Attached is a recent cloudflare test of verizon, taken about 1.2 miles from a rural tower. While the bandwidth was low for reasons unknown, the latency was... quite tolerable. Perhaps they are figuring this out.

I have no idea how atlas calculates latency or loss. Over what interval? In the starlink case especially, I look at a typical plot, and I hope that if more people looked at details like that, they would despair of using averages ever again.

On Wed, Sep 20, 2023 at 8:05 AM Livingood, Jason via Starlink <starlink@lists.bufferbloat.net> wrote:

As some of you may recall I recruiting some US-based Starlink users to host RIPE Atlas probes as part of my role as a RIPE Atlas Ambassador. I just got word that NetForecast conducted a performance study that looked at RIPE Atlas data using some of the data from those and other probes. They looked at homes across the US with 5G fixed wireless, LEO satellite, and DOCSIS.

They have posted a first report of this study at https://www.netforecast.com/wp-content/uploads/FixedWireless_LEO_CableComparisonReport_NFR5148-1.pdf. I am looking forward to other researchers looking into the RIPE Atlas data – I know a few are working on this.

Jason

_______________________________________________
Starlink mailing list
Starlink@lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/starlink