6 ms·

With radar, it's actually worse! Received signal strength is proportional to the fourth power of distance (1/r^2 on the way to the target, then 1/r^2 on the wa

With radar, it's actually worse! Received signal strength is proportional to the fourth power of distance (1/r^2 on the way to the target, then 1/r^2 on the way back to the radar). That's part of the reason 5G is such an issue, despite the ostensibly-large guard band. You need really good filters on the cellular 5G equipment to make sure it doesn't leak into the radar band.

anfilt 5y ago
Let's not also forget about drift on these filters. Those components will experience drift from age. That drift can also be accelerated from quiet a few factors such as heat ect...
sandworm101 5y ago
And many hundreds of emitters. All those towers and phones pumping out 1s and 0s will occasionally line up in a way thay looks like a higher/lower frquency. The aircraft looking down sees across multiple cells. All sorts of strange combinations might occur. Thats what guard bands are for.
nwallin 5y ago
> With radar, it's actually worse! Received signal strength is proportional to the fourth power of distance (1/r^2 on the way to the target, then 1/r^2 on the way back to the radar). I think that's specifically true when you're pointing a radar at an aircraft, but is not true for radar altimeters. When you're pointing a radar at an aircraft, the angular size of the aircraft is 1/r^2, so 1/r^2 of the emitted radiation hits the aircraft. Once emitted from the aircraft, the angular size of radar receiving dish is 1/r^2, so 1/r^2 of that radiation is detected. With a radar altimeter, you're pointing the radar at the ground. The angular size of the ground is 50% of the sphere, whether you're one foot above the ground or 100,000 feet. (...ok, maybe it's like 49.5% of the sphere...) As long as you're pointing it down-ish, 100% of the emitted radiation hits the ground. There is only one 1/r^2 relationship, on the return leg.
- lutorm 5y ago
  Well, 100% of the emitted radiation hits the ground in terms of total power, but the radar is looking for the _closest_ reflection, which is some relatively small solid angle straight down. It doesn't change your conclusion though, that solid angle is independent of the distance to the ground.
  - nwallin 5y ago
    > Well, 100% of the emitted radiation hits the ground in terms of total power, but the radar is looking for the _closest_ reflection, which is some relatively small solid angle straight down. This fixed angle paints a larger and larger circle as altitude increases. If you double your altitude, you quadruple the size of the surface that's reflecting radar back at you. A bunch of stuff cancels out and you end up with 1/r^2 at the receiver, which is what we started with.
    - steerablesafe 5y ago
      An alternative explanation is that you treat the ground as a flat mirror, and the radar detects the distance of your mirror image. The distance between you and your mirror image is (2r), so the signal strength you receive from your mirror image is ~1/(2r)^2, still follows 1/r^2 .
- fennecfoxen 5y ago
  I’m not arguing with the idea most of the signal gets to the ground, but, I’m worried that you get specifics quite wrong. Planes don’t want a reflection from 50% of the sphere. They want it from directly below. Other land is further from the plane and obscures the measurement. But they’re not broadcasting to the whole sphere either, because the radar instrument is not an isotropic radiator. That antenna has gain. It’s louder / more sensitive in a specific direction. Probably by a lot. You can’t make perfect antennas, but you don’t need any signal at all from most of the sphere, so you can gain many decibels. The angle of the beam you do have is probably fixed, and being fixed would imply that the area swept by the core of the beam is indeed proportional to the square of the altitude. But I don’t think two trips through an n^2 distance is n^4 unless I’m missing some of the physics. Shouldn’t it be (2n)^2 and this still proportional to the square of the altitude not to the hypercube? (And affected by scattering and bad reflectivity from the terrain, which seems likely to be independent of altitude?) This is different for detecting random entities in free space where most of the first n^2 trip really is lost. So it’s like you say but for different reasons??!?!
hilbert42 5y ago
"That's part of the reason 5G is such an issue, despite the ostensibly-large guard band." As I keep repeating here, if excessive or debilitating interference is being experienced between services under normal operating conditions then there has been a failure of Spectrum Management planning - absolutely no question about it. OK, we know such interference has happened, so who was responsible for the planning fuck-up? I've no argument about filters, RFI etc., what most are saying about the engineering issues makes sense. My argument is that the spectrum planning has failed and that much of the reason can be put down to deregulation in the 1980s when many governments essentially washed their hands of Spectrum Management (succumbed to commercial pressures and overly loosened engineering standards and now we're paying the penalty). It seems to me (given the international nature of the problem with non-US registered aircraft flying into US airspace, etc.) that we need to go back to the core of the problem and acknowledge these past 'regulations' errors before the problem can be fixed properly. Perhaps all radar altimeters will have to be changed (respecified with much tighter filters, better intermod figures, etc.) but given the fuck-up who should pay? After all, it's not the FAA's fault nor that of the aircraft industry (see my comments in other posts about the ITU and WRC). As I see it, this is a political issue, not a technical one.

