**7.1 Illustrative example - Pilot certification of telematic applications using GNSS**

Pilot tests of GNSS signal reception conditions and their evaluation were carried out during September and November 2009 at e-Ident laboratory in Prague [21], the laboratory for electronic identification systems and communications. The signal was simulated with Spirent GNSS Simulator, type GSS 8000. For pilot testing the GPS positioning system was selected as a reference GNSS system.

The routes were generated by creating NMEA messages (a special file was created for each route), every case is located in Prague area and its nearby neighbourhood. NMEA log files of routes for pilot testing were recorded during September in real test rides in cars, with the GPS unit. This data was processed and adjusted according to requirements for pilot testing in laboratory.

For virtual gate passage tests the hardware equipment of universal telematic mobile unit as OBU (On-Board Unit) was used. Installed DEFT application (Dynavix, EFC, Fleet, Toll) was also developed on testing hardware device.

For testing were chosen 2 following testing sections. Both sections are situated in the west of Prague, where both directions of travel through the section are considered and prepared for testing (Fig.6):


Test sections have been recorded into the log file in GPS unit placed on the windscreen of testing vehicle. This data was then processed into various NMEA files (each represents one testing section and one direction of travel).

The NMEA records were cut and connected into series. Series of measurement consist of 10 experiments, 5 of them in one direction and 5 of them in the opposite direction, so that it can test the entire sequence, i. e. all 10 passages through the defined segment. Before each series testing, there was simulated passing of the initialization section (to get fix of GPS signal and calibrate the equipment).

Fig. 6. Initialization (blue) and testing (red) section on highway R5, western part of Prague

Virtual gates locations were defined by geographical coordinates and the direction was determined by the starting azimuth of drive. For testing there was set azimuth angle of 90° (from the road axis, 45° on each side).

Virtual gate passage was detected with software by the principle of measuring the distance from defined positions of the virtual gate in the outer circle of radius r1, which takes the value of the distance travelled by a vehicle at time t1 = 2 s, respectively, crossing the border of inside circle of radius r2, which takes the value of the distance that the vehicle travels at time t2 = 1 s (Fif.7).

Fig. 7. Radius of detection circles

20 Modern Metrology Concerns

Pilot tests of GNSS signal reception conditions and their evaluation were carried out during September and November 2009 at e-Ident laboratory in Prague [21], the laboratory for electronic identification systems and communications. The signal was simulated with Spirent GNSS Simulator, type GSS 8000. For pilot testing the GPS positioning system was

The routes were generated by creating NMEA messages (a special file was created for each route), every case is located in Prague area and its nearby neighbourhood. NMEA log files of routes for pilot testing were recorded during September in real test rides in cars, with the GPS unit. This data was processed and adjusted according to requirements for pilot testing

For virtual gate passage tests the hardware equipment of universal telematic mobile unit as OBU (On-Board Unit) was used. Installed DEFT application (Dynavix, EFC, Fleet, Toll) was

For testing were chosen 2 following testing sections. Both sections are situated in the west of Prague, where both directions of travel through the section are considered and prepared for

 section of road II/605 around the bridge over the highway R1; to initialize the GNSS unit, previous section of road II/605 (in the direction from the city centre) was used;

 section of highway R5 near the Metropole shopping centre; to initialize the GNSS unit, previous sections of highways R1 and R5 were used; section transit speed between 80

Test sections have been recorded into the log file in GPS unit placed on the windscreen of testing vehicle. This data was then processed into various NMEA files (each represents one

The NMEA records were cut and connected into series. Series of measurement consist of 10 experiments, 5 of them in one direction and 5 of them in the opposite direction, so that it can test the entire sequence, i. e. all 10 passages through the defined segment. Before each series testing, there was simulated passing of the initialization section (to get fix of GPS signal and

Fig. 6. Initialization (blue) and testing (red) section on highway R5, western part of Prague

**7.1 Illustrative example - Pilot certification of telematic applications using GNSS** 

selected as a reference GNSS system.

also developed on testing hardware device.

testing section and one direction of travel).

section transit speed between 50 and 70 kph;

in laboratory.

testing (Fig.6):

and 100 kph;

calibrate the equipment).

