Starlink launches constellation, improves comms latency

After launch of 23 craft for growing satellite constellation, Elon Musk-led US spacecraft manufacturer reveals steps taken to improve quality of experience and service

On 10 March, the world watched SpaceX’s Falcon 9 rocket take 23 Starlink communications satellites to low-Earth orbit, but perhaps what was not observed so clearly was the work the engineering teams of the satellite comms firm has been doing in improving the performance of its satellite constellation network, with the goal of delivering a service with stable 20 millisecond (ms) median latency and minimal packet loss.

The world’s first orbital-class reusable rocket, the two-stage Falcon 9 was designed and manufactured by the Elon Musk-led American spacecraft manufacturer for the reliable and safe transport of people and payloads into Earth orbit and beyond.

Reusability allows SpaceX to re-fly the most expensive parts of the rocket, which in turn is claimed to drive down the cost of space access.

The launch at 19:05 was the 11th flight for the first-stage booster supporting this mission, which previously launched Crew-5, GPS III Space Vehicle 06, Inmarsat I6-F2, CRS-28, Intelsat G-37, NG-20, and now five Starlink missions. The mission taking off at 21:09 was the 17th flight for the first-stage booster supporting the mission, which previously launched Sentinel-6 Michael Freilich, DART, Transporter-7, Iridium OneWeb, SDA-0B, and now 12 Starlink missions.

The Starlink constellations are designed to make high-speed internet available almost anywhere on Earth, but the supplier is now also focusing on quality of service in addition to footprint. In the UK, Starlink high-speed internet offers for businesses start at £96 per month, encompassing fixed sites, maritime, aviation and land mobility. The current packages offer 40-220+ Mbps download speeds, 8-25+ Mbps upload and 20-60 ms latency.

Yet Starlink now says that over the past month it has “meaningfully” reduced median and worst-case latency for users around the world.

Latency in the Starlink network is said to be driven by several factors, the principle ones being physical speed-of-light propagation from the user to the satellite and back to the ground; ground latency from the gateway sites to the internet connection point driven by ground network layout; and what Starlink called “dumb stuff” driven by non-physical limitations in its system, such as unneeded processing delays, unoptimised buffers or unnecessary packet drops that force retries.

Read more about space communications

Looking at remediations to these issues, Starlink said that while laser connectivity was essential for connecting the most remote locations on Earth and for routing around congestion in the network, it was “making strides to ensure that latency-sensitive traffic can flow over the shortest path possible”.

To address ground latency from the gateway sites to the internet connection point driven by ground network layout, Starlink has committed to adding six internet points of presence in the US this year, and optimising gateway locations and planning algorithms to ensure traffic can land as close to its destination point as possible.

While recognising that fronthaul latency is an inherent part of shared wireless systems, the company said there was significant room for optimisation, and as a result, this had been a major focus in the past several months.

In terms of the “dumb stuff’, buffers across the Starlink network have now been amended to reduce “bufferbloat”, and queueing algorithms are said to have been improved to increase capacity on gateway links from the ground to satellites.

