GPS L1-L2 Bias Determination

Abstract : The GPS satellite L1-L2 bias T sub gd is determined using data from a TI-4100 receiver operating at Millstone Hill in Westford, MA. Pseudorange L1-L2 differences are computed. This difference is a measure of the total electron content (TEC) path delay along the line of sight plus the L1-L2 satellite and receiver bias. The difference of two measurements at the same time and satellite position cancels the TEC path delay and possible receiver bias and is a measure of the difference in satellite bias of the two satellites; i.e., Sv(a)L1-L2 - Sv(b)L1-L2. such direct measurements are not often available. Therefore, measurements at the same time are converted to 'zenith' TEC values using a mapping function. These difference are used in a least squares solution to determine the difference in bias between GPS satellites. Individual bias differences are obtained with an accuracy of 0.1 ns (o.285 TEC units) or better. By choosing one satellite as the reference, the individual satellite biases can also be obtained. The stability of each satellite bias is also estimated by examining a time series of the differences.