Deliverable 8.1: Report on the evaluation of possible synchronisation/calibration solutions
In the full EISCAT_3D system it is critical that the antenna pattern and phase delay of each element is accurately known at all times.
Deliverable 8.1 presents the results of the evaluation of the synchronization system proposed for EISCAT_3D developed and evaluated by National Instruments, and a number of different calibration options applicable to daily operation of the EISCAT_3D system evaluated by Luleå University of Technology.
The good initial synchronization of the transmitters and receivers used in EISCAT_3D will allow the system to be operated without applying additional calibration. However, non-negligible performance improvements are expected to be achievable by calibrating for not only residual timing errors in the synchronization system, but also for cable delay variations, temperature dependent phase delay of the LNAs, as well as the antenna pattern of individual antennas and their positions. The investigated calibration methods are: celestial radio sources, radar reflections from objects in Earth orbit, local calibration signal injection, global cable-based calibration signal distribution and local calibration towers.
It appears entirely feasible to achieve excellent calibration accuracy using any of the methods presented in this report, but that a combination is required to calibrate both the whole bandwidth of the receiver arrays and the transmitter array. Radar reflections from objects in earth orbit is not only the only method identified so far that can be used to measure full transmitter antenna pattern (including far field phase), but also more efficient than using celestial sources for receiver calibration, but only applicable in the frequency band we are transmitting. Unless accurate antenna models can be used to predict the receiver radiation pattern over the whole bandwidth based on only transmission band measurements, we will also need to perform measurements using celestial radio sources as a complement for the rest of the bandwidth. We do not wish to rule out using local (without the signal distribution network) signal injection as a diagnostic tool for online (receiver antenna) return loss and coupling measurements, but monitoring changes in receiver antenna radiation pattern based on the previous two methods could be sufficient. Similarly, building a limited number of masts in some proximity to the transmitter for probing the field from the transmitting antennas could be useful for measuring variations in phase delay and perhaps a limited number of parameters of a antenna model on shorter time-scales than reflections from objects in orbit around Earth will allow.