The geographical coordinates defining the location of the virtual gates were deducted from the publicly available maps on Internet, the azimuth was derived from the test drive data. Number of defined virtual gates was not restricted. For each section have always been defined 2 positions of virtual gate in each direction:


For the section on road II/605 the location of the second gate (parallel communication) was defined at a distance of approx. 20 m from the axis of the road. In the case of section on highway R5 the virtual gate for parallel communication is located at a distance of approx. 30 m from the axis of one belt of the highway.

The virtual gate passage detection testing (with usage of GNSS simulator) was carried out at a total of 26 series of 10 measurements (Fig.8):


The tests were carried out at first with the default GNSS signal without any restrictions and with the defined parameters of the route. After that further test scenarios were developed and the simulated signal was influenced in the following way:


Fig. 8. Simulation output of the position on road II/605

The simulated signal was influenced at first for whole measurement series, this approach was later changed to influencing the signal for various tests in the series, always another way, such as switching on and off the satellites at a defined angle above the vehicle, respectively above the ground (road plane). Signal influence has been prepared based on detailed description of possible signal influence. For the tests the spherical characteristics of the receiving antenna were chosen, with an open top of the antenna (i. e. position of the antenna on the roof of the vehicle).

Running of the tests has been recorded in the "log file" stored directly in the hardware unit. Based on this data further processing took place already on PC. The hardware unit recorded the following data of a virtual gate passage:


Each measurement result was then classified as "passed" or "failed" according to the following categorization:

	- gate is identified on the running road;
	- gate is not identified in other nearby road;
	- any upstream gate is not identified;
	- more passages through the same gate during one test are not evaluated;

Results of carried out testing are listed in Tab.4 [22]. Based on the measurements, it was demonstrated, that the results of the passage identification varies for different parameters of the GNSS signal, environment and other influences on the signal reception. The relatively low percentage of successful running gate identification may be due to high sensitivity software in the OBU. It can be assumed that for usage of OBU for telematic applications the higher success rate for negative detection of upstream or neighbouring gates will be demanded.

 simulation of built scenes of surroundings – highway, city, suburbs – which contain a typical set of ground clutter, multipath signal transmission and signal shading;

The simulated signal was influenced at first for whole measurement series, this approach was later changed to influencing the signal for various tests in the series, always another way, such as switching on and off the satellites at a defined angle above the vehicle, respectively above the ground (road plane). Signal influence has been prepared based on detailed description of possible signal influence. For the tests the spherical characteristics of the receiving antenna were chosen, with an open top of the antenna (i. e. position of the

Running of the tests has been recorded in the "log file" stored directly in the hardware unit. Based on this data further processing took place already on PC. The hardware unit recorded

Each measurement result was then classified as "passed" or "failed" according to the

more passages through the same gate during one test are not evaluated;

Results of carried out testing are listed in Tab.4 [22]. Based on the measurements, it was demonstrated, that the results of the passage identification varies for different parameters of the GNSS signal, environment and other influences on the signal reception. The relatively low percentage of successful running gate identification may be due to high sensitivity software in the OBU. It can be assumed that for usage of OBU for telematic applications the higher success rate for negative detection of upstream or neighbouring gates will be

ID and a description of the gate (defined positions and azimuth).

simulation of various predefined atmospheric changes.

Fig. 8. Simulation output of the position on road II/605

antenna on the roof of the vehicle).

virtual gate passage time;

following categorization:

**failed** in other cases.

demanded.

the following data of a virtual gate passage:

 **passed**, if all the following conditions are true: gate is identified on the running road; gate is not identified in other nearby road; any upstream gate is not identified;


Table 4. Results of pilot testing, depending on the segment, direction and type of test

Within implementation of a sufficiently large count of measurements the resulting values in Tab.4 show the probability of conformity of the tested OBU properties with the desired properties of the measured applications.

Proposed test protocol is divided into a part of recording the individual measurements and a part of evaluation of the application as a whole, including evaluation of system parameters. It is already possible on the basis of pilot testing to summarize the partial results and define partial requirements for selected system parameters of tested applications:

