Analysis of Radio-Shaded Areas in the Geoje Island Sea Based on the Automatic Identification System (AIS)

Abstract

An automatic identification system (AIS) is often installed on merchant ships and fishing boats to prevent collisions and ensure safe navigation. The location information of ships transmitted from AIS equipment can help maritime traffic control prevent accidents. The southern coast of Korea comprises a complex coastline with numerous fishing boats and transit vessels. In particular, the Tongyeong and Geoje Islands include high-altitude mountains and islands, resulting in several radio-shaded areas where AIS signals cannot be received, owing to geographical effects. However, only a few studies have explored this region and performed practical experiments on the reception status of AIS locations in radio-shaded areas. In this study, we performed an experiment in the Geoje Island Sea on the southern coast to analyze the impact of high terrain on the reception rate and status of automatic identification devices. Two identical pieces of AIS equipment were installed to generate multiple radio waves, and the location data transmitted via different antennae were compared. The experimental analysis forms the basis for identifying the exact location of ships in the event of maritime accidents, facilitating rapid rescue. Moreover, the accuracy of the location transmitted by the AIS equipment can aid in detecting the cause of accidents.

1. Introduction

The primary function of the automatic identification system (AIS) is to prevent collisions by displaying the location and navigation information of nearby ships. The Maritime Traffic Control Center uses AIS data collected in the control area to monitor and control ships efficiently.

The General Information Center on Maritime Safety and Security (GICOMS) monitors ships using AIS, cellphones, and satellite information, and provides information on system connections to the National Security Council, National Intelligence Service, Ministry of Public Administration and Security, and Korea Coast Guard Agency [1].

The AIS equipment accurately extracts and displays the global navigation satellite system (GNSS)-based locations to control transit vessels in vessel traffic services (VTS) and to avoid collisions with other vessels. In particular, the accuracy of the GNSS location, determined by the AIS equipment installed on the ship, is essential for investigating the cause of accidents and facilitating rescue missions. Additionally, the reception rate and reception status of AIS location information transmitted from the ship and received by the VTS or GICOMS via sea and land repeaters are important factors.

Table 1. Automatic identification system (AIS) Class A nominal reporting interval (ITU-R M.1371-4).

Dynamic Condition of the Ship	Nominal Reporting Interval
Ship at anchor or moored and not moving faster than 3 knots	3 min
Ship at anchor or moored and moving faster than 3 knots	10 s
Ship moving at 0–14 knots	10 s
Ship moving at 0–14 knots and changing course	3.3 s
Ship moving at 14–23 knots	6 s
Ship moving at 14–23 knots and changing course	2 s
Ship moving at >23 knots	2 s
Ship moving at >23 knots and changing course	2 s

summarizes the general reporting interval of AIS Class A equipment prescribed by the International Telecommunication Union (ITU-R M1371-4), and

Table 2. AIS Class B nominal reporting interval (ITU-R M.1371-4).

Dynamic Condition of the Ship	Nominal Reporting Interval
Ship not moving faster than 2 knots	3 min
Ship moving at 2–14 knots	30 s
Ship moving at 14–23 knots	15 s
Ship moving at >23 knots	5 s

summarizes the standard AIS Class B reporting interval. The reporting interval varies from 2 s to 3 min depending on the navigation status and speed of the ship.

The comparison of V-pass data received from the VTS and AIS Class B equipment [2] indicates that, in the case of a transmitting device on a fishing boat on the coast of Tongyeong, the AIS data received from the VTS are typically transmitted every 30 s to 1 min in regular intervals. However, the interval between reception locations may be irregular owing to no location transmission for 2 min to 7 min and 30 s, despite the V-pass data being transmitted every 30 s. Although the experimental vessel transmits the location accurately via an antenna, no radio wave system can receive 100% of the location information that is transmitted from all ships owing to certain challenges, including the number of ships operating simultaneously in the same area, the type of equipment, wireless interference, the limitations of communication systems, and the distance from land relay stations. The experiment compared the status of the transmission and reception of V-pass equipment, which is the most commonly used location-transmission device in fishing boats along with AIS equipment; however, it was limited to the area around Bijindo Island on the southern coast of Gyeongsangnam-do and did not reflect the geographical environment.

A study explored the degree of location error of each piece of location-transmitting equipment by extracting the GNSS location data stored in the V-pass and AIS equipment installed on a silverfish fishing boat [3]; the GNSS location data were stored in the AIS equipment. Among the three pieces of equipment, which included the electronic chart display and information system (ECDIS), AIS, and V-pass, the differential global positioning system (DGPS) location of the ECDIS was recorded relatively accurately; however, an error occurred occasionally in the form of fluctuations for 10–30 s or the location appearing out of position for 8–30 s, with a maximum error of 28 m. The location error of the equipment with the GNSS antenna was larger than that of the equipment with the DGPS antenna. In particular, the maximum location error of the V-pass equipment was 168 m, which was the largest among the three pieces of equipment; the AIS equipment exhibited an error of 51 m. This study provided a basis for using the GNSS location error of V-pass and AIS equipment installed on small fishing boats without using the DGPS antenna, facilitating the responses to fishing boat accidents.

In comparison with the information collected by other navigation devices such as automatic radar plotting aids [4], the collected AIS information may be inaccurately displayed owing to human errors and technical defects, which can adversely affect the VTS control. Jeon and Jeong [5] analyzed the problems of AIS data error caused by the omission of information input by the user and the analog-to-digital converter error in the prerequisite defense information provided by the navigator; they suggested an improvement plan for the information input. Kim et al. [6] proposed AIS error data and a field-correction algorithm to improve the reliability and accuracy of ship information received from the AIS equipment. Nguyen et al. [7] developed a method for interpolating the loss of AIS location data missed by a ship. Additionally, Jung et al. [8] analyzed the priorities according to the utilization of AIS information and determined that dynamic information was used the most, confirming the need for further research to increase the reliability of AIS information [9].

Several studies on AIS equipment discuss the problems in using the equipment and the subsequent utilization of the collected information [8,10,11,12,13,14,15,16,17]. In these studies, the reception rate of the radio waves varied depending on the type of equipment, geographical environment, location of the repeaters, and ship traffic; in certain cases, the location was not received for a specific time period. Moreover, various errors can occur in the received location data, indicating that the location transmitted by the ship during an accident may not be accurate.

Therefore, the exact location of ships should be identified during accidents to facilitate rapid rescue. Furthermore, the accuracy of the GNSS location transmitted by the AIS Class B equipment installed on most domestic ships must be directly evaluated to investigate the cause of the accident. In this study, two identical pieces of AIS equipment were installed at the same location on an experimental ship. The location data transmitted via different antennae were compared, one of which was set to record the location of the vessel every 5 s and compared with the location data received by the maritime police VTS and geographic information system (GIS).

2. Materials and Methods

2.1. Test Details and Procedure

An experiment was conducted in the waters near Geoje, Gyeongsangnam-do, on 20 September 2023. Three data points were compared, including two data points transmitted through a land repeater with different antennae and maritime mobile service identity (MMSI) numbers for two automatic identification devices (AIS equipment, AIS-50N; Samyoung ENC, Busan, Repubilc of Korea).

In general, the AIS equipment receives satellite signals transmitted by the GNSS, determines the current location of the ship, and transmits the location to a base station on land using a set radio wave. Various technical approaches and transmission and reception methods may be used by the equipment depending on the type of manufacturer. Therefore, we used two identical models of AIS equipment from the same manufacturer to exclude technical differences and to ensure a relatively accurate experiment.

The experimental ship used was Chambada, a marine survey ship belonging to Gyeongsang National University. The ship was 23.14 m long, 4.7 m wide, and 2.18 m deep, equipped with a tonnage of 36 tons, a maximum speed of 20 knots, and two diesel engines of 600 horsepower. The weather was clear during the experiment, with an average temperature of 23 °C, 1–4 m/s southwest wind, and a wave height of 0.5–1.0 m.

2.2. Experiment Location and Method

In addition to the AIS equipment (referred to as CHAMBADA AIS) installed in the Chambada experimental ship, additional AIS equipment of the same model (referred to as STUDY AIS) was installed in the steering room. An antenna was installed in a location similar to the Chambada antenna to minimize positional errors. The storage interval of the AIS equipment was set to 5 s to compare the AIS location data transmitted from the ship and received by the base station on the land while rotating the surrounding sea around Geoje Island on the electronic diagram of the ship collision reproduction system [18,19]; this is a system used by the Korea Coast Guard to analyze the cause of accidents during the real-time reproduction of collisions by inputting the ship’s navigation details at the time of collision. This system was developed under the National Research and Development Project of the Korea Coast Guard from 2014 to 2017 and is currently used by the scientific investigation community of local offices to analyze the cause of accidents. In this study, AIS navigational data obtained via maritime experiments were compared and analyzed.

2.3. Experimental Equipment

The AIS-50N model of Samyoung ENC is a Class B type AIS device with its own DGPS function. It enables the real-time monitoring of the nationality and navigation information of other small vessels sailing in coastal waters. Regulations on the installation of ship-location-transmission devices, except fishing boats, are stipulated in the Ship Safety Act. Fishing boats of more than 10 tons are required to be equipped with AIS equipment. Because most fishing boats in coastal waters operate on islands or near coasts, location transmission is not likely to occur because of geographical problems. Therefore, an experiment was conducted to confirm the topographical effect of radio wave reception using two pieces of AIS Class B equipment of the same model.

3. Results

3.1. Comparison of AIS Data Received by VTS and GIS of the Korea Coast Guard

Figure 1 depicts a comparison of the data of the existing Chambada AIS equipment received from the Tongyeong Coast VTS in the Geoje Island maritime experiment, data of the STUDY AIS equipment installed in the laboratory, and data recorded in the two pieces of AIS equipment stored in the GIS of the Korea Coast Guard on an electronic map. GIS provides various types of information (e.g., maps, charts, and pictures) by integrating and managing the location and attributes of objects with geographical locations, such as ships.

The AIS data of the Tongyeong Coast VTS are indicated as green dots every time they are received, and the interval between receptions is irregular. Among the AIS data received from the two pieces of AIS equipment stored in the GIS of the Korea Coast Guard, the red spot represents the laboratory equipment, whereas the brown spot represents the existing AIS equipment installed in the Chambada.

The entire experiment lasted approximately 5 h, from 10:00 to 15:00, wherein the ship departed from the Tongyeong Port, sailed counterclockwise with Geoje Island in the center, and sailed close to the island and coast to detect the radio-shaded area of the AIS in the Tongyeong and Geoje Islands that include numerous islands and mountains. Although the entire Geoje Island was circled, no AIS location signal was received in certain sections depending on the area and the equipment used. Although the equipment used was of the same model, the brown and red marks overlapped in some sections; however, no overlapping of location data was observed.

The Chambada data (green) received from the Tongyeong Coast VTS center, the STUDY data (brown) stored in the GIS, and the CHAMBADA data (red) were easily distinguished between the area where the AIS location data were received and the area where they were not received.

Figure 2 depicts a part of the entire voyage that begins from the Tongyeong Port and sailing near Hansando, west of Geoje Island. The AIS (red) of the Chambada passed Hansando Island from 10:33:42, the STUDY data (brown) did not receive any AIS location signal from 10:36:27, and the first data of the VTS reception (green) were received from 10:31:56. The CHAMBADA data (red) were not received because the location signal was cut off at 10:33:42.

Figure 3 depicts some of the data points indicated in Figure 1, wherein the ship sails east of Geoje Island. Figure 2 indicates that the data were not received from the point passing through Hansando, whereas the STUDY AIS data (brown) were received near Galkotdo and the CHAMBADA AIS data (red) were received up to Jisimdo.

Figure 4 depicts some of the data points illustrated in Figure 1, wherein the ship sails north of Geoje Island. Both the STUDY AIS data (brown) and the CHAMBADA AIS data (red) exhibit a close reception interval in the northeast region, and the reception interval widened near northwest Gajodo.

Table 3. AIS location data reception rate according to the equipment and system.

Sort	Time	Total Time	No. of Data Points	Target Value	Reception Rate	Receiving Interval
Chambada	10:08:42–14:54:42	4 h 46 min 00 s	135	572	23.6%	127 s
Study	09:58:58–14:55:27	4 h 56 min 29 s	158	592	26.69%	112 s
VTS data	10:31:56–11:29:48	57 min 52 s	348	347	100.3%	9.97 s
AIS equipment	10:08:42–14:48:02	4 h 39 min 20 s	3352	3352	100%	5 s
Average					62.65%	63.49 s

lists the data reception rates for each piece of equipment. The CHAMBADA and STUDY data received location data for approximately 4 h and 46 to 50 min, which was the duration of the experiment. The reception rates were approximately 23.6% and 26.99%, calculated based on 30 s for 135 CHAMBADA data points and 158 STUDY data points, respectively. The 348 VTS data points indicate a 100% reception rate at 10 s, when the location data were received for approximately 57 min. A total of 3352 data points were stored every 5 s in the AIS equipment for approximately 4 h 39 min and 20 s.

The average reception interval for each piece of equipment and system, including the non-reception interval, was 127, 112, and 9.97 s for CHAMBADA, STUDY, and VTS data, respectively.

Table 4. AIS location data reception rate according to the equipment and sections that do receive data.

Sort	Time	Total Time	No. of Data Points	Target Value	Reception Rate	Receiving Interval
Chambada	10:08:42–14:54:42 -(10:33:42–12:19:44)	4 h 46 min 00 s -(1 h 46 min 02 s)	135	359	37.6%	79.9 s
Study	09:58:58–14:55:27 -(10:36:27–11:33:58)	4 h 56 min 29 s -(57 min 31 s)	158	478	33.05%	90.7 s
Average					35.33%	85.3 s

presents a recalculation of the reception rate by subtracting the section that did not receive data from the total reception target value (Table 3) and the section that did not receive data from the CHAMBADA and STUDY AIS equipment. The reception rate was calculated by subtracting the section of the CHAMBADA data (red) at 10:33:42–12:19:44 (a total of 1 h 46 min and 2 s) identified in Figure 1, Figure 2 and Figure 3; the reception rate was 33.05% when the section of the brown data (total of 57 min and 31 s) was subtracted. This reception rate was calculated under the assumption that data were received every 30 s from AIS Class B equipment.

The average reception interval for each piece of equipment and system, excluding the non-reception interval, was calculated to be 79.9 and 90.7 s for the CHAMBADA and STUDY data, respectively. In Table 3, the average reception interval was reduced compared with the calculation including the missed reception interval; however, the interval was eight to nine times higher than the average reception interval of 9.97 s of the VTS data.

3.2. Comparison of Storage Data in the AIS Equipment of Chambada with the Data Received by VTS

The ship departed from the Tongyeong Port and sailed as close to the island and coast as possible to identify the shaded areas of the AIS land antenna and VTS radar, turning counterclockwise around Geoje Island.

Figure 5 depicts a comparison between the data received from the Tongyeong Coast VTS in the Geoje Island sea experiment (Chambada) and the navigational records stored in the AIS equipment installed in the Chambada on an electronic map. The AIS data of the Tongyeong Coast VTS were indicated using green spots with every received signal and the AIS data were represented as blue dots every 5 s.

Figure 6 depicts an enlarged image of the lower part of Geoje Island with a VTS green spot. An abnormal phenomenon was observed in five areas (A to E). In certain sections, the location of the experimental ship was not received for a certain period; in other sections, the ship was stationary and a location error indicated by another sea point appeared for a while.

In zones A and B, only a part of the navigation was displayed in some places, while certain places had no navigation at all. The navigation was not displayed in chronological order, and the movement of the vessel along the route and its return to the previous location was confirmed near Hansando Island.

Figure 7 is an enlarged view of zone A (Figure 6). Here, the AIS data recorded in the AIS equipment (blue spot) moved sequentially with respect to time along the route, whereas the AIS data of the VTS moved from 10:31:56 to 10:32:06 and then reversed near the starting point. This resulted in an abnormal phenomenon in the form of a position error with respect to time, in which the 10:32:19 position appears.

Figure 8 illustrates an enlarged view of zone B (Figure 6). Here, after moving from 10:36:28 to 10:40:07, the location of 10:40:19 is displayed against the route, and a position error similar to that indicated in Figure 7 is identified.

Figure 9 depicts an enlarged view of zone C (Figure 6), where the experimental ship is displayed at a completely different location from the blue spot data in the AIS equipment. The ship followed the route from 11:03:08 to 11:03:35, moved to a different location at 11:04:36, and 13 s later, at 11:04:49, the ship returned to the original route before moving in a different direction from 11:04:59 to 11:06:07. Finally, it returned to the route at 11:06:16.

The aforementioned observation is similar to the phenomenon where the radar shows that, if the direction of the movement of the target is not confirmed, the target moves at the same speed as the previously recognized course until it recognizes it; Figure 9 estimates this as a position error caused by the tracking function of the radar.

Figure 10 depicts an enlarged view of zone D (Figure 6). Similar to zone C, a position error caused by the tracking function indicates the incorrect location for a certain period, and the radar-vector-jumping phenomenon occurs, in which the location jumps to another ship in the moving direction.

Figure 11 depicts an enlarged view of zone E (Figure 6), where the ship follows the AIS equipment record drag and then moves out of the track for approximately 2 min and 30 s from 11:19:46 to 11:22:08. The ship moves back to the position near 11:24:18. A similar phenomenon occurs from 11:28:47 to 11:29:48, and the ship moves to approximately 11:26:09 in the direction of nearly 38°, forming a straight track before returning to the position at 11:26:15. Subsequently, the phenomenon of the position error caused by the radar tracking function occurs repeatedly, moving the ship to 11:28:38 and then returning it to 11:28:47.

We observed that the course and speed were not recorded and the ship was stationary in the same position for approximately 2 min from 11:22:08 to 11:24:07, for nearly 24 s from 11:26:15 to 11:26:39, and approximately 1 min from 11:28:47 to 11:29:48.

3.3. VTS Radar Video

Figure 6 depicts the phenomenon of deviation and jumping without connecting the track of the experimental ship in the lower three areas of Geoje Island, where several ships, including fishing boats, small merchant ships, and guide ships, pass or operate. Although not included in the control area of the Tongyeong Coast VTS, this experiment aimed to identify an area where the location transmission and reception of the AIS equipment is not smooth and where numerous accidents occur along the coast.

First, the radar video data of the experimental ship from the Tongyeong Coast VTS were received and compared with the storage data of the AIS equipment. Figure 12 indicates that the radar tracking system of the VTS failed to track the experimental ship, which sailed in the direction of approximately 135–119° for nearly 1 min. The line indicating the AIS position and track and that indicating the position and track of the radar appear simultaneously on the screen; however, the actual track is displayed differently.

Figure 13 depicts zone D (Figure 6) using radar video data. The radar location and course of the experimental ship were sailing at 136° and abruptly changed to 90°, indicating that the radar tracking system failed to track the experimental ship, which continued sailing at 136°. Subsequently, the position was shifted to another fishing boat operating at Sobyongdaedo, resulting in a vector jumping error that left the ship for approximately 2 min and 30 s and displayed the ship at a completely different position.

Figure 14 depicts zone E (Figure 6) using the radar video data when the ship passes near Galgoddo. The location and course of the AIS and radar data are displayed differently. The radar location of the experimental ship exhibits a phenomenon where the tracking fails while the ship sails at 82°, remains stationary for approximately 2 min, moves to 37°, shifts to the location of the AIS, and moves to 37° in a different direction for 1 min and 50 s.

3.4. Allocation of Ship Traffic and AIS Slots

The AIS equipment stores the position coordinates determined using its own GNSS antenna in the VTS center and those obtained from the Ministry of Oceans and Fisheries via the relay antenna installed on an island or on land using very-high-frequency (VHF) radio waves. Figure 15 illustrates the location of the VTS center, the location of the AIS land relay antenna, and the location of the sea radar site (R/S) used for assisting the VTS control. The VTS close to Geoje Island, where the sea experiment was conducted, includes the Tongyong Coastal VTS, Masan Port VTS, Busan New Port VTS, and Busan Port VTS. Located closest to the west of Geoje Island, the Tongyong Coastal VTS receives radar data detected from Maemuldo R/S and Yokjido R/S at the bottom of Geoje Island, which can be used for ship control along with the VTS radar data. Four AIS repeaters around Geoje Island were installed in Yonghwasan, Silido, Gamcheonhang, and Eomgwangsan, indicated as blue triangles in Figure 15.

Figure 16 depicts the large-section classification areas of the ship’s location information obtained from the statistical information service website of the Ministry of Oceans and Fisheries, GICOMS. The western section of Geoje Island is classified under Sector No. 98, the southern and eastern sections are classified as Sector No. 99, and the northern section is classified under Sector No. 5099. Most of the experiments were conducted in Sector No. 99. The Tongyeong Port, which was the departure point of the experimental ship, is located in Sector No. 98, and certain routes pass through Sector No. 5099 north of Geoje Island.

The sea experiment was conducted on 20 September 2023, from 10:00 to 15:00 h.

Table 5. Hourly traffic of AIS vessels.

Sort	10:00 h	11:00 h	12:00 h	13:00 h	14:00 h	15:00 h	Total
98	136	130	131	141	143	133	814
99	207	203	212	202	204	202	1230
105	33	32	34	33	28	25	185
106	46	43	45	40	39	37	250
5099	170	174	179	179	179	179	1060
Total	592	582	601	595	593	576	3539

(Source: GICOMS).

lists the hourly traffic volumes of ships equipped with AIS equipment and passing through Sector Nos. 98, 99, 105, 106, and 5099 near Geoje Island. In the western part of Geoje Island (Sector No. 98), a total of 814 ships passed from 10:00 to 15:00 h. Sector No. 99, which comprises most of the routes of the Geoje Island experiment, such as the south and east sides of Geoje Island, exhibits the highest traffic volume with a total of 1230 ships in the same time period. Sector No. 5099 north of Geoje includes the Masan, Gohyeon, and Tongyeong routes, with a total of 1060 ships, which is higher than that observed in the area near the Tongyeong Port, the western part of Geoje. If the sea area was expanded to the Tongyeong Coastal VTS control zone, the traffic was high close to the trade port, with 185 and 250 ships in Sector Nos. 105 and 106, respectively.

According to [20], unmanned relay stations, such as radar, weather, and direction detectors, transmit information to the VTS center via R/S and relay stations to control transit vessels in areas far away from or outside the network distance within the VTS control zone. In the case of the Masan Port VTS, R/S exists in Dolsum, Somodo, Silido, Anjeong, and Jisepo; unmanned relay stations are also present in Ansan and Geojedo Daebongsan. In the case of the Tongyong Coastal VTS, R/Ss are present in Yokjido, Maemuldo, and Samcheonpo, with Yokjiko and Maemuldo functioning as relay stations. The AIS land relay antenna delivers the AIS information of the ship to the VTS, which divides the AIS information of multiple ships into two channels and allocates the information of approximately 2250 ships to the slots. The International Association of Navigation Marks (IALA) [21] reported that communication overload occurs when the communication slots used by the AIS in one channel exceed 50% (when more than 375 ships operate within 50 miles of the communication range of the base station).

Assuming that all 375 ships with AIS Class A equipment in the VTS control zone sailed below 14 knots, 100% of the slots were allocated because a single ship must receive locations every 10 s and receive six data points per minute. In other words, 375 ships must allocate a total of 2250 data points to the slots. However, if a few ships move faster than 14 knots or 23 knots, location information must be received and distributed every 6 or 2 s, respectively. A total of 3750 slots are required if all ships move faster than 14 knots, and up to 11,250 slots are required if all ships move faster than 23 knots. Therefore, when this overload occurs, the base station sequentially omits or adjusts the location information of the Class B equipment by increasing the reception time interval.

During the sea experiment, nearly 300 ships passed through the Tongyong Coastal VTS and Masan Port VTS control zones, which was less than 375 ships (below the 50% recommended by the IALA); however, calculations indicated that only 75 Class A ships with a speed of 14 knots or higher can send 2250 messages, resulting in 100% overload.

3.5. Radio-Shaded Area

Figure 17 depicts the radio-shaded area where the AIS location information was not properly received during the experiment. Typically, the AIS location information varies in the reception period depending on the equipment (Class A or Class B), and varies in units of 2–30 s depending on the navigation status and speed. In this experiment, most of the information was received every 30 s using Class B equipment, and the area where the location was not received for several tens of minutes and the area where the locations were received at an irregular period of more than 30 s were marked on the electronic chart.

In Figure 17, the VTS data exhibit a narrower and denser reception interval than other data intervals, which are a mixture of radar location information with periodically incoming AIS information. As indicated in Figure 18, the information was divided into the following three types: AIS, AIS and C-SCOPE extraction tracking (CET), and CET, where AIS received only AIS signals, CET received only radar data, and AIS and CET indicate that AIS and CET data were received simultaneously. According to a VTS official, CET is a C-SCOPE equipment and a radar tracking device from Kongsberg that receives radar and AIS data from Maemuldo or Yokjido, where an R/S is installed and is used for VTS control.

Of the 348 test data points received from the Tongyeong Coastal VTS, 35, 93, and 220 data points were classified into AIS, CET, and AIS and CET categories, respectively. While 63% of the data were a mixture of AIS and CET, 27% of the data received were only radar tracking data (CET). According to the latitude and longitude of the location data that received only CET, the time, location, course, and speed were all recorded to be identical. This resulted in a tracking failure, in which the radar tracking system failed to track the experimental ship that was stationary at the final location for a certain period and was marked before eventually disappearing. The lost radar location recovers the original location of the ship after a certain period using the location information of the AIS or other auxiliary equipment.

Figure 19 depicts a map of the location where the AIS land relay antenna was installed and the high-elevation topography around it. Although the Yonghwasan repeater is located on Mireuksan Mountain, it is installed at a height of approximately 280 m rather than at the top. Consequently, the blocking and interference of radio waves may occur owing to the geographical features of Mireuksan Mountain (458 m), Hansando (295 m), and Geoje Island’s Garyongsan (569 m), Sanbangsan (507 m), and Nochasan (557 m). Because most of the signals from Tongyong and the west side of Geojedo were received through the Yonghwasan repeater, we confirmed that no location data were received for a certain period in the terrain where radio waves were obscured.

Additionally, the location data were completely absent because the slot allocation was not received for 1–2 h owing to the communication overload caused by the excessive number of surrounding ships.

3.6. Graph of Radar Data

Figure 20 depicts the graphical representation of only a portion of the AIS equipment data obtained from the experimental ship and stored every 5 s that matches the VTS data section, indicating that changes in latitude and longitude move relatively smoothly.

Figure 21 illustrates the VTS data as a graph. In comparison with Figure 20, the latitude and longitude changes move irregularly and unnaturally rather than in the form of a smooth curve. The yellow squares indicate where the vector-jumping or position-jumping phenomenon occurred in the radar data (Figure 6).

Figure 22 depicts a graph that overlaps the graphs illustrated in Figure 20 and Figure 21. Overall, the movement of latitude and longitude is similar; however, the part where the radar error appears in the VTS data significantly deviates from the actual track, and the maximum deviation distance at this point is approximately 2.4 km on the graph.

4. Discussion

In this study, we performed a sea experiment near Tongyeong and Geoje Islands, which include islands and mountains with varying topographical features. The location and distance of land repeaters and the number of nearby navigation vessels were simultaneously investigated to extract relatively accurate data.

Because the reception rate of the radio waves related to the amount of data received from the AIS equipment of the ship relies on the performance of the equipment manufacturer or model, we conducted the experiments several times in the same area using various types of equipment to evaluate the reception rate considering radio wave interference. Furthermore, the model was used to simultaneously test several pieces of the same equipment to validate the results. The location was determined using the GNSS antenna of the AIS equipment, and the antennae of the two pieces of AIS equipment were installed close together to minimize the alignment error induced by the difference in antenna location.

The location data transmitted from the AIS equipment were delivered to the base relay station. Although the base station stores information in 2250 slots, the location may not be stored because of disorders or interference caused by various factors, such as radio wave communication problems and geographically shaded areas.

Therefore, the VTS receives radar location information from the R/S and data from the AIS equipment separately. The radar and AIS data received from this R/S are not stored in the GICOMS of the Ministry of Oceans and Fisheries but are managed by the VTS itself. According to the ship traffic control operation rules, the retention period of data, such as radar images, is 60 days.

According to the antenna height [4] of each AIS base station (

Table 6. Antenna height of the AIS base station in the island area.

Base Station	Antenna Installation Height (m)	Height of the Antenna (m)	Total
Heuksando	356	35	391
Nakpo	90	36	126
Geomundo	250	25	275
Yonghwasan	275	35	310
Eomkwangsan	493	20	513
Average	292.8	30.2	323

), the Yonghwasan base station in Tongyeong is 275 m above sea level, and the antenna height is 35 m. This implies that the actual antenna is 310 m above sea level. The Eomkwangsan base station at Busan Port is 493 m above sea level and the antenna is 20 m above sea level, indicating that the actual antenna would be 513 m above sea level. The average antenna installation height was 292.8 m, and the average antenna height was 30.2 m, indicating that the installation height of the AIS relay station antenna was not extremely high.

Typically, the AIS propagation distance in an open atmospheric state is 50 miles; however, previous studies [22,23,24] have reported that no atmospheric duct exists up to 80 km, and that the surface and elevated ducts appear after 500 and 200 km, respectively.

The four AIS land relay antennae in Geoje Island, the experimental area, were located at Yonghwasan, Silido, Eomkwangsan, and Gamcheonhang, within 50 miles of the VHF reach. An addition of four VTS centers resulted in a total of eight base stations. The distance between each relay antenna from the Yonghwasan relay antenna was a minimum of 34 km and a maximum of 66 km, which was closer than the VHF reach of 50 miles. Therefore, the experimental area overlapped the reach of three AIS repeaters. Geoje Island includes mountains that are over 500 m high and islands that are over 200 m high. This implies that radio waves do not reach several areas owing to topographical reasons, which may be related to the installation height of the relay antenna. If the antenna is extremely high, radio waves can be transmitted and received to a distant area without being affected by the surrounding terrain. However, a disadvantage is that a shaded area may exist below the antenna. If the installation height of the antenna is low, the shaded area under the antenna may be reduced; however, more shaded areas may exist because of the high terrain.

According to [4], factors that affect the reception rate of AIS data include slot occupancy, fading, and slot interference. However, the IALA [21] recommends that a smooth exchange of AIS data can be realized only when the VHF data load is managed within 50% with respect to slot occupancy. Regional slot occupancy must be analyzed by adding the message-receiving and collision slot occupancies based on frames of one minute. However, calculating an accurate slot occupancy is difficult because only the receiving slot occupancy can be analyzed using the message stored and utilized in GICOMS. Therefore, when the AIS receiver receives data, the error correction code is used to determine the presence of an error in data points; if an error exists, the data point is not stored. Therefore, the data point containing an error because of slot interference is omitted, resulting in an empty slot.

According to the AIS terminal used in ships [25,26,27,28,29,30], the communication range decreases when the communication load increases because of the automatic adjustment function of the slot, thereby excluding the allocation of communication slots for relatively distant ships. However, the AIS device in the base station that monitors the location of nearby ships may experience a slot collision, during which the location information from the ship cannot be received at fixed time intervals. This may result in a capture effect, wherein only strong signals are allocated slots, leading to a situation where normal traffic control services cannot be provided by the port control center or the situation room system.

The experimental area was within 50 km of the AIS antennae in Yonghwasan, Silido, and Busan New Port. Initially, the area was assigned to the Yonghwasan repeater slot when passing the west side of Geoje Island, and the location reception was omitted as it passed through the shaded area near Hansando Island. The prolonged missing time was attributed to the failure of the slot allocation of the experimental ship during the process of changing the antenna to the Busan New Port repeater when moving east of Geoje Island.

5. Conclusions

Because most fishing boats operate near islands or coasts, the location of the ship is often not transmitted owing to issues such as geographical radio transmission and reception problems and the priority slot allocation of controlled ships (e.g., merchant ships). Therefore, we performed an experiment on the southern coast of Geoje Island, a terrain with several high-altitude mountains and islands, to investigate the impact of geographical features on radio wave transmission.

Two identical pieces of AIS Class B equipment were installed on the experimental ship to compare the data received by the Korea Coast Guard’s VTS and GIS. We determined that the location of one piece of equipment was not received for approximately 57 min and that of the other piece of equipment remained unknown for approximately 1 h and 46 min. Furthermore, the location was stored only in the radar data of the VTS. The radar data received from the Tongyong Coastal VTS center comprised an hour of missing location data; however, these were the radar tracking data received from the unmanned or maritime R/S and the location data obtained from the AIS equipment installed in the R/S rather than the AIS information of the ship obtained via the land relay antenna. Some amount of radar data could not be tracked and appeared in a completely different location or were displayed in the same location without movement for a certain period.

The locations of the two pieces of AIS equipment of the same Class B model were recognized as separate pieces of equipment by the relay antenna because the storage time did not overlap. Moreover, the data received from the Tongyeong coastal VTS center were separated from the data received by the GIS with respect to time. However, this served as a complementary measure for the AIS ships passing through the topographically shaded areas.

Depending on the height of the ship antenna, the communication status, interference, geographic impact, and the number of navigation ships may not be completely transmitted when radio waves are transmitted to the VTS. Furthermore, the communication slot allocation may have been omitted owing to communication overload occurring during the process of allocating the transmitted information to the slot.

During the sea experiment, the ship passed through the Tongyeong Coastal VTS and Masan Port VTS control areas. Approximately 300 ships existed in the Tongyong Coastal VTS control area, which was less than the 50% or fewer ships (375) recommended by the IALA. However, based on the calculation that 75 Class A ships with a speed of 14 knots or higher could send 2250 messages, resulting in 100% overload, the number of relay antennae was deemed insufficient compared with the number of passing ships. Therefore, additional land AIS base stations must be installed near shaded areas, such as Hansando Island, for more accurate ship control.

We used marine survey ships that are similar to fishing boats in size to determine the effect of topography on AIS radio transmission in high-altitude islands and mountains. In the future, actual fishing boats that operate close to the coast or move by changing their routes must be investigated for a more accurate analysis.