1. Introduction
Automatic Identification System (AIS) is an advanced system imposed by the IMO (International Maritime Organization) dedicated to providing automatic traffic control and collision avoidance services in the maritime transportation sector. This technology enables ships to monitor the surrounding traffic and to share their status for safety purposes [
1,
2,
3]. Originally, AIS technology was dedicated to being used in terrestrial-based vessel monitoring applications (also known as terrestrial-based AIS), having a visibility range up to 40 nautical miles, and is operated based on the SOTDMA (Self Organized Time Division Multiple Access) communication protocols. The goal of this application can be attained since the terrestrial-based AIS is capable of allocating and sharing the available airwaves on the AIS frequency (161.975–162.025 MHz) to perform the ship-to-ship communication [
4,
5]. Due to this capability, within the last few years, a huge interest has arisen in terms of monitoring the vessels from space, i.e., space-based AIS in order to gain surveillance of the vessel traffic globally [
6].
In recent years, the space-based AIS technology has become an essential component for maritime wide-area surveillance since the swath width provided by the space-based AIS can be over 5000 km, which is much larger than terrestrial-based AIS. For example, due to the existence of curvature of the earth, the line-of-sight for ship-to-ship communications is limited to only 40–60 km, depending on antenna height [
7]. Due to this advantage, some government space agencies and private companies such as USA, Norway, China, India, Indonesia, and also European Union are installing AIS receivers into their satellites in order to explore the capability of this technology.
In 2007, the U.S. tested space-based AIS tracking using TacSat-2 microsatellite. However, due to the simultaneous receipt of many signals from the satellite footprint, the received signal was corrupted [
8]. Two years after this unsuccessful mission, in July 2009, the U.S launched AprizeSat-3 and AprizeSat-4 with AIS receivers which are capable of receiving U.S. Coast Guard’s test beacons off of Hawaii in 2010 [
9]. The demonstration of space-based AIS technology was continued by the Norwegians, which successfully launched a nanosatellite (AISSat-1) into polar orbit on July 2010. AISSat-1 was developed in order to improve the surveillance of maritime activities in the High North. On April 2011, Indian Space Research Organization also launched Resourcesat-2 bringing a Satellite-AIS (S-AIS) payload to provide the monitoring of maritime traffic in the Indian Ocean Search & Rescue (SAR) zone [
10,
11].
Specifically, in Indonesia, the space-based AIS has been installed on LAPAN-A2 microsatellite and was launched successfully in September 2015 [
12]. This satellite has been developed and operated by National Institute of Aeronautics and Space of Indonesia (LAPAN) until the present day. This satellite was equipped with a space-based AIS module capable of collecting vessel messages from about 650 km altitude with a 6-degree inclination. The LAPAN-A2 microsatellite is a unique satellite since it was placed in equatorial orbit; hence, it is capable of monitoring Indonesian maritime territory and other equatorial countries 14 times a day. This condition is in line with the purpose of installing the AIS system that is to monitor vessel activity crossing Indonesian waters. Therefore, for Indonesia, which consists of 5.8 million square kilometers of marine territory, the existence of this satellite is beneficial, especially in terms of maintaining maritime surveillance. Besides space-based AIS payload, Automatic Package Reporting System (APRS), Voice Repeater (VR), and multispectral camera were also installed to this satellite in order to support amateur communication and remote sensing purpose [
13]. The illustration of the LAPAN-A2 microsatellite and its mounted payloads is provided in
Figure 1 below.
Since LAPAN-A2 is the first satellite designed by LAPAN that has brought AIS payload, advanced in-orbit research is needed in order to study the performance of the AIS payload. Various researches and experiments related to the performance of the AIS payload are carried out in order to design a better generation of space-based AIS systems in the future. The performance of a space-based AIS system represents its ability to collect vessel messages which can be observed from quantitative and qualitative aspects. Quantitative AIS performance can be measured by comparing the number of messages received by the space-based AIS system to another AIS system such as terrestrial-based AIS [
14]. The other one, qualitative AIS performance, can be assessed by comparing the portion of valid messages to all received messages containing erroneous data [
15].
