posted 10-12-2013 07:33 PM ET (US)        

One beautiful afternoon in September 2013, for some unknown reason, and in the usual way of browsing the world-wide web, I happened to come across a website that listed the highest points in every county of Michigan. I used the resource of that website to locate the highest point in my county (Oakland). The summit of Oakland County happens to be not very far from my home, about ten miles, and it is located in an area that I have occasionally driven past, without ever realizing my proximity to the high spot. I became curious, and decided I would drive out to that area, to see what the view might be like and also to check radio reception with my AIS receiver.

There is a county road that goes very close to the summit, which is located on private property adjacent to the road. The owner of the property is a very old and venerable organization: the Benedictine Monks. About 60 years ago they had discovered this high spot and bought the property. I assume their interest was in being as close to heaven as possible. They built a monastery there, and a group of Benedictine monks live there. I was able to get very close to the summit by sitting in my car in their parking lot which was just a few hundred feet away from the exact high point, and on the downslope toward the Detroit side. This should have given me excellent reception of any AIS signals to the SE, covering Lake St. Clair, the Detroit River, and Lake Erie. I brought with me my portable AIS station.

My portable AIS station consists of the small em-trak R100 receiver that can run from a USB port, a MacBook Pro laptop, and a GlobalSAT BU-353 USB GPS receiver. For use in the car, I employ a mag-mount five-eights-wavelength whip antenna on the car roof. I spent a few minutes at the monastery receiving AIS signals. The results were disappointing.

Instead of a plot of dozens of ships, I saw nothing on my chart display. This indicated I was not receiving any ship AIS signals. Concerned, I began to monitor the NMEA output of the receiver, and, surprisingly, there were quite a number of AIS sentences being received. I was not sure why the chart plotter was not displaying any of these vessels. It had been a while since I had lashed all of his gear together into an AIS system, and I was concerned something might be malfunctioning. To help diagnose the problem, I captured a dozen or more of the NMEA sentences coming from the receiver, and stored them in a file.

The story now goes into hiatus for about a month, until late one evening I recalled the file of NMEA sentences captured by my AIS receiver at the monastery. These sentences are sent with a very compact encoding of data, and they are not human-readable. The data has to be decoded by a rather complicated method. Using an on-line AIS decoder website, I fed one of my mystery sentences into the decoder. The output was a big surprise.

Here is my input sentence:

!AIVDM,1,1,,B,403Ow4QunLkE:r3JNtH<Pa?0285t,0*3C

And here is the decoded output:

Packet Type = AIVDM
CHANNEL = A
Message Type = 4, Base Station Report
Repeat Indicator = 0, Default
User ID = 3669778
Year (UTC) = 2013
Month (UTC) = 9
Day (UTC) = 25
Hour (UTC) = 19
Minute (UTC) = 20
Second (UTC) = 40
Position Accuracy = 1, A DGPS-quality fix with an accuracy of < 10 ms
Longitude = -83.14031 West
Latitude = 42.2846066666667 North

This message was very interesting. It did not come from a vessel but from a base station. Using the decoded latitude and longitude, I located the position of the base station using GOOGLE EARTH. Its satellite view clearly showed a large tower at the transmitted position, and the tower was very familiar to me.

The mysterious AIS signal is coming from this location.

I had seen this tower many times as I drove across the Rouge River on Interstate-75. But I never associated the tower with an AIS base station. The next day I went on a road trip to investigate.

The tower's base is located near the Rouge River in an industrial area. I was not familiar with the streets in that area, but with the aid of Goggle Earth, I noted the ironically named thoroughfare that would lead me to the tower: Pleasant Street.

Pleasant Street was not very pleasant. Coming up from Jefferson Avenue, my progress was blocked by what appeared to be a very long train train stopped on the tracks. I had to take a detour, leading me to visit a few blocks that I would probably never otherwise have visited in the normal course of my life, and, to which, I probably will never return. Eventually, I was able to approach the base of the tower. With my first close-up glimpse, I saw why there might be an AIS signal emanating here: the top of the tower held a nine-element vertical direction-finding array which I concluded was most likely part of the RESCUE21 radio system of the U.S. Coast Guard.

The top-mounted nine-element vertical direction-finding antenna is typically used in the RESCUE21 System. This antenna seems to have an extra element, a tenth vertical mounted in the center. This may be part of a separate antenna system. The next antenna is a 4-element vertical collinear; I belive this is used for VHF receive. Below that is an unknown antenna. This might be used for NAIS or perhaps for UHF communication with other USCG resources. The lower 4-element collinear is believe to be used for VHF transmit.
(Based on some comments I found from a USCG technician's Facebook page--jimh)

These direction-finding antennas are key to the RESCUE21 radio system being able to get very fast fixes on received radio signals. In addition to the top-mounted DF array, there were three side mounted VHF antennas whose elements looked like they could be cut for the marina band. As further evidence, the stacked dipoles were all mounted to favor an azimuth pattern of broadside toward the general area of Lake St. Clair, the Detroit River, and Lake Erie. And these antennas looked very new, having a bright reflective surface, which would be atypical for anything that has been in this industrial area for more than a few months. Another clue came, again, from Google Earth.

The aerial view of the tower available from the imagery on Google Earth reminded me of what I had seen atop this big tower for years: a pair of vertically-polarized Yagi antennas pointing southwest, probably used for some sort of point-to-point VHF radio service. The shiny new DF array and the similarly new and shiny side-mounted stacked dipoles are not seen in a 2010 aerial view. This was more evidence that the present-day antennas could be linked to the recent installation of RESCUE21 radio service in the Great Lakes.

posted 10-13-2013 06:46 PM ET (US)            

Might there be a similar setup by me to check?

posted 10-13-2013 09:57 PM ET (US)            

--Lakeport, north of Port Huron; I had already located this site visually.

--Sarnia (a Canadian station)

--Grande Point (a Canadian station East of Lake St. Clair)

However, those stations were not the one I was actually looking for; that station is listed as CASCO. I think that refers to Casco township, in St. Clair County. I did see a big tower located on the top of a landfill site. I could not find the access road--didn't have Google Earth working in the car to help me.

I did record a lot of AIS sentences which could not be decoded by the on-line decoder. These AIS sentences are perhaps subject to revision and updating, and the decoder software may just not be aware of the latest updates to AIS NMEA protocol.

I also read that the USCG maintenance personnel visit these sites frequently to update the software that runs the equipment. That also suggests they might be sending some sentences that are very new.

posted 10-14-2013 09:16 AM ET (US)            

"This message is to be used by fixed-location base stations to periodically report a position and time reference."

http://gpsd.berlios.de/AIVDM.html#_type_4_base_station_report

According to the Wikipedia article on AIS:

"Base stations have a complex set of features and functions which in the AIS standard are able to control the AIS system and all devices operating therein. Ability to interrogate individual transponders for status reports and or transmit frequency changes."

posted 10-15-2013 12:14 PM ET (US)            

--Lakeport

--Sarnia

--Grande Point

--Wayne

--Leamington

posted 10-16-2013 11:53 AM ET (US)            

The scale of the tower is easy to judge by inclusion of my car in this image. The tower is an impressive structure.

Some radio tower sites can be hard to access by car, as they often tend to be located some distance off from decent roads and in rugged terrain. RFF LAKEPORT was just a few feet from a good country road. I could drive right up to it.

The RESCUE21 antennas are side-mounted well below the top of the tower. This view shows the familiar nine-element vertical direction-finding array, and the two vertical collinear arrays below it. I estimate their height to be about 200-feet above ground.

The USCG RESCUE 21 antennas share the tower with other users, and the top antennas are probably for cellular telephone service. There were also several microwave dish antennas, perhaps also part of the cellular installation for relaying data to other sites.

Side mounting usually distorts an antenna's pattern, but in this instance that is not a problem. The RESCUE 21 antennas are mounted on the East tower leg, facing the lake. Also, you can see that the collinear arrays are all oriented in the same direction. This causes their pattern to become non-circular and to favor the direction of the lake.

The USCG equipment is contained in its own, separate building. The sign on the door clearly identifies this location: RFF LAKEPORT.

At the base of the tower a large area is enclosed in a barb-wire-topped chain-link fence. Inside the fence are two equipment buildings, electrical power distribution panels, a generator, a propane tank, and supports for bringing transmission lines from the tower to the equipment buildings.

Five large transmission lines enter the USCG equipment building at Lakeport. The outer conductor of the transmission lines are bonded to a large earth ground before entering the building structure. This is done to encourage any lightning-strike currents to seek ground before they come into the equipment area.

