Foreflight caves and adds ADS-B traffic

denverpilot said:
The issue... way down at the low-level, is ADS-B is bandwidth-limited, severely. If they'd have chosen a data rate and technology (which triggers another huge set of problems in spectrum utilization) that could have sent ALL the data they wanted the service to provide to all aircraft within the coverage area of the antenna, the whole problem with the "hocky puck" goes away.

They did the hockey-puck, because the system can't support the whole area.

Once we hit a point where every aircraft has ADS-B and the hocky pucks are everywhere... we see which ADS-B sites channel-overload first...

There's a heirarchy of what will be dropped out of the data transmissions and in what order, and how they'll back off on updates of things less critical, once you hit that point, in the spec...

All that nice Marketing fluff about ADS-B, doesn't mention the "lesser" products will eventually be hard to get around busy hubs... right where most folks will ... oh, you know... probably want them. :)

Traffic is pretty high on the list, but weather... watch the updates slow to a crawl, eventually.

By then, the system will be well-entrenched enough to ask for a few more Billions to "upgrade" it... and they'll get 2000's tech to replace the 90's tech... in 2020 or so... haha...
Click to expand...
Well, Nate, it sounds like you figured out the Govt's acquisition system quite well! :D

That is at least almost exactly how acquisition works in the DoD.
 
Well said and analyzed Nate. I think this UAT ground infastructure is a waste of good money that could have been used to keep ATC controllers and TSA employees. After all everything that the UAT offers already exist in the form of XM\WX and TCAS without penalything the taxpayer. And it gets promoted by the "Free subscription" phrase while the tax payer pays 3 billions for it. If you can not afford an XM\WX $50/month subscription how can you afford a tie down for your plane. If the airlines are not equipping for UAT (they already have WX radar and TCAS) how is the general flying public (airlines) is going to benefit. When the next budget cut comes this will take the same path as Loran which was less money and benefited a wider public.

José
 
So it's basically a worthless feature without ADS-B out?
 
teethdoc said:
So it's basically a worthless feature without ADS-B out?
Click to expand...

In general, it seems so. I have heard mixed reviews in very busy areas (think NY, SF, LA). Mine comes next week, and I am curious to see if there are enough ADS-B Out planes in the SFO area to get me traffic. I am not keeping my hopes very high.
 
teethdoc said:
So it's basically a worthless feature without ADS-B out?
Click to expand...

Not completely. You just can't 100% rely on it as a primary information source.

If it saves someone's butt, fine... And it probably will, but just about any tool used in a way unintended by the designer can do that, aviation or non-aviation. :)

My experience with things like the Zaon (sp?) has been mixed. Co-owner has the non-3-dimensional model of that thing. And it's not ADS-B, of course... But it's a similar "odd use not originally intended" type of device...

In busy areas it whines too much and you naturally start unintentionally ignoring it. In open areas, when it squawks you sit up and pay attention.

So far, it's squawked about crop-dusters a couple thousand feet below, and we spotted them, which we'd never have done without it, but it hasn't saved our bacon yet.

In the terminal areas, it has an utter conniption-fit about joining a busy pattern and probably detracts from safety since you end up losing some outside the window attention and time looking at it briefly to see what it's forsaking out about.
 
denverpilot said:
Not completely. You just can't 100% rely on it as a primary information source.
Click to expand...

I'd say that it stands a good chance of detracting from safety. If a pilot assumes there is no traffic just because the ADS-B display sez there is nothing out there then there's a problem.
 
They did the hockey-puck, because the system can't support the whole area.

Once we hit a point where every aircraft has ADS-B and the hockey pucks are everywhere... we see which ADS-B sites channel-overload first...
Click to expand...
Interesting. I assumed that they implemented the hockey puck strategy in order to motivate ADS-B Out installations. It's certainly clear that when ADS-B Out participation is high the ground stations will be broadcasting essentially all the traffic, at which point the hockey pucks will functionally cease to exist and certainly will not be saving any bandwidth.

But this is an easy analysis to do and it's hard for me to believe that they are installing a system that by design will be overloaded. Also, given that each traffic report can't be more than a few hundred bits of data a few times a minute it seems like there ought be plenty of channel capacity.

Nate, have you actually run some numbers? What do you think the capacity of the channel is and what do you think the load looks like? I obviously don't have any data to contradict what you're saying but just looking at the logic of the situation makes me wonder how your assertions can be true.

(Others: please spare me the anti-government rants and conspiracy theories.)
 
I'd say that it stands a good chance of detracting from safety. If a pilot assumes there is no traffic just because the ADS-B display sez there is nothing out there then there's a problem.
Click to expand...
Agreed. Mode S traffic has this problem now. When I am flying out near the edge of the TRACON radar coverage it is easy to forget that there might be traffic a little below me that is below radar coverage and hence not showing up on my fish finder.
 