anfilt 5y ago

Let's not also forget about drift on these filters. Those components will experience drift from age. That drift can also be accelerated from quiet a few factors such as heat ect...

sandworm101 5y ago

And many hundreds of emitters. All those towers and phones pumping out 1s and 0s will occasionally line up in a way thay looks like a higher/lower frquency. The aircraft looking down sees across multiple cells. All sorts of strange combinations might occur. Thats what guard bands are for.

nwallin 5y ago

> With radar, it's actually worse! Received signal strength is proportional to the fourth power of distance (1/r^2 on the way to the target, then 1/r^2 on the way back to the radar). I think that's specifically true when you're pointing a radar at an aircraft, but is not true for radar altimeters. When you're pointing a radar at an aircraft, the angular size of the aircraft is 1/r^2, so 1/r^2 of the emitted radiation hits the aircraft. Once emitted from the aircraft, the angular size of radar receiving dish is 1/r^2, so 1/r^2 of that radiation is detected. With a radar altimeter, you're pointing the radar at the ground. The angular size of the ground is 50% of the sphere, whether you're one foot above the ground or 100,000 feet. (...ok, maybe it's like 49.5% of the sphere...) As long as you're pointing it down-ish, 100% of the emitted radiation hits the ground. There is only one 1/r^2 relationship, on the return leg.

lutorm 5y ago
Well, 100% of the emitted radiation hits the ground in terms of total power, but the radar is looking for the _closest_ reflection, which is some relatively small solid angle straight down. It doesn't change your conclusion though, that solid angle is independent of the distance to the ground.
- nwallin 5y ago
  > Well, 100% of the emitted radiation hits the ground in terms of total power, but the radar is looking for the _closest_ reflection, which is some relatively small solid angle straight down. This fixed angle paints a larger and larger circle as altitude increases. If you double your altitude, you quadruple the size of the surface that's reflecting radar back at you. A bunch of stuff cancels out and you end up with 1/r^2 at the receiver, which is what we started with.
  - steerablesafe 5y ago
    An alternative explanation is that you treat the ground as a flat mirror, and the radar detects the distance of your mirror image. The distance between you and your mirror image is (2r), so the signal strength you receive from your mirror image is ~1/(2r)^2, still follows 1/r^2 .
fennecfoxen 5y ago
I’m not arguing with the idea most of the signal gets to the ground, but, I’m worried that you get specifics quite wrong. Planes don’t want a reflection from 50% of the sphere. They want it from directly below. Other land is further from the plane and obscures the measurement. But they’re not broadcasting to the whole sphere either, because the radar instrument is not an isotropic radiator. That antenna has gain. It’s louder / more sensitive in a specific direction. Probably by a lot. You can’t make perfect antennas, but you don’t need any signal at all from most of the sphere, so you can gain many decibels. The angle of the beam you do have is probably fixed, and being fixed would imply that the area swept by the core of the beam is indeed proportional to the square of the altitude. But I don’t think two trips through an n^2 distance is n^4 unless I’m missing some of the physics. Shouldn’t it be (2n)^2 and this still proportional to the square of the altitude not to the hypercube? (And affected by scattering and bad reflectivity from the terrain, which seems likely to be independent of altitude?) This is different for detecting random entities in free space where most of the first n^2 trip really is lost. So it’s like you say but for different reasons??!?!

hilbert42 5y ago

"That's part of the reason 5G is such an issue, despite the ostensibly-large guard band." As I keep repeating here, if excessive or debilitating interference is being experienced between services under normal operating conditions then there has been a failure of Spectrum Management planning - absolutely no question about it. OK, we know such interference has happened, so who was responsible for the planning fuck-up? I've no argument about filters, RFI etc., what most are saying about the engineering issues makes sense. My argument is that the spectrum planning has failed and that much of the reason can be put down to deregulation in the 1980s when many governments essentially washed their hands of Spectrum Management (succumbed to commercial pressures and overly loosened engineering standards and now we're paying the penalty). It seems to me (given the international nature of the problem with non-US registered aircraft flying into US airspace, etc.) that we need to go back to the core of the problem and acknowledge these past 'regulations' errors before the problem can be fixed properly. Perhaps all radar altimeters will have to be changed (respecified with much tighter filters, better intermod figures, etc.) but given the fuck-up who should pay? After all, it's not the FAA's fault nor that of the aircraft industry (see my comments in other posts about the ITU and WRC). As I see it, this is a political issue, not a technical one.