The significance of guidance data for telemetry and control systems cannot be overstated. Fully utilizing limited input data to generate high-quality guidance data plays a critical role in the efficiency of the entire telemetry and control system and, to some extent, acts as protection for the receiving equipment. This paper addresses the issue of varying input data rates from different sources by designing a guidance generation algorithm for low data rate sources. It is based on the sliding window least squares method, performing two rounds of fitting. The first fitting keeps the data rate unchanged, while the second fitting increases the data rate to meet the guidance requirements. Simulation tests have proven that the two rounds of fitting can achieve more accurate and smoother guidance data under limited window length conditions.