Ya'll got it backwards. The hockey puck goes with the aircraft that have ADSB installed and it lights up those that don't have it. As more and more get ADSB, fewer occasions of the hockey puck are needed to light up non participants. If every aircraft had ADSB, all the traffic would be air to air and the ground station isn't involved at all. Also, say that twelve aircraft light up one mode C recalcitrant, there is only one broadcast from the ground. With ADSB, the B is for broadcast and ground TISB is not addressed to a specific hockey puck.

With dual frequency receivers becoming the norm, there is less and less need for ADSR. Sometime after 2020, the ground stations won't be transmitting much traffic. There is no bandwidth issue whatsoever on UAT. 1090ES is different and has greater chance for congestion, with ground radars, TCAS, and Multilateration systems in the mix in addition to the ADSB traffic, steps have been taken to address congestion. With UAT, the cells are non directional, many if not most of the 1090ES cells are split into four directional segments and power is adjusted to meet the client requirements.

Also, with the version of ADSB that is required for 2020+, the ground stations generate much fewer broadcasts because the airliners who generate most of the traffic today will not have ADSB In installed and the ground station will send them nada.
 
airdale said:
Nate, have you actually run some numbers? What do you think the capacity of the channel is and what do you think the load looks like? I obviously don't have any data to contradict what you're saying but just looking at the logic of the situation makes me wonder how your assertions can be true.
Click to expand...

I had found an article about it... I'll see if I can find it. One of the downsides of using all these "mobile" devices nowadays is that I'm rarely at the "real" laptop... haha... and I think I saved that one, but I'm not sure.

Basically the transponder-based "extended squitter" system was overloaded the second it was deployed. If I recall, UAT can handle roughly 100 aircraft if ONLY doing traffic, and none of the other data muxed into the UAT channel (weather, etc.).

There was a field test and some math run by someone in California (being they have a heavier traffic density than anywhere else, they had more curious engineers than anywhere else, basically...) so a google search of UAT and California test may find it...

The UAT 978 MHz channel rate is basically 1Mb/s. It's not going to saturate right away, but it'll get there, if everyone in a busy area is "participating".
 
The LA Basin 2020 traffic assumptions have been modeled with UAT and are included in the UAT MOPS in an appendix to RTCA DO 282B. Although this is a public document, it is not free and costs $670 in electronic form. You can purchase it from RTCA.

The UAT design assigns the first 200 MS of the for FISB and the rest is available for traffic, both airborne and GBT TISB. There are 3200 slots per second available. There is no interference between FISB and TISB as they do not use the same portion of the transmission space available. The FISB space is allocated into 32 channels so that adjacent cells don't overlap one another. There are buffers of unused spectrum that support up to a 200 KM transmission through the atmosphere and they are allocated to reduce data collisions between adjacent slots. Within the 3200 available slots for position data, each user uses one slot per second with a pseudo random allocation of the slot by each user on subsequent broadcasts. Data collisions will occur, but within one second the data is updated on its new random slot.
 
Yeah. I found this (free) report, too:

http://adsbforgeneralaviation.com/w...SBS-Description-Doc_SRT_47_rev01_20111024.pdf

"The overall purpose of this document is to describe the services provided by the Surveillance and Broadcast Services System (SBSS) over the Air Interface to ADS-B Equipped aircraft. Accordingly, this acts as a ‘users guide’ for ADS-B avionics vendors, customers, and users to the services that the SBSS provides to ADS-B equipped aircraft. ... "

It describes the frame structure and a lot of other technical details. I'm too lazy to study it in detail but it looks like the design will handle a few thousand TIS-B traffic reports coming from each transmitter. Interestingly, reports to clients in a terminal environment are more frequent than reports to cllients enroute. Very logical.

The host site http://adsbforgeneralaviation.com/ is maintained by an individual but it looks like it has a lot of information that is pretty germane to our small bugsmasher world.

BTW thanks for pointing out that the ground stations don't have to report participating traffic. I hadn't thought that one through.
 
Very interesting Nate. Thanks for the explanation. Question, how is the mutual interference at the aircraft side is handle. Example: At 30,000 feet I could be in view of 20 UAT stations all transmitting traffic and weather. This is the same problem with cell phones and unlike UAT single frequency (978MHz) cell phones use multiple frequencies.

José
 
Piloto said:
Very interesting Nate. Thanks for the explanation. Question, how is the mutual interference at the aircraft side is handle. Example: At 30,000 feet I could be in view of 20 UAT stations all transmitting traffic and weather. This is the same problem with cell phones and unlike UAT single frequency (978MHz) cell phones use multiple frequencies.

