion of the extrinsic information and the received systematic observations in the soft outputs of decoding. In this way, turbo decoding can trace the changes of channel and obtain improved performance. Simulation results show that the proposed scheme can lower the error floor of BICM-ID.