In order to study the in-orbit performance of the LAPAN-A2 AIS system, in this paper, a quantitative LAPAN-A2 AIS performance in terms of message reception has been performed based on its AIS antenna orientation. Since LAPAN-A2 is also equipped with another payload, i.e., multispectral camera, in some cases, in order to point a camera lens to a specific target, the satellite has to be maneuvered into the desired angle. This maneuver will set the AIS antenna position to be oriented, causing a shift angle towards its default position. Since the difference in antenna orientation produces variations of covered ground area, we have hypothesized that satellite maneuver can affect the number of receive message and the message reception performance. Therefore, in this paper, the measurement of LAPAN-2 AIS receiver performance was carried out by varying the antenna orientation. Since this research is intended to study the impact of different AIS antenna orientations on its message reception performance, there is no ground truth or other AIS data involved in this research. This data may be required if one desires to compare LAPAN-A2 AIS performance to other AIS systems.
In case no specific maneuver is required, the LAPAN-A2 satellite is set to nadir pointing mode. In the nadir pointing mode, the spacecraft’s camera or
z-axis is oriented to the center of the Earth, and this mode is defined as the nominal pointing of LAPAN-A2 satellite. During this nominal pointing, LAPAN-A2 AIS antenna is oriented to the south direction or 90° from its nadir direction. Since the antenna is mounted fixedly on the satellite’s body, to provide different antenna orientation, we have maneuvered the attitude of the satellite so that the AIS antenna will orient to 135° and 180° from nadir direction or 45° and 90° from its flight direction. All AIS messages collected during those maneuvers are analyzed and compared to messages collected in the nominal pointing (nadir pointing). The AIS datasets collected during the maneuvers are then analyzed to study the impact of antenna orientation on its reception performance quantitatively. The details of AIS antenna orientation and flight configuration are shown in
Figure 2 below.
Finally, the result of this research is organized to produce an essential recommendation related to orienting and positioning the AIS antenna. Even though this research has been performed on LAPAN-A2 satellite, the result is suitable to be used as a reference by other typical small satellites which already exist in the orbit. Moreover, the result can also be used in determining the right AIS antenna position and orientation for any satellite being developed which plans to carry similar AIS antenna to LAPAN-A2 or other typical omnidirectional antenna.
3. Results and Discussion
All datasets used in the comparison are produced by following the procedure provided in
Figure 6 (
Section 2.3.1). The complete properties of the AIS dataset collected under nadir and maneuvered conditions are shown in
Table 1.
As shown in
Table 1, for each maneuver, a total 33 h of continuous acquisition has been performed. The 33 h of duration was allocated to provide 11 h of acquisition before the maneuvering, during the maneuvering, and after the maneuvering, respectively. It is also shown that each maneuver was separated by different time distances, e.g., 2 days between Run-1 and Run-2 and 1 day between Run-3 and Run-4. This time separation has been chosen by considering other missions that have to be accomplished by LAPAN-A2 satellite. However, the existence of this separation is not an issue since the observation is not intended to be performed on a similar ship at all but rather to the ship on a similar location. Therefore, as long as the area covered by the satellite between the maneuvers is similar, the start and end time of the observation can be chosen freely. In addition, the problem might be faced if one uses overlapped datasets in which the message recorded in a dataset is re-used by another dataset.
3.1. AIS Messages Comparison
This section describes the comparison of total message number and unique MMSI received under nadir and maneuvered conditions. For detailed comparisons, each dataset has been categorized based on the AIS type messages according to the ITU definition [
18]. The collection of AIS messages is categorized in
Table 2.
The resulting activity of before, during, and after the 90°-maneuver and 45°-maneuver, respectively, is shown in
Figure 9 and
Figure 10. Those figures consist of the comparison of total AIS messages and each categorized message according to
Table 2 excluding Class A static and other messages. The exclusion of those types of messages is due to its low reception performance resulting in none of the received messages.
In
Figure 9, it is shown that the total number of messages received during the maneuvering, i.e.,
D90, (orange bar) are not significantly improved compared to before the maneuvering, i.e.,
D0 (dark-blue bar). This trend is almost consistent with other categorized AIS data. This phenomenon could be understood since the antenna coverage patterns are quite similar while performing nadir and 90°-maneuvered, as provided in
Figure 4a,c.