Looking up from the south-face of the tower, you can see that the collinear arrays are oriented slightly to the northeast. The fiberglass-encased UHF receive-transmit antenna is not oriented toward the lake; it is probably used for communication to other land mobile radio stations like law enforcement or first responders.

The Coast Guard has clearly labeled some of their power distribution equipment on the site.

posted 10-17-2013 12:22 PM ET (US)            

I have contacted the developer of PolarView regarding handling of Type-4 messages. He has this enhancement on his ToDo list for the software.

It might be interesting to get some data from users of other types of AIS receiver and chart plotter combinations regarding the plotting of AIS base stations. It would seem that if one is in a region of the USA coast or on the Great Lakes, there should be a RESCUE21 site that is sending an AIS base station message. If the chart plotter is aware of this category of AIS message, it should present the AIS base station as a target on the chart display.

posted 10-17-2013 02:49 PM ET (US)            

I was part of an IT project to increase the accuracy of land surveys. Within our county, we have five or six passive receiving stations that continuously record where they "think" they are. This data is uploaded to one of our servers and can be accessed by a survey crew for correction values.

When the receivers (Leica) we're in place, USGS came and plotted coordinates to 1/10,000 of a Lat/Lon. I asked for some sort of reference that I could relate to, he pulled out a dime and said that it's the diameter of a dime.

posted 10-17-2013 03:29 PM ET (US)            

For example, the reported latitude varies from 43.12439 to 43.12445, or about 0.00006-degree. At latitude 43-degrees, one degree of latitude is 364412.8-feet

0.00006-degree x 364476.6-feet/1-degree = 21.9-feet

This is an error of about plus-or-minus 10.9-feet. That would be a typical error for a GNSS receiver with a precision fix augmentation such as from an space-based augmentation system (SBAS) from the FAA's Wide-Area Augmentation System (WAAS).

posted 10-17-2013 03:40 PM ET (US)            

An error of 1/10000th of a degree is 0.0001-degree, which at 43-degrees latitude is a distance of 36.4-feet, so that is one darn big dime that USCG guy is carrying around.

A dime is a distance of less than 1-inch, roughly, so we could say it was about 0.0825-feet. In terms of one degree of latitude at 43-degrees, this would be

1-degree/364476.6-feet x 0.0825-feet = 0.0000002-degree

which is two-ten-millionths of a degree.

posted 10-18-2013 10:40 AM ET (US)            

In the course of looking for data about towers, I did find an interesting listing of towers by township. Cf.:

http://www.homefacts.com/fcctowers/Michigan/St.-Clair-County/ Casco-Township.html

posted 10-18-2013 11:24 PM ET (US)            

The two sets of four-element stacked collinear dipoles are fed by separate transmission lines; one is used for transmit and one for receive. In both installations there is a third antenna, a smaller fiberglass vertical, and that would account for a third feed line. This is likely a UHF receive-transmit antenna.

That leaves two transmissions line to consider. Perhaps the nine-element vertical direction-finding array uses two transmission lines to bring the elements of the antenna do a specialized radio that can compare carrier phase between two antennas to get more precise bearings.

The notion of a nine-element direction-finding array is interesting in itself. I have no idea how it might actually work. The system apparently has very good resolution and can give an azimuth bearing with very good accuracy.

posted 10-19-2013 10:03 AM ET (US)            

http://www.navcen.uscg.gov/pdf/AIS/ais_data_by_station.pdf

Also see

http://www.navcen.uscg.gov/pdf/AIS/ITU-R_M1371-3_AIS_Msg_4.pdf

and

posted 10-23-2013 08:46 PM ET (US)            

--RFF CASCO

--RFF FORESTER

--RFF HALFWAY CORNERS

I had made guesses about the location of these three sites. I had to guess because these three are not transmitting as AIS Base Stations, so I could not get their exact position from their AIS Type-4 message sent every ten seconds; they were silent. But using some good estimates and with the help of the FCC registered tower database, I found very likely candidates for all three locations.

All my estimates were correct: I found the RESCUE21 sites without any problem. I have a bunch of images I captured from each site. I will post a few later.

