Jillian Carson-Jackson | Manager Maritime Communications and Vessel Tracking at AMSA |
This document reports on the VDES field trial that was performed in Brisbane from June to September 2015. The trial focussed on the performance of the new VDE channel in open and urban environments as well as the impact of interference from other VDES channels. The data collected during the trial has been made available for further analysis.
TRIAL OVERVIEW
A VDES trial was established in Brisbane with the express purpose of gathering data from a live system using some of the VDES modulation systems in the Preliminary Draft New Recommendation (PDNR) ITU-R M.[VDES] specification.
The Brisbane VDES trials have four high level phases:
- Prototype implementation
- Lab trial
- Field trial
- Reporting
The trials focus on the performance of the physical layer of the VDE terrestrial component as defined in PDNR ITU-R M.[VDES]. The VDE terrestrial channels support various modulations and various bandwidths ranging from 25k Hz to 100 kHz.
Using a COTS (Commercial Off The Shelf) Software Defined Radio (SDR) system, the VDES prototype was implemented during May 2015. The prototype feature set includes the latest VDES changes that were incorporated at the IALA e-Nav Comms working group meeting in May 2015. Following the completion of the VDES prototype implementation, lab trials were performed in South Africa to characterise the performance of the VDES prototype system.
During June 2015 the VDES prototype system was used to perform field trials in Brisbane, Australia. A transmitting base station was installed at the AMS offices in Pinkenba, Brisbane and VDES receivers were installed on two vessels. One vessel, the Tangalooma ferry, travelled down the Brisbane River and on to Moreton Bay and provided a suitable platform to test the range of VDES. The other vessel, an AMS vessel, was directed to travel up the river into the Brisbane urban area to evaluate the performance of VDES in areas where obstructions are present. Various recordings were made of the received RF signals during the trial and will be made available for further analysis.
Following the initial trial in June, an extended trial was executed over three months, collecting more data from the Tangalooma ferry.
Effective and robust voice communication and data transfer
| Organisation |
|---|
Australian Maritime Safety Authority |
Australian Maritime Systems Limited (AMS) |
IMIS Global limited (IMIS) |
Stone Three Venture Technology (Pty) limited (Stone Three) |
The field trails were successful in identifying the factors that affect that VDES system when using the modulation schemes described. The impact of the morphology in ports that are located inland and / or that may not have significant line of sight propagation paths is substantial and is illustrated below.
During post processing the BER and PER of the various trial runs were calculated. To provide an indication on the performance and reliability of the VDE modulation methods, packet error rates (PER) are evaluated at various distances from the base station. These results are averaged over a set of transmissions. To compare the performance of each modulation scheme, the range of modulation methods are grouped together per bandwidth. The estimated range predictions are based on the calculations done during the VDES Lab trial and are described in the VDES Lab report.
The data throughput of the system was also evaluated. This is displayed in both the BER based throughput, based on the amount of correct bits received, and PER based throughput, based on the amount of bits in packets received without errors.
The PER based results provide an indication of the system throughput without any error correction whilst the BER throughput results can be used as a basis to evaluate the system throughput with an error correction layer.
The results from the two vessels are evaluated separately as they have different height and route profiles.
In order to satisfy the agreed VDES trial brief and provide the best possible data, the system used for the VDES field trial was designed to test the physical layer of the VDES terrestrial component. The system supports a range of modulations and bandwidths. These were tested by transmitting a set of predefined messages from a base station located at the AMS offices and receiving these transmissions on the selected vessels. Raw recordings were made of the transmissions to enable post-processing of the data using various methods and algorithms. The following figure gives an overview of the complete VDES field trial system.
[[{"fid":"633","view_mode":"default","fields":{"format":"default","alignment":"","field_file_image_alt_text[und][0][value]":false,"field_file_image_title_text[und][0][value]":false},"type":"media","field_deltas":{"1":{"format":"default","alignment":"","field_file_image_alt_text[und][0][value]":false,"field_file_image_title_text[und][0][value]":false}},"link_text":null,"attributes":{"height":401,"width":538,"class":"media-element file-default","data-delta":"1"}}]]
METHODOLOGY USED FOR DATA COLLECTION
During the VDES lab trial, a loopback system was implemented that tested the complete system on a continuous basis. For the VDES field trial the loopback system was divided into four stages, as detailed below.
The VDES trial focussed on the performance of the modulation methods as defined in PDNR ITU-R.M [VDES]. One of the objectives was to compare these modulation methods with the GMSK modulation method used in AIS. In order to compare these results directly, GMSK was implemented using the same symbol rate as the other methods. It is important to note that in the 25 kHz bandwidth use case the default symbol rate is 19200 Symbols/s (19200bits/s), which is twice the data rate used in AIS.
The VDE terrestrial shore to ship band (VDE-1-B) was used for VDE channels during the VDES trial. This band consists of the following four channels:
- 2024 @ 161.800 MHz
- 2084 @ 161.825 MHz
- 2025 @ 161.850 MHz
- 2085 @ 161.870 MHz
As the trial system cycled through various modulations at various bandwidths, the centre of these channels was used as the centre frequency (161.8375MHz).
The VDES field trial in Brisbane provided a unique opportunity to evaluate the performance of VDES in a range of environments. From the analysis of the results the following can be concluded:
- Performance of VDES in open environments
- The overall range performance of GMSK and Pi/4 QPSK was about 4km shorter than estimated, relating to 3dB of additional loss at some stage in the system. This is most likely due to the specific demodulator implementation and its handling of various non-AWGN conditions.
- The various 8PSK combinations performed better than estimated, combined with the highest data rate, it also provides the best data throughput.
- From the throughput results there are clear stages where one modulation method provides better throughput performance than the other. At short ranges the highest order system gives the best performance, but at longer distances the alternative methods provide better throughput.
- This information may be incorporated in the process of selecting a suitable modulation and coding sequence for a specific transmission.
- Performance of VDES in urban environments
- In the presence of large obstructions all modulation methods suffer from extensive loss of transmissions.o The lower order methods are more reliable in these conditions, but they do not provide a significant advantage over the higher order methods.
- For example, in a short range environment with a lot of obstructions (for example in a complex port environment), using GMSK at 25 kHz will be a bit more reliable than using 8PSK at 100 kHz, but the GMSK signals will also suffer from complete packet loss under shadowing conditions.
- Under some circumstance, it should be considered to make use of higher speed methods with smaller packets sizes and to rely on retransmission at a shorter interval to improve the link reliability.
COMPARISON OF MODULATION METHODS
As the VDE channels of VDES support multiple modulation methods, the VDES field trial provides insight on the performance of each method under different conditions. The ideal method to be used in a specific scenario may also be determined by the use case.
For example, if a vessel had to operate in an area under 18km from the base station, 8PSK @100kHz would provide the best data throughput, however at longer distances the lower order modulation methods would provide the best data throughput.
For example in safety related applications, the narrowband, low order modulations would provide a more reliable link.
FUTURE WORK
The VDES trial data set has been made available to all IALA members to support further development and analysis of VDES. This data set includes both transmitted and received messages as well as the raw modulated signals.
Future work that can be addressed by using the VDES trial results and data set includes:
- Calculating the performance of the VDES using various FEC code rates based on the measurements done during the VDES trial.
- Identifying error patterns that may be present in the recoded data. For example the presence of burst errors.
- Evaluation of various equalizers to improve the system performance under non-AWGN conditions.