Real Time TEC Monitoring using Triple Frequency GNSS Data: A Three Step Approach
Spits, J.; Warnant, R.
Royal Meteorological Institute of Belgium

In the next few years, triple frequency Global Navigation Satellite Systems (Galileo and modernized GPS) will be operational. The availability of a third frequency will not only allow to improve the positioning accuracy but also to develop improved real time Total Electron Content (TEC) monitoring techniques.
The objective of our work is to develop such an improved TEC monitoring technique, of which the accuracy will not be affected by code hardware delays. This paper describes our first attempts to develop this technique.

As the third frequency is not yet available, this technique can only be tested on simulated data. For this purpose, we have developed a software allowing to simulate the measurements that will be made on the signals which will be emitted by the Galileo system and by the modernized GPS constellation. The objective is to simulate precise and realistic code and phase measurements, what implies that we have to minimize the differences between simulated and real data.
Therefore, in a first step, we developed a double frequency GPS simulation software. In this context, some utilities were developed in order to model the perturbations affecting GPS signals, what allows to introduce these effects in the simulated data. We were then able to compare real data with simulated data for different stations and days. We have shown that the differences between real data and simulated data can be explained by the uncertainties remaining in the modeling of the perturbations affecting GPS data.
In a second step, the software has been adapted in order to add the third frequency. Subsequently, we also developed the triple frequency Galileo software on the same basis.

Then, the GPS and Galileo simulated data can be used to test a triple frequency technique for TEC monitoring. We applied a method which is divided in three steps. GPS frequencies are L1, L2 and L5 which corresponds to L1, E5b, E5a for Galileo.
The goal of the first step is to resolve the so-called extra widelane (EWL) ambiguities by differencing the code and phase combinations of L2 and L5. The EWL combination wavelength equals 5.861m for GPS and 9.768m (almost double) for Galileo. We have estimated the influence of the time group delays, code multipath effects and code noise on the determination of these ambiguities, and we have concluded that it is possible to fix them at their integer values.
In the second step, we use the EWL fixed ambiguities to resolve the widelane (WL) ambiguities, which is a combination of L1 and L2 ambiguities. The WL combination wavelength equals 0.862m for GPS and 0.814m for Galileo. Unfortunately, this resolution is not possible due to the influence of residuals effects (mainly ionospheric effects). However, we obtain a first estimation of the WL ambiguities which is used in the next step.
In the third step, we solve the one-epoch system of 2 dual frequency Geometric Free (GF) phase combinations (2 equations – 4 unknowns) by using the information we have from EWL and WL ambiguities, so we obtain a system of 2 equations with two unknowns: the TEC and the ambiguities on L2. Due to the unfixed WL ambiguities, the system is not correctly solved. But as these ambiguities are integer numbers, we can fix them at their correct values if we have a rough estimation of the TEC. In this context, it is important to note that a change of 1 cycle in the WL ambiguities causes a change of approximately 12 TECU in the TEC values. The estimation of the TEC is obtained by using the usual dual frequency method. We have shown that this estimation is sufficient to resolve the WL ambiguities, so we can precisely monitor the TEC.

In conclusion, the main advantage of our method is that we have to resolve integer ambiguities (EWL and WL) in place of non-integer ambiguities coming from the GF combination, what is easier and more efficient. The second is that even if code measurements are used in the first step of the method, code biases do not affect the accuracy of the final reconstructed TEC. As a result, the accuracy of the final reconstructed TEC will be much better (at least one order of magnitude) than in the past with the double frequency TEC technique.

Up to now, the method has been tested on GPS and Galileo simulated data. However, as the Galileo GIOVE-A navigation signal will be soon available to users, we will be able to test and validate our TEC monitoring technique on real data.

Furthermore, we will be able to exploit the third frequency and the more precisely computed TEC in order to ameliorate the modeling of the ionospheric effects for the needs of real time positioning techniques based on phase measurements and requiring ambiguity resolution.