However, there is an interesting point that after the maneuvering (light-blue bar), the number of total received messages increases by about 29,07% on the average compared to during the maneuvering. We have investigated into the reason and found out that this condition is caused by the procedure used to set the satellite back to nadir after performing a certain maneuver. Before the maneuver, the current satellite flight configuration is nadir, as shown in
Figure 4a. In case an operator wants to change the current configuration to 90°-maneuvered flight configuration, it just needs to perform a straight forward maneuver by rolling the satellite. Moreover, to achieve 45°-maneuvered from the current nadir flight configuration, the pitch and roll maneuver have to be performed. Changing satellite orientation from nadir to 45°-maneuvered or 90°-maneuvered flight configurations can be achieved during a single satellite pass period. However, based on satellite attitude information, bringing back the satellite attitude to precise nadir flight configuration from a maneuver condition needs several adjustments; hence, it can be completed for about two to three satellite pass periods due to the limited access of a ground station. In this transition time, the AIS antenna orientation varies ±10° towards its flight direction. As a consequence, after the maneuver, the satellite is not immediately set to its nadir flight configuration. Therefore, all of the AIS datasets acquired after the maneuvering consist of a higher number of AIS messages compared to the number of messages acquired before the maneuvering, since these datasets are acquired under the combination of nadir and maneuver flight configuration.
Based on
Figure 10, it is confirmed that the performance of an AIS receiver during the 45°-maneuvered increases significantly compared to before and after the maneuver, except for Long-Range AIS messages type. For Long Range AIS on Run-3, although the 45°-maneuvered is capable of receiving a higher number of messages than before the maneuvering, this reception is still lower compared to after the maneuvering. On Run-4, the message reception of Long-Range AIS before the 45°-maneuvered is similar to the message reception during the 45°-maneuvered. We believe that the exception for Long Range AIS is caused by the small order of the number of received messages. Therefore, there is not enough data to make a statistically relevant conclusion. Excluding the Long Range AIS, the total number of messages received by making 45°-maneuvered increases around 31.11% to 239.29% compared to before and after the maneuver. One might have assumed that the increasing AIS message reception during the 45°-maneuvered and 90°-maneuvered is influenced by the different satellite ground track produced during the 11 h of maneuver. To deal with this assumption, the visualization of all ground tracks produced by the satellite during the observation is provided in
Figure 11 and
Figure 12.
The ground tracks provided in
Figure 11 and
Figure 12 were calculated during the 90°-maneuvered and 45-maneuvered observation period using the SGP4 algorithm [
21]. It is shown that the ground tracks before the maneuvering, during the maneuvering, and after the maneuvering are quite similar; hence, it is assumed that the AIS message reception should be quite similar. This phenomenon eliminates the assumption and confirms that the increasing AIS message reception is purely influenced by satellite maneuvering.
3.2. Class A and B Composition
Besides analyzing the total number of received messages and other AIS data categories, in this work, an analysis to observe the impact of satellite maneuvering to the message reception of certain AIS classes has also been conducted. Because LAPAN-A2 receiver is capable of receiving both class A and class B types, in this case, we divide AIS message composition based on those two classes. Since the AIS message transmitted from a class A transponder will always be prioritized, the AIS message transmitted from a class B transponder will not be received until a slot on the AIS channel is available. Due to this schema, only a small portion of class B messages are recorded by the LAPAN-A2 space-based AIS receiver. This phenomenon can be used to study the effect of satellite maneuvering on AIS receiver sensitivity level on a small portion of AIS messages, i.e., class B AIS messages.
Practically, class B AIS messages can be recognized by their message identity (message ID) assigned by “18” and “19” values while the others can be classified into class A AIS. The total number of class A and class B messages received during the observation are provided in
Figure 13.
Figure 13 shows that by using Run-1 and Run-2 datasets, the averaged message reception of class A AIS during the 90°-maneuvered is increasing by 17.17%, while for class B, the average message reception increases by 18.34% compared to before the maneuvering. It is also shown that Run3 & Run4 (45°-maneuvered) are capable of increasing the number of received class A & B messages. It is also confirmed that the averaged message reception is increasing by 208.80% and 175.93% for class A and B, respectively, compared to before the maneuvering. This condition indicates that by applying the 45°-maneuvered flight configuration, the LAPAN-A2 AIS receiver is capable of receiving more class A and class B AIS messages than the nadir flight configuration.
In this work, a filtering process was also performed in order to observe only unique MMSI detected during each maneuvering period. As shown in
Figure 14, during the 45°-maneuvered, the averaged detection performance of unique MMSI is consistently increasing by a factor of 33.34% compared to before the maneuver. Whereas the 90°-maneuvered is capable of increasing the averaged detection performance of unique MMSI by a factor of 13.43% compared to before the maneuver. By comparing 45°-maneuvered and 90°-maneuvered, it can also be seen that 45°-maneuvered are better in overall AIS receiver performance.