José
Click to expand...

The system is designed to handle interference out to approximately 200 NM. Assuming an antenna is at 100 MSL, the line of sight at 200 NM is over 35000 MSL. At these distances, the receiver will be able to discriminate the closer ground stations.
 
John Collins said:
The system is designed to handle interference out to approximately 200 NM. Assuming an antenna is at 100 MSL, the line of sight at 200 NM is over 35000 MSL. At these distances, the receiver will be able to discriminate the closer ground stations.
Click to expand...

John's got it. One of the reasons you don't see many UAT ground stations on top of tall mountains.

UAT has pretty decent bandwidth overall. Time will tell. The other likely "fixes" will include augmenting and moving sites and changing power levels to handle dense areas better.

Similar to the mega-growth phase of cellular...

It's fun to analyze. It'd be REALLY fun to watch UAT handle say, if every aircraft at OSH were participating... That's a real-world that'd be a system crusher. Heheh.
 
Current GBT sites typically have 4 directional service volumes for 1090 MHz, but a single omni-directional antenna for UAT. The UAT could easily be updated to break it into separate service areas as traffic volumes get higher.
 
John Collins said:
The system is designed to handle interference out to approximately 200 NM. Assuming an antenna is at 100 MSL, the line of sight at 200 NM is over 35000 MSL. At these distances, the receiver will be able to discriminate the closer ground stations.
Click to expand...

That is assuming only the ground UAT stations are on 978MHz. But keep in mind that airborne 978 MHz transponders will also contribute to the mutual interference. And there is greater range between two planes (300nm) at 10,000 feet than to the ground. In an area of 200nm x 200nm with a density of 1 plane per 10 sq.nm you could be easily listening to 4,000 airplanes plus the ground UAT stations. What most likely is going to happen is that the airborne receiver will be overwhelmed trying to sort out the crowd. The same problem that happens with airborne cell phones. In a crowded environment you will probably get a broken weather picture with long lagging times.

José
 
Last edited:
John Collins said:
Why?:confused:
Click to expand...

As the number of stations (ground & airborne) increases the chances of getting complete weather data in a short time decreases due to the mutual interference. The receiver needs to wait for another string of clean data because the pevious one was corrupted by interference. This is not uncommon in surveillance ATC radars or TCAS. The difference is that weather data strings are much longer than a Mode C reply. You could get another Mode C data string within milliseconds while for weather it could be minutes before a clean string is received. Even today weather from UAT stations at times comes in a broken fashion. The problem with broken weather display is that it can mislead the pilot in believing that there is no weather in a broken area.


José
 
Last edited:
Piloto said:
As the number of stations (ground & airborne) increases the chances of getting complete weather data in a short time decreases due to the mutual interference. The receiver needs to wait for another string of clean data because the pevious one was corrupted by interference. This is not uncommon in surveillance ATC radars or TCAS. The difference is that weather data strings are much longer than a Mode C reply. You could get another Mode C data string within milliseconds while for weather it could be minutes before a clean string is received. Even today weather from UAT stations at times comes in a broken fashion. The problem with broken weather display is that it can mislead the pilot in believing that there is no weather in a broken area.

José
Click to expand...

So the fact that the first part of the UAT message space (first 200 Msec) is fully dedicated and non overlapping with any of the portion assigned to traffic (800 Msec) isn't relevant. Furthermore, the fact that the FISB data is divided into one of 32 channels with each cell on its own channel to protect against adjacent cell collisions for up to 2.5 times the cell width, also isn't relevant.

I suggest you review the technical details of the technology before you make such unfounded claims.
 
John Collins said:
So the fact that the first part of the UAT message space (first 200 Msec) is fully dedicated and non overlapping with any of the portion assigned to traffic (800 Msec) isn't relevant. Furthermore, the fact that the FISB data is divided into one of 32 channels with each cell on its own channel to protect against adjacent cell collisions for up to 2.5 times the cell width, also isn't relevant.

I suggest you review the technical details of the technology before you make such unfounded claims.
Click to expand...

Another interference source is that the DME band overlaps the 978MHz channel. Also the onboard transponder replying on 1090Mhz with 200watts minimum is another source. A transponder interrogated by TCAS and ATC could easily be jamming the UAT signal 1,000 times per second.

BTW is not 32 channels but time slots

José
 
Last edited:
In an area of 200nm x 200nm with a density of 1 plane per 10 sq.nm you could be easily listening to 4,000 airplanes plus the ground UAT stations.
Click to expand...
True enough. And if it was one plane per sq. nm then it would be 40,000 airplanes. One airplane per 100 sq. nm then 400. The math is easy. But let's try for a realistic density number:

IIRC from tours, the maximum number of airplanes that an ARTCC sector can be working is 16. I don't know what it is for a TRACON but I doubt that it is more. So your 4,000 airplane scenario might require as many as 250 controllers/sectors for 40,000 sq. nm.

