The uniformity of coverage and the spatial distribution of Quality of Service. A thread. 🧶🐈 ”Y’know if we do this right, we will end mobile dead-zones. We will have connectivity everywhere”. - @JRosenworcelFCC How exactly, technically, will @AST_SpaceMobile do that? 1/n

This is a population density map of the US. Generally speaking at these population density peaks towers / small cells are, and will be, cost efficient means of coverage. But most of the map is sparsely populated. We will discuss that part. 2/n

For rural tower coverage lowband (sub 1GHz) is essential. In free space it propagates further than higher spectrum. 2/n

Terrain and foliage add to the signal path loss. As terrestrial signals work. The entire fresnel zone needs to be clear of obstructions or reflecting signals will cause destructive interference. This is rarely the case. Fresnel zone most often compromised. 3/n

This results in rural towers having very good signal near. But for most of the area covered and certainly so near cell edges (halfways between two towers) signal to noise and interference ratio is bad and spectral efficiency is low. Causing dead-zones and low data rates. 4/n

Adressing that with more densely packed towers (which wirks in cities) is very costly in a rural setting. Adressing it with more prime lowband - far reaching- spectrum is also very costly and still doesn’t reach the dead zones. 5/n

Theoretical blanket coverage using rural towers looks a bit like this. Imagine a flat USA without much foliage or mountains. Midwest/agricultural plain. Then the peaks (dark green) are at tower location. The troughs (light green) are halfway between. Chart: CatSE 6/n

Here is another visualisation. The nails is where throughput is high. The bed (most parts) is where it is not. 7/n

AST SpaceMobile’s satellites. Called BlueBirds generate thei cells by illuminating them with a highly directive beam. It can be highly directive because of many elements on its huge area. (90x90 antenna elements for a Block1 lowband BlueBird.) Like lower right image. 8/n

Notice the rounded tip of the beam. That is how signal strength is distributed within the cell. That rounded curve, at its tip, doesn’t drop of rapidly. 9/n

You get a distribution of signal strength (Actually SNIR and thus Spectral Efficiency, but anyway) that looks like this coming from a fully deployed SpaceMobile constellation. Chart: CatSE 10/n

The actual throughput will fluctuate a bit, as LEO satellites are not static the way towers are. AST SpaceMobile will illuminate cells from multiple angles in a dynamic system. Fully deployed it will illuminate each cell from multiple satellites. 11/n

But this type of distribution of coverage is a good representation of the mean signal strengths. If you want to imagine the dynamic picture imagine that these peaks in blue oscillates a bit up and down and side to side. 12/n

So how does these two distributions compare? 13/n

If you deploy both networks there will be a lot of area (yellow in this chart) where throughput is better with a satellite. This will include all dead zones, except subterranean and such. There will also be islands where towers would outperform. Chart: CatSE 14/n

We haven’t talked economics yet. But let’s do that. What does this mean for rural coverage? Remember this? It is a visualisation of demand. It is fractal. A rural landscape also has it’s peaks like this. Like a local little town. 15/n

What AST SpaceMobile allows is for a hybrid coverage of rural areas that looks a like this. Blanket sat-coverage giving ”good enough” coverage _everywhere_ as represented by the bed. And excellent coverage at select high demand locations through towers, where profitable. 16/16

More on the subject, here. Prior DD. https://t.co/dhQkgLJSBP

And more here. Prior DD. https://t.co/ZrdrpfIYiX

The signal path length you experience varies 1–2x under the sats. Distribution of coverage within a sat-cell is due to the HPBW or 3dB drop from beam center to beam edge With towers distance varies 1000x+ and signal drop is ~50dB. Signal is subject to Fresnel drop from terrain.

While you are here… Let’s look closer at these 3D beams. Business will go to the most efficient system. Three aspects speak to increased efficiency by higher directivity. -SNIR (SE) -Beam cell Area (Area SE) -Mapping resolution (tower integration) AST has highest directivity

All users within a cell gets to share the bandwidth. Small cells =good. High Signal to interference ratio (as in intereference from your own sidelobes) = higher modulation= more throughput per unit of bandwidth. Smaller sidelobes and cells also means you can map better to towers

I’ll end with this from my 📌 tweet. A Lynk cell visualised in yellow the size of a small US state casts an interference pattern far and can not map to towers. An AST beam, in red, is the size of a large tower cell. It stays earth fixed and does not cast wide interference.

yes