C🅰️tSE Profile picture
Dec 25, 2022 12 tweets 6 min read Read on X
$ASTS satellite #BlueWalker3 in orbit testing is ongoing.

It is the cutting edge of 5g Non-Terrestrial-Network NTN, Direct-To-Handset, DTH, technology.

It has a lot of interesting design choices.

Selected, in tech, standard and regulatory terms as it was purpose built.

1/n
Previous schematic of the transparent satellite AST uses is one of two NTN DTH architectures supported by the global 3GPP 5g standard.

The alternative AST opted out of has parts, or all, of the base-station placed onboard the satellite, whereas ASTs is on earth.

2/n
Starlink DTH is an add-on functionality to an existing architecture built for another purpose. It uses the other architecture.

This has several implications.

For handoff and multi connectivity.

But importantly ..

3/n
Having the eNodeB or base-station on the ground like AST does allows access to it and it allows use of COTS cutting edge hardware like Nokia single-RAN Airscale.

AST has been working with Nokia Bell Labs for two years to integrate assuring best possible technology.
4/n
Satellite communications payload is another choice where AST stands out. It is typical GEO (high altitude) satellite specs, but placed in LEO (low altitude).

A bit like putting an ocean-liner in a pond it might seem.

5/n
That allows for more narrow beams using less power, yet creating high sihnal strengths.

The win is three fold. Good mapping/resolution. High spectrum reuse and high modulation.

Resulting in good coverage and high throughput.

6/n
Low Earth Orbit has these benefits. Low RTT latency is one. Essential to work with 5g.

But also problems. Latency is still higher than terrestrial towers. Doppler shift needs to be handled and Farraday rotation of the polarization.

These have tweaks that can be tuned.

7/n
But tuning takes time.

I will quote @steve_larrison here.

”Testing is a process, not an event.”

Initial results < final results.

But there is more to architecture.

AST BlueBird servicelinks are 20 degrees above horizon.

8/n
That is another extreme design choice made possible by the huge 20 meter square array. As you need all of it to make a so narrow and strong beam.

The BlueWalker3 test satellite will generate such smsll cells 30-35 degrees above horizon as its phased array is smaller.

9/n
$AST uses fixed cell sites and steerable beams. Beams that are made more narrow and elliptical near the edge of the satellite firld of view in order to cast circular beamcells of uniform size onto the surface of the earth. It does not use moving cells.

10/n
The backhaul, or feeder link, field of view is even wider. It works down to 10 degrees above horizon as its antenna is a steerable 70 cm dish connecting to huge steerable dishes at the gateways.

11/n
There is much more, to tell, about the design choices made.

One is that the satellite is FPGA, or software defined. It means they can patch and tweak a lot during in orbit tests.

I will just summarize, for now, that the choices made are very wise as you scrutinize them.
12/12

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with C🅰️tSE

C🅰️tSE Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @CatSE___ApeX___

Jan 23
In Latin, the concept of drawing an inference from what is not shown or mentioned is most commonly expressed as:

Argumentum ex silentio

$ASTS

🧠🐈‍⬛

1/n Image
When the last BlueBirds deployed there was two patterns.

One was a spike in drag and the level of drag settling on a higher level afterwards.

The other and first sign was a delay of TLE data longer than usual.

In a way the first -inconclusive- sign was the absence of data.
/2
As back then TLEs was delayed more than 2 full days. Which is /was unusual.

/3 Image
Read 10 tweets
Jan 8
Deploying a 200 sqm array within two weeks of launch is a high-stakes move. Historically, large structures wait longer for "checkout."

Is the upcoming FM2 launch giving us a faster deployment?

There is now FM1 data.

1/ Image
Here is drag data.

/2 Image
Image
Here is the relative change of drag (schematically) that you’d expect from detumbling, unfolding and the reorienting back to edge on flight.

3/ Image
Read 9 tweets
Jan 6
I was asked to comment on Space-x last Ex parte letter that they have filed to the FCC.

So here is the picture:

$ASTS has asked authority to launch full constellation beyond 25 satellites.

Space-x wants to delay and complicate that.

They keep filing all the way to sunshine
1/ Image
It’s extremely uncompetitive behaviour and a bit immoral as what Space-X has begged be implemented onto AST is the same type of regulations they see as an obstacle when applied to themselves.

What they ask that AST shall not be allowed to is what they themselves do.
Golden rule? Image
It’s important to grasp that the next 20 satellites and the Block1s are approved already.

So this pen-fighting is about about satellites to launched beyond Q1 2026.

/3 Image
Image
Image
Image
Read 20 tweets
Jan 2
_One way to increase Area spectral efficiency is lowering constellation altitude.

That way comes at two costs: The number of 🛰️satellites required on orbit increases and their orbital dwell ⏳time decreases both affecting the replenish rate 🛠️adversely as:

🛠️ = 🛰️/⏳

🧶🐈‍⬛

1/n Image
Let’s do a SpaceMob thing and look at this from first principles.

A satellite has a field of view. FoV.

That Field of View projects a footprint on earth.

The footprint increases with angle of the field of view and increases with altitude.

2/ Image
Image
Within this field of view the satellite creates beams.

They also have an angle called beamwidth and a footprint called beamcell.

There are many of these beams and beamcells within the satellite footprint.

3/ Image
Image
Read 31 tweets
Nov 25, 2025
🚨 $ASTS IS INCREASING ITS PATENT MOAT ESP. ON DOW ORBITS 🚨THIS PATENT COVERS THE SIGNAL-PROCESSING METHODS — SELECTION COMBINING, DIVERSITY COMBINING, AND MIMO — THAT ENABLE RELIABLE DIRECT-TO-CELL CONNECTIVITY FROM LEO SATELLITES TO STANDARD HANDSETS.

1/ Image
THE PROBLEM ADDRESSED: END USER DEVICES MAY RECEIVE MULTIPLE SATELLITE SIGNALS (MULTIPLE PATHS, SUB-ARRAYS, OR MULTI-SATELLITE LINKS) WITH DIFFERENT DELAYS, DOPPLER, AND SNRS.

THE PATENT SPECIFIES HOW TO CHOOSE AND COMBINE THESE SIGNALS TO MAXIMIZE LINK RELIABILITY.

2/
SELECTION COMBINING: the system monitors multiple receive branches and selects the best branch using quality metrics (e.g., SNR/BER estimates, channel quality indicators). selection lowers complexity and power for the handset when one branch dominates

3/ Image
Image
Read 13 tweets
Aug 11, 2025
Quarterly report thread. $ASTS

Excellent initial impression.

August delivery of FM1 !

5 launches before Q1 -26

1.5Bn+ at hand

🧶🐈‍⬛

1/

stocktitan.net/news/ASTS/ast-…
Signed two additional early-stage contracts for the U.S. Government end customer, bringing the total to eight contracts to date with the U.S. Government as an end customer.

This is huge.

The rate at which company adds DoD contracts is staggering. It’s not in analyst models.

2/
Service Rollout: Nationwide intermittent service in the US by end-2025, followed by UK, Japan, and Canada in Q1 2026. Expected revenue: $50-75M in H2 2025 from government and commercial customers. Supports full voice, data, and video at up to 120 Mbps peak speeds per cell.

/3
Read 27 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(