city. During the test, a GPS receiver antenna of 
the master station was set up on the roof of 
Asia Air Survey Building. The vehicle was 
driven along major city roads during day-time 
traffic. The distance between master station 
and measurement vehicle is about 2.0 to 10 km. 
Before the test run starts, the INS initial 
alignment for about 8 minutes was carried out. 
The HISS was executed in the 
INS/DGPS/Odometer mode. Namely, if the GPS 
receiver could be utilized in differential mode, 
and also the values of PDOP (Positional 
Dilution of Precision) were less than 2.5, the 
INS/DGPS mode was executed. Otherwise, the 
INS/Odometer mode was used. 
4.1 Positioning Accuracy 
As the reference system, the DGPS with carrier 
phase observation (Trimble4000SSE) was used. 
During the test, 5 to 7 satellites were observed 
and the PDOP factor varied between 1.8 to 3.0. 
Figure 3 shows the comparison of 3D position 
by HISS solution with those by kinematic GPS 
solution. On the whole, the differences were 
less than 5 meters in latitude, longitude and 
height respectively. As shown in figure, height 
error changes largely between 5400 and 5500 
UTC time. That is the reason why the HISS 
data does not synchronize with the kinematic 
GPS data strictly. So, their differences seem to 
become larger when the vehicle rises suddenly 
on elevated road. But, in the another test under 
the same condition, the horizontal position 
error of 8 meters was observed. 
4800 5000 5200 5400 5600 
UTCTime(sec) 
Figure 3 HISS Positioning Accuracy 
A comparison between the DGPS solution and 
the HISS solution in latitude is shown in 
Figure 4. It indicates that the integrated 
solution is much smoother and removes 
peculiar GPS noise. 
1 51 101 151 201 251 30 
Time epochs (20msec) 
Figure 4 Difference between HISS and DGPS 
4.2 Attitude Accuracy 
Before evaluating the road alignment 
measurement by the HISS, a static test of the 
attitude was conducted. As the reference 
system, a high precision INS (TA7601) which 
accuracy is 0.02 degree in attitude was used. It 
was installed to approximately parallel with 
the HISS on the measurement vehicle. Roll, 
pitch, and heading angles were measured at 
the 17 different locations with different 
inclinations. The experimental results are 
shown in Figure 5. As is evident from Figure, 
the differences between HISS and TA7601 are 
about 0.02° for roll, 0.05° for pitch and 0.15° for 
heading. And the distinct time-dependent error 
was not confirmed. It is assumed that their 
differences are the HISS misalignments with 
respect to the TA7601. Their standard 
deviation are 0.012° for roll, 0.015° for pitch 
and 0.011°, the attitude accuracy of HISS is 
estimated be equivalent to the accuracy of 
TA7601. 
Time (minute) 
Figure 5 Static Accuracy of Attitude 
4.3 Road alignment measurement 
The HISS can output the attitude at high data 
rate with high accuracy over short periods of