Abstract:  Tail-biting is a technique to convert convolutional codes into block codes. It eliminates the bit rate loss caused by the zero return state and avoids the performance degradation caused by tail-cutting. It has obvious advantages in short code transmission. Aiming at the complexity of the existing decoding algorithms of tail-biting convolutional code (TBCC) over large and convergent, a low complexity TBCC adaptive cyclic Viterbi (VA) decoding algorithm is proposed. The algorithm adjusts the number of iterations adaptively according to the change of the channel so that the tail-biting path converges to the best. By comparing the block error rate and decoding iteration times of different decoding algorithms, the simulation results show that the performance of TBCC is obviously better than traditional convolutional codes. Compared with the similar cyclic VA algorithm, the improved algorithm simplifies the stop rule and reduces the number and complexity of decoding iteration without reducing the performance. At low SNR, the average number of iterations of the improved algorithm is reduced by about 4 times compared with the traditional wrap-around Viterbi decoding algorithm (WAVA).