Flight Data Assessment of Tightly Coupled PPP/INS Figure 6. Example of the convergence benefit of INS integration during the half hour of the real-time PPP solution when compared to the postprocessed ambiguity-fixed PPP reference solution. ters takes nearly 30 min. The convergence benefit is also shown in Figure 7, which shows convergence of the carrier-phase bias estimates during the first 15 min of the forward-filter solution. In Figure 7, phase-bias errors for each of the satellites tracked at the start of the data set are estimated with respect to their final steady-state estimates. For each satellite, the tight-INS solution consistently enhances that filter's ability to quickly converge upon the carrier-phase bias. Figure 8 shows an example of troposphere zenith delay estimation error convergence during the first 15 min of the filter that used real-time products with and without tight-INS integration. The errors are calculated with respect to the Vienna mapping function (VMF) [49] grid delay estimation, which shows that the integrated solution converges to the correct troposphere delay more stably when compared to the solution without INS. Finally, when comparing positioning performance after a 30min convergence, as shown in Table 3, it becomes apparent that the positioning estimation performance is nearly identical between the filters with and those without INS. This substantiates that INS's primary benefit is to reduce the PPP solutions' initial convergence. Figure 7. Example of GPS carrier-phase biases converging during the first 15 min of the real-time filter with and without INS integration. Phase bias errors are estimated with respect to their final estimated steady-state values at 3.5 h into the flight. 18 IEEE A&E SYSTEMS MAGAZINE AUGUST 2017