Direct estimation of partial channel impulse response for code acquisition and tracking

The conventional acquisition and tracking algorithms, based on the correlation of the received and the locally generated pseudo-noise (PN) sequences, provides maximum likelihood (ML) estimations. In practice, the correlation is usually computed partially, and results in a suboptimum solution. We propose an algorithm based on a recursive weighted least squares (RWLS). Here, the correlation window can be made arbitrarily large. This increase in the effective width of the correlation window results in improved performance due to the increased processing gain. For time-varying channels, however, an arbitrarily large window may not be desired and this effect is controlled through the forgetting factor in the proposed RWLS algorithm