Design Article
Record and replay in global navigation satellite system testing
Julian Thomas, managing director, Racelogic (Buckingham, UK)
2/24/2012 8:55 AM EST
Extended simulation
Sometimes just simulating the GPS signal is not enough, as there may be other sensors involved in the successful operation of the system under test. Not only that, but actual motion may be required to stimulate embedded gyro sensors in order for the system to operate in a realistic fashion. For example, it is very common for a built-in Satellite Navigation system to use gyro and wheel speed information to supplement the GPS signal for poor satellite visibility areas such as urban canyons and tunnels. The gyro signal and wheel-speed data are often available on the CAN bus of the vehicle, which is an efficient way of connecting sensors around the vehicle, but makes simulation more difficult.
In order to accommodate for this kind of test scenario, LabSat has various input and output capabilities which are flexible enough to cover a number of different configurations.
On the input side, there is a high speed digital input which is sampled on every GPS sample, providing a tightly synchronized record of the digital data. The recorded digital data is then reproduced on the output port at the same time as the GPS RF data is replayed.
In order to convert real world signals into a digital format which can be recorded by Lab-Sat, there are a number of different conversion modules available such as RS232, RS422, RS485 and CAN. In some cases, the gyro sensor may be built into the main navigation unit, which makes it very difficult to record and replace the data from this sensor. In these cases, physically rotating the unit can be the only option. In order to overcome this difficulty, LabSat has a built-in GPS engine, which is used to monitor the output during the replaying of GPS data. The data from this GPS engine can be transmitted on a USB or serial link, and configured to control a precision turntable to reproduce the heading measurement corresponding to the direction in which the GPS antenna would be traveling.
In operation this is very effective, as the device under test is turned through exactly the same angle as it would have been in the real world. This means you can use pre-recorded scenarios, or artificially generated scenarios, the results are the same, and the device under test receives realistic signals from the internal gyro.
Nothing beats testing in a real world scenario, however, the ability to reproduce over and over again exactly what happened on the test drive, is invaluable in the development and testing of integrated GPS-based systems. An engineer can build up a library of useful test drives, and mix this with a number of artificial scenarios which are normally difficult to achieve in real world testing, such as crossing meridian lines and year rollover tests. From then onwards, most of the functional testing can be carried out on the bench with a high level of repeatability, in a very controlled environment.
Synchronizing with video
One of the disadvantages of testing with real world data on the bench is that you cannot see the exact conditions which were experienced by the receiver when the data was recorded. You may have satellite dropouts or multipath caused by tall buildings, bridges or even passing traffic, but when it is replayed on the bench, it would be difficult to know why this was happening.
To overcome this situation, Racelogic can supply a synchronized video system which is fitted to the car at the same time, which records video alongside the GPS RF data. When the data is replayed on the bench back at the office, the video plays alongside, fully synchronized at all times, allowing you to observe the exact conditions which were present during the original recording.
About the author:
Julian Thomas is managing director of Racelogic.
He can be reached at: julian.thomas@racelogic.co.uk
Visit Racelogic
Courtesy of EETimes Europe
Sometimes just simulating the GPS signal is not enough, as there may be other sensors involved in the successful operation of the system under test. Not only that, but actual motion may be required to stimulate embedded gyro sensors in order for the system to operate in a realistic fashion. For example, it is very common for a built-in Satellite Navigation system to use gyro and wheel speed information to supplement the GPS signal for poor satellite visibility areas such as urban canyons and tunnels. The gyro signal and wheel-speed data are often available on the CAN bus of the vehicle, which is an efficient way of connecting sensors around the vehicle, but makes simulation more difficult.
Figure 3: A setup on a precision turntable can reproduce GPS heading to test the unit’s gyro response
In order to accommodate for this kind of test scenario, LabSat has various input and output capabilities which are flexible enough to cover a number of different configurations.
On the input side, there is a high speed digital input which is sampled on every GPS sample, providing a tightly synchronized record of the digital data. The recorded digital data is then reproduced on the output port at the same time as the GPS RF data is replayed.
In order to convert real world signals into a digital format which can be recorded by Lab-Sat, there are a number of different conversion modules available such as RS232, RS422, RS485 and CAN. In some cases, the gyro sensor may be built into the main navigation unit, which makes it very difficult to record and replace the data from this sensor. In these cases, physically rotating the unit can be the only option. In order to overcome this difficulty, LabSat has a built-in GPS engine, which is used to monitor the output during the replaying of GPS data. The data from this GPS engine can be transmitted on a USB or serial link, and configured to control a precision turntable to reproduce the heading measurement corresponding to the direction in which the GPS antenna would be traveling.
Figure 4: A typical vehicle CAN connection
In operation this is very effective, as the device under test is turned through exactly the same angle as it would have been in the real world. This means you can use pre-recorded scenarios, or artificially generated scenarios, the results are the same, and the device under test receives realistic signals from the internal gyro.
Nothing beats testing in a real world scenario, however, the ability to reproduce over and over again exactly what happened on the test drive, is invaluable in the development and testing of integrated GPS-based systems. An engineer can build up a library of useful test drives, and mix this with a number of artificial scenarios which are normally difficult to achieve in real world testing, such as crossing meridian lines and year rollover tests. From then onwards, most of the functional testing can be carried out on the bench with a high level of repeatability, in a very controlled environment.
Synchronizing with video
One of the disadvantages of testing with real world data on the bench is that you cannot see the exact conditions which were experienced by the receiver when the data was recorded. You may have satellite dropouts or multipath caused by tall buildings, bridges or even passing traffic, but when it is replayed on the bench, it would be difficult to know why this was happening.
To overcome this situation, Racelogic can supply a synchronized video system which is fitted to the car at the same time, which records video alongside the GPS RF data. When the data is replayed on the bench back at the office, the video plays alongside, fully synchronized at all times, allowing you to observe the exact conditions which were present during the original recording.
About the author:
Julian Thomas is managing director of Racelogic.
He can be reached at: julian.thomas@racelogic.co.uk
Visit Racelogic
Courtesy of EETimes Europe
Follow us at Military/Aerospace Designline!
|
|
|
|
|
Navigate to related information