New scenario for gps cycle slips detection and fixation using single frequency data

Abstract

GPS high accuracy applications are based on the phase measurements. Such phase measurements are frequently subjected to cycle slips. So, special attention should be paid to the detection and fixation of cycle slips. Previous works concerned with this issue dealt with single frequency data was faced with the low sensitivity of these models as most of them are based on differencing the change in both code and phase differences between consecutive epochs. Although the resulted test quantity is free of most GPS biases, it is contaminated by twice the ionospheric error due to its opposite sign in both code and phase equations. In this paper, a new test quantity is proposed. This test quantity, which was denoted as CPR, is defined as the ratio between the code and phase differences between consecutive epochs. CPR values exhibited very smooth manner between consecutive epochs. Different simulated cycle slips are introduced to different data sets. CPR graphs yielded detection sensitivity of one cycle for sampling rates up to 15 seconds and two cycles for higher sampling rates (up to 30 seconds). Concerning the fixation process, two scenarios are introduced in this paper which are backward and forward scenarios. In backward scenario, the 1 st corrupted CPR value is estimated. This estimated CPR is used backwardly to get the number of the slipped cycles by estimating the phase difference between the epoch of cycle slip and its preceding epoch. The same scenario was followed forwardly by estimating the 2 nd corrupted CPR and going by it in a forward model. Both scenarios yielded slightly different float solutiosn and the same fixed solutions. Finally, estimating the corrupted CPR values was tried using weighted mean of uncorrupted CPR values. Weighting of CPR is taken inversely proportional to the time gap between the estimated and the used values. Results proved that weighting of the most recent seven CPR values yielded the best float solution in both backward and forward scenarios.