42 
particularly difficult for the reception of the GPS signal 
(Fig. 11). 
Fig. 11. Measurement of a point with the 
GPS+GLONASS RTK technique. 
The tunnel is about 2 km long and its course is rather 
tortuous, both in planimetry and in altimetry. In fact, we 
measured differences in height between successive 
vertices only a few hundred meters apart of up to 18 
meters. A preliminary marking out performed to obtain 
information for the realisation of the working plan was 
followed by a rigorous measurement of the vertices of the 
polygon. A planimetric precision of 2 cm and an 
altimetric precision of 3 cm was required for this 
measurement \ The survey was performed with a mixed 
technique, GPS+GLONASS and traditional. The satellite 
technique was used in the static and fast static mode. The 
planimetric orientation of the GPS+GLONASS survey 
was carried out with some trigonometric vertices of the 
first order network and a vertex of the IGM95 GPS first 
network. For calculation of the orthometric heights of the 
tunnel’s course, we performed an altimetric link, by 
trigonometric levelling, of some datum points in the area. 
During construction of the tunnel, after the surveys had 
been performed, we proceeded to control the excavation 
operations. This was carried out by sampling some 
3 
These precisions are required whenever the excavation operations in a 
tunnel are performed with an in-line tunnelling machine. 
vertices of the polygon, selected on the basis of their 
position along the course (start of tunnel, end of tunnel 
and where the conduit intersected various constructions 
and existing works: a railway link, a cave, a hospital, 
etc.). The control was performed with the traditional 
topographical technique and the GPS+GLONASS RTK 
technique; the established precisions were close to those 
indicated in the measurement of the vertices. The RTK 
survey was carried out with the same instrumentation 
described in section 1. The radio link connection was 
realised with a Satel AS2X with selectable frequency and 
directional antenna; this apparatus permits larger ranges 
than the conventional ones and thus is particularly 
suitable in the presence of vegetation. The threshold 
filters for the fixes were selected at 4 cm for planimetry 
and 3 cm for altimetry. The acquisition interval was set at 
1" and the cut off at 5°. The measurement session lasted 
one hour. For the co-ordinate transformations, we used 
the parameters calculated during the static 
GPS+GLONASS survey. Despite the difficult operating 
conditions, there were always 5 satellites visible, with a 
PDOP never greater than 3. The cycle losses were 
frequent, but the apparatus was always re-initialised in the 
minimum possible time. The traditional survey was 
performed with a NIKON DTM850 integrated station. 
Table 2 summarises the mean values (and range) of the 
absolute differences between the co-ordinates of the 
sample points measured with the GPS+GLONASS RTK 
instrumentation and the co-ordinates of the same points 
measured with the traditional instrumentation. 
Differences in the 
Mean value 
Range 
co-ordinates 
(cm) 
(cm) 
5N 
3.5 
9.0 
ÔE 
2.5 
5.0 
8H 
2.0 
3.0 
Table 2. Differences in the co-ordinates measured with 
the RTK technique and with the integrated theodolite. 
The results of this test revealed the following aspects: 
• the acquisition capacity of the GPS+GLONASS RTK 
system even in the presence of vegetation and 
obstacles to signal reception; 
• minimum times of re-initialisation; 
• the possibility of operating with a fix filter that is very 
low in relation to the operating conditions; 
• precisions of positioning comparable to those 
obtained with traditional topographical 
instrumentation. 
CONCLUSIONS 
The availability on the market of GPS+GLONASS 
receivers on the L1-L2 frequency has led to appreciable 
advantages in operational geodesy. In this paper, we have