We also notice that in all trials under 45°-maneuvered and 90°-maneuvered flight configurations, the detection performance after the maneuver gives a significant increase of AIS message detection compared to before the maneuver. However, 45°-maneuvered shows an overall increasing of MMSI number detection.
3.3. Individual Message Reception Performace
As described in the previous section, according to IMO regulation, several types of ships should transmit their AIS signals containing the status of their ship every several minutes. By using this positive assumption, we have also performed the ship tracking in order to observe the individual message reception performance of particular ships as explained in
Section 2.3.2 by using the
D0,
D45, and
D90 datasets. The ship tracking is required to assess the impact of different antenna orientations specified by its flight configuration to the number of the message received by the space-based AIS system. The visualization of the ship tracking results are provided in several figures below.
From
Figure 15, it can be seen that a similar ship is detected in the South American Ocean having 100% re-detection probability. This ship has a different AIS message reception during the 45°-maneuvered and 90°-maneuvered observations. A ship with MMSI number 735058041, an Ecuador oil tanker ship, was detected 35 times before the maneuver, 22 times during the 90°-maneuver, and 43 times after the maneuver. At different times, this ship was detected 26 times before the maneuver, 35 times during the 45°-maneuvered and 27 times after the maneuver. From that data, the 45°-maneuvered is capable of providing higher message reception compared to the 90°-maneuvered and nadir flight configuration.
In
Figure 16, in an ocean near the Philippines that has 50–60% re-detection probability, we also made a tracking of a class A ship with MMSI 374131000. This bulk carrier ship with a Panama flag was detected one time before the maneuver, four times during the 45°-maneuvered, and three times after the maneuver. This ship was detected four times before the maneuver, one time during the 90°-maneuvered and four times after the maneuver. According to this ship reception data, in the area that only has 50–60% re-detection probability, the 45°-maneuvered is also better than the 90°-maneuvered in terms of message reception performance.
We have also performed some sampling of class B ships in the area which has 100% re-detection probability around the Pacific Ocean, as provided in
Figure 17. Class B ship usually uses small transmit RF power, around 2 watts, so detection of this type of ship from space is a challenge. In the area of the Pacific with 100% re-detection probability, class B could be detected quite well. One sample of Taiwanese fishing ship with MMSI 416004043 has been detected 26 times before maneuver, 50 times during 90°-maneuvered, and 51 times after maneuver. This fishing ship has also been detected 23 times before maneuver, 100 times during the 45°-maneuvered and 50 times after the maneuver. According to this message reception data, in this case, the 45°-maneuvered leads to better performance compared to the 90°-maneuvered. This result may be caused by the effect of Faraday rotation of the AIS signal received in space. Furthermore, radio wave reflection from the seawater surface; & radiation pattern of ship-borne VHF antenna may fully match its radiation pattern with the satellite’s antenna [
22]. The other way to increase the performance could be done by placing a directional antenna to narrow the coverage on a higher orbit satellite, especially in the equatorial region and also there is a study to develop 5-element dual-polarized array AIS antenna to provide full Earth coverage [
23,
24].
4. Conclusions
Within this paper, the analysis on the impact of varying the orientation of monopole λ/4 AIS antenna on its message reception performance has been presented. The analysis has been carried out by involving AIS datasets collected under 45°-maneuvered and 90°-maneuvered flight configuration applied on LAPAN-A2, an equatorial orbiting microsatellite. In this case, the 45°-maneuvered and 90°-maneuvered, respectively, refer to the satellite attitude configuration in which the antenna is oriented 45 and 90° towards the satellite flight direction. Based on the analysis over those datasets, it is confirmed that the highest message reception performance can be achieved by orienting the monopole λ/4 antenna 45° toward the flight direction. This antenna orientation gives a significant increase in overall AIS messages reception, including class A and class B type ships. From this finding, it is recommended for a small satellite or other satellite type carrying AIS payload to mount its monopole antenna orientation 45° towards its orbital flight direction.
Further study may be done by using other orientation angles or a combination of another type of antenna and different frequency. It also might be interesting to observe the performance of dual-polarization antennas for space-based AIS.