Abstract: To enhance the pseudo-range measure performance based on code synchronization of short burst DSSS signals, such as the RDSS inbound signal of Compass, a new kalman tracking algorithm based on iteration is proposed. After a code and carrier forward tracking based on initial values provided by acquisition or coarse tracking, a back estimation is followed to get more precise values of initial parameters of the received signal at the beginning point. By utilizing these precise initial values, forward tracking based on the same received signal is carried out again. By doing forward tracking and backward estimation iteratively, more and more precise tracking results are achieved. At the beginning of each forward tracking or backward estimation, an optimization of code discriminator is carried out considering the precise of initial carrier and code phase, which improves the final precision of code synchronization any more. For the convenience of optimization, a performance estimation algorithm of dot product power code discriminator for banded signal is proposed. Compared with several traditional tracking algorithms, simulation results and practical adoption during RDSS signal tracking of Compass show the excellence of our proposed algorithm.