Given that the Oakland ARTCC controls 140,000 (non-oceanic) sq. mi. with 18 sectors, it could be that your one plane per sq. nm. scenario is a bit over the top. Doing the math on Oakland would say that they would be overloaded with one plane per 500 sq. mi.

Obviously, all airplanes are not IFR and hence not a load for the centers but the point is still the same: Even if only 1/10th of the airplanes were IFR that would still be only one plane per 50 sq. mi., just 20% of your deliberately-alarming hypothetical scenario.
 
airdale said:
True enough. And if it was one plane per sq. nm then it would be 40,000 airplanes. One airplane per 100 sq. nm then 400. The math is easy. But let's try for a realistic density number:

IIRC from tours, the maximum number of airplanes that an ARTCC sector can be working is 16. I don't know what it is for a TRACON but I doubt that it is more. So your 4,000 airplane scenario might require as many as 250 controllers/sectors for 40,000 sq. nm.

Given that the Oakland ARTCC controls 140,000 (non-oceanic) sq. mi. with 18 sectors, it could be that your one plane per sq. nm. scenario is a bit over the top. Doing the math on Oakland would say that they would be overloaded with one plane per 500 sq. mi.

Obviously, all airplanes are not IFR and hence not a load for the centers but the point is still the same: Even if only 1/10th of the airplanes were IFR that would still be only one plane per 50 sq. mi., just 20% of your deliberately-alarming hypothetical scenario.
Click to expand...

For real time aircraft density go to http://www.flightradar24.com/ and count how many airplanes are 100nm on the clear. BTW the radar site does not include those VFR planes talking to nobody.

Question: what is the added benefit of the UAT for those with XM\WX or weather radar and traffic detection capability (TAS/TCAS)?

José
 
For real time aircraft density go to http://www.flightradar24.com/ and count how many airplanes are 100nm on the clear. BTW the radar site does not include those VFR planes talking to nobody.
Click to expand...
Instead of doing kindergarten stuff, let's keep coming at it analytically.

The continental US ARTCCs' approximate annual traffic in 2020 is estimated to be about 49,000,000. (http://www.faa.gov/data_research/aviation/IFR_Forecasts/media/artcc10.pdf) Assume they are all concentrated in a 12-hour day and the average flight duration is 2 hours. That would result in a system load of about 23,000 aircraft. The contiguous US is about 3,000,000 sq. mi. (http://en.wikipedia.org/wiki/Contiguous_United_States), so the density would be one airplane in 130 sq. mi. That's fairly consistent with the rougher calculation that said Oakland's overload point is an airplane per 50 sq. mi. -- the estimates check.

At one per 130 sq. miles, your hypothetical 200 x 200 nm. area would contain about 400 airplanes. Maybe it's 200 and maybe it's 800, but it ain't 4,000.
 
Tried the traffic feature today, after updating the Stratus firmware a couple of days ago. Pretty cool!

Foreflight_130407.jpg


By the way, note that N72PV is shown as being 900' below me. I was at 900' AGL when that screenshot was made. N72PV was on the ground taxiing at KVUO.
 
Pilawt said:
Tried the traffic feature today, after updating the Stratus firmware a couple of days ago. Pretty cool!

Foreflight_130407.jpg


By the way, note that N72PV is shown as being 900' below me. I was at 900' AGL when that screenshot was made. N72PV was on the ground taxiing at KVUO.
Click to expand...

Do you have ADS-B out?
 
Two aircraft in the Atlanta area and only one of them airborne is not what I would consider an ADS-B rich environment. Keep looking out the windshield for the other 99 aircraft.
 
That's promising, I'm still looking forward to seeing how well traffic works in the KSFO airspace without ABS-B out.
 
John Collins said:
Two aircraft in the Atlanta area and only one of them airborne is not what I would consider an ADS-B rich environment. Keep looking out the windshield for the other 99 aircraft.
Click to expand...
Atlanta ... ? This is Portland, and yes, I am aware of the system's limitation. I'll take all the traffic information I can get.
 
Pilawt said:
Atlanta ... ? This is Portland, and yes, I am aware of the system's limitation. I'll take all the traffic information I can get.
Click to expand...

Understood, sorry for the error on location. As I previously stated, it won't lie if it does show traffic, rare as that might be, it is the other 99% that it doesn't report that are the issue. As long as you are aware of its enormous limitations without ADSB Out, good.
 
You must log in or register to reply here.
Back
Top