I also had a great lunch with Don Mac in Port Huron, as I was between RFF CASCO and RFF FORESTER. We enjoyed some fine food at the VINTAGE TAVERN in Port Huron.

posted 10-26-2013 01:35 PM ET (US)            

To find the location, name, and MMSI of all the Canada radio sites on the Great Lakes took me about five minutes. I soon found a document from the government of Canada that listed all their marine radio sites for the entire country, gave their exact latitude and longitude, and included the marine mobile service identity of each site. In contrast, after days and days of searching for this sort of information about RESCUE21 sites in the USA, I have still come up with nothing. I have also made three telephone calls to the USA Coast Guard. Except for one call, which turned out to a wrong number and was answered by a very polite and helpful Bosun's Mate on a USCG Vessel, I only reached answering machines. The messages I left have all failed to solicit a return telephone call. I can't understand why this basic information seemed so hard to find.

However, while searching for information about RESCUE21, I did come across some amazing reports from the Government Accountability Office about the cost of RESCUE21, how the contracts have been administered, and how the project has been managed. These reports document significant cost overruns, even in an age when citizens of the USA are accustomed to hearing about government spending exceeding initial estimates.

According the GAO July 2012 "Report to the Ranking Member, Committee on Homeland Security and Governmental Affairs, U.S. Senate", GAO-12-629, the original life-cycle project cost of RESCUE21 was said to be $250-million. This cost has now grown by $2.4-billion to an estimated project life-cycle cost of $2.66-billion. This represents a cost overrun of more than 960-percent.

The RESCUE21 project has been the prior subject of a number of GAO reports, which all found that the costs were underestimated, implementation of the project was behind schedule, and key features of the proposed system were being dropped because they could not be implemented or could not be implemented in the timeframe needed to meet delivery requirements.

The RESCUE21 system is composed of about 425 radio sites. I believe that expenditures up to now have reached about $850-million. This gives us the simple math that each site would tend to cost on average about $2-million. For the $2-million one would have to acquire the land, build a tower, install the equipment, and maintain it for the life-cycle of the project. Based on my personal visits to five RESCUE21 tower sites, I found that all of these sites were using space on an existing tower, at an existing site, and making use of existing site improvements. This means we are down to about $2-million per site just to put up the antennas on an existing tower, install a pre-fab equipment building, furnish the equipment, and maintain it for the project life cycle.

Not all the planned sites have been installed. I suspect that the hard-to-do sites are the ones left undone. All the easy sites have been built. The leasing of tower space and facilities at existing improved sites has probably speeded the process of building out the system significantly. The government did not have to do any of the heavy lifting at the sites I visited; they were all very nice, improved, ready-to-move-in sites for leasing tower space and facilities. I suspect that part of the stunning cost overrun estimates for the project is the ongoing cost of leasing all of these facilities. Now, to be fair, of the five sites I visited, three were on towers owned by the State of Michigan, but I do not expect that the USCG is getting that tower space for free. I would expect they have to pay their way onto those towers. They certainly must be paying their way onto the other two towers I visited, operated by private companies.

The GAO reports also mention that a key element of the initial proposal for the system has been dropped or moved to a later implementation date: vessel tracking. This seems hard to understand. If you visit the website MARINETRAFFIC.COM you can see vessel tracking implemented--on what is more or less a hobbyist website. And there are many commercial companies that collect--with their own private satellites orbiting the earth--AIS data from vessels all over the planet and provide tracking data. But the implementation of RESCUE21 has appeared to drop the vessel tracking portion of the system. The USCG has said it will implement vessel tracking using other means.

The projected cost overruns of RESCUE21 perhaps explains why the USA Coast Guard has been making such a public relations fuss about the acceptance of each new RESCUE21 site or sector. There is no end to web articles by the USA Coast Guard and other boating sites that document the ceremonies undertaken by officers of Coast Guard and public officials to roll out a new RESCUE21 installation. I think the Coast Guard is on the offensive with this public relations campaign before someone notices that the cost of this project is almost 1,000-percent over the initial estimate.

posted 10-28-2013 04:35 PM ET (US)            

When one takes a drive these days into the country side, one of the most commonly seen structures is no longer just a barn or a silo, but a communication tower. Farmers seem to have been once again very fortunate to find a new source or revenue from their land: leasing of small parcels to operators of communication towers. The communication tower operators, in turn, lease space on their sites for antennas and equipment to many types of companies involved in wireless communication. While the general public probably refers to these sites as "cell towers" and thinks of them as being part of the cellular radio network of their wireless telephone providers, the towers also host many other customers, including government agencies. And, in many cases, the government agencies themselves are the operators of the tower sites, either directly our through contracts with companies to manage the sites for them. Here we take a brief look at what goes into constructing a site.

The first step in building a tower site must be to obtain the land or a lease for the land. This probably involves careful research into the location, the elevation and radio horizon, and consideration of any encumbrances such as environmental and zoning restrictions. Because these towers are typically more than 200-feet tall, coordination with the Federal Aviation Administration (FAA) must be done. Once a site has been chosen and all the legalities worked out, construction can begin.

The best site is often not immediately adjacent to a good road, so typically some sort of access road has to be constructed. Next, the tower foundations have to be dug and filled with reinforced concrete. This usually involves a lot of heavy trucks and equipment on site. When the tower foundations are done, the tower is erected. This may require a crane to come to the site. Many towers are sectional and are erected in 20-foot increments. Others are tubular and use a crane to lift long and large sections into place.

Once all the heavy construction is done, the site is then usually graded level. Underground electrical conduits are trenched and installed. Then a layer of sheeting is put in place to suppress growth of vegetation. Atop this a thick layer of gravel laid so the site will have good drainage. A fence is installed around the tower base. The site is ready for its first client, their antennas, transmitters, and receivers.

posted 11-05-2013 11:16 AM ET (US)            

Nationwide Automatic Identification System or N-AIS is a project to enhance AIS in the USA. For more information, see

http://www.navcen.uscg.gov/?pageName=NAISmain

posted 11-06-2013 10:00 AM ET (US)            

quote:

---

Since September 2007, the Coast Guard has had an interim NAIS capability at 58 U.S. ports and 11 coastal areas to receive data from out to 24 nautical miles offshore. The system currently receives 64 million AIS messages per day from approximately 7,900 unique vessels.

---

If we calculate the number of AIS messages the system would receive in five years of operation, i.e., from 2007 to 2012, there would be about 116-billion AIS vessel position messages available. This is with the system just being only partially deployed, in the build-out phase to the nationwide system, and not fully operational. When fully constructed, one might expect that such a nationwide system could intercept and record the AIS data from even more transmissions.

The USCG says:

quote:

---

The [N-AIS] system combines AIS data--such as vessel location, source and speed--with other government information and sensor data to form a holistic view of maritime vessel traffic near the continental U.S. and its territorial waters.

---

The "other government information and sensor data" might be interesting to speculate about. I have observed that in the area of the St. Clair River, Lake St. Clair, and the Detroit River, there are towers with surveillance camera all along the USA side of the border. I presume that these cameras are part of what is described above as "sensor data".

A recent local newspaper article discussed further methods of keeping watch on boat traffic, including the use of drone aircraft and sophisticated aerial look-down RADAR. See

posted 11-16-2013 11:14 AM ET (US)            

Holistic means in a manner following holism. Holism is:

quote:

---

...a theory that the universe and especially living nature is correctly seen in terms of interacting wholes (as of living organisms) that are more than the mere sum of elementary particles.

---

Cf.: http://www.merriam-webster.com/dictionary/holism?show=0&t=1384617843

I don't think that there is anything particularly holistic about gathering AIS data and presenting it to law enforcement agents in combination with data from other information gathering systems or databases. Maybe holistic has entered the argot of governmental regulatory and law enforcement agencies and is a buzzword for them.

Buzzword, noun.

--an important-sounding usually technical word or phrase often of little meaning used chiefly to impress laymen

Cf.: http://www.merriam-webster.com/dictionary/buzzword

Argot, noun.

--an often more or less secret vocabulary and idiom peculiar to a particular group

Cf.: http://www.merriam-webster.com/dictionary/argot

posted 11-16-2013 01:11 PM ET (US)            

They initially had, uh, "issues"--I know you love that word--in having any other government agency wanting to place anything additional on their towers. It took a suit by Oakland County along with a tower engineer's analysis to dispute the state position regarding tower wind loading was all wet.

Did I ask you, based on your excellent calculations, what typical receive range of the Rescue21 might be?. I'm wondering if the sites are located close enough to develop a plot, from a typical small vessel that's transmitting 8-feet off the water, on a ten-year-old VHF transceiver.

And, I'm going to guess that the overall costs are going to be closer to three-billion-dollars. I'd be curious on how many rescue ops are initiated per year where the lat and lon is an unknown.

posted 11-16-2013 04:27 PM ET (US)            

The antennas at other co-located sites where I found RESCUE21 may have been for commercial cellular carriers. As I recall the site at RFF FORESTER was probably like that. I will have to review my photographs.

I have also learned, since my visit to all these RESCUE21 sites, that certain of them have N-AIS facilities. I will have to double check the photographs to see if any separate antennas show up for the N-AIS radios. The N-AIS implementation goal is to have a receive range of 50-miles. That is probably based on a Class-A AIS transmitter, with 12.5-watts and mounted on a larger ship, on which the antenna would probably be at least 20-feet above the water. The N-AIS goal is for a transmit range of about 25-miles. I believe those distances are measured from the shore, not from the site. Some of the sites are located a few miles inland from the shoreline.

I will also have to review my photographs to see if I can find evidence of an additional antenna and transmission lines for the N-AIS radios. It seems like most of the sites I visited only had five transmission lines coming into the shack. I will have to compare the N-AIS sites with the non-N-AIS sites to see if there are any additional transmission lines. If there are not, then the AIS transmitter must be sharing an antenna with the RESCUE21 transmitter.

The RESCUE21 coverage goal is for about 20-miles off the coastline, assuming a transmitter power of 1-Watt with a transmitter antenna height of 2-meters above the water.

Tom--glad you jumped into the discussion. I think your word choice is better.

posted 11-18-2013 03:21 PM ET (US)            

Now we have to predict the path loss change with distance. There are several methods for doing this. The most favorable is to view propagation over water as free-space path loss. See

Calculating Path Loss of Water
http://continuouswave.com/ubb/Forum6/HTML/002143.html

Other models of radio signal propagation suggest higher path loss. See

Radio Propagation Over Water
http://continuouswave.com/ubb/Forum6/HTML/003380.html

If we use the most optimistic analysis, an increase of 14-dB in power will increase the path distance by a factor of five. If we use a more conservative analysis, we will only increase the distance by a factor of about 2.2. To see this explained in more detail read

RESCUE 21 VHF Radio Coverage
http://continuouswave.com/ubb/Forum6/HTML/003322.html

There is also likely to be an improvement in the signal if the antenna is raised. Actually, in some models (Egli's, for example), adding antenna height is estimated as being extremely influential on reducing path loss. The RESCUE21 coverage is based on an antenna height of 2-meters, or a little over 6-feet. In your question the transmitter antenna is 8 to 10-feet. The added height will add some distance to the path. It is hard to say precisely how much. This, again, depends on what model is used to predict path loss. The most favorable says that if the height is doubled, the path loss will decrease by half.

posted 11-18-2013 06:15 PM ET (US)            

Unless the distressed vessel is located pretty much equidistant between the two stations, the distressed vessel signal might not reach to the second site and therefore provide a (somewhat) accurate fix.

posted 11-20-2013 12:11 PM ET (US)            

--the accuracy of the beam heading solution of the radio direction finding antenna that produces a line of position, and

--the angle at which the two lines of position cross.

I do not have any information on the exact antenna that is being used for radio direction finding in RESCUE21, but it does seem to resemble in a general way the Rhode & Schwarz VHF/UHF direction finding array shown at

http://www.rohde-schwarz.com/en/product/ add050sr-productstartpage_63493-10506.html

The Rhode & Schwarz antenna has nine elements, and has a generally similar physical appearance. Let us assume that the RESCUE21 radio direction finding antenna is similar in its electrical properties to the Rhode & Schwarz antenna.

The Rhode & Schwarz antenna literature

http://cdn.rohde-schwarz.com/dl_downloads/dl_common_library/ dl_brochures_and_datasheets/pdf_1/ ADDx_multichannel_bro_en_0758-1106-12_v0400.pdf

says (on page 13 of the above) that the accuracy of the beam heading is about 1.6-degrees RMS.

The literature also says that Rhode & Schwarz radio direction finding antennas all use nine elements and competitors' antennas tend to use five elements. This also suggests more support for the RESCUE21 antenna being a Rhode & Schwarz antenna, or, if not Rhode & Schwarz, then one having similar specifications. Rhode & Schwarz describe their antenna as being "super resolution" antennas.

If we assume a signal is received from a transmitter at a range of 30-miles from the RESCUE21 site, a beam heading with an accuracy of 1.6-degrees suggests that the line of position at a distance of 30-miles might be off by as much as 0.8-miles (i.e. error = tan(1.6*) x 30-miles).

The best situation for getting a fix from two lines of position occurs when the lines of position cross at a 90-degree angle. As the crossing angle changes from 90-degrees, the ambiguity of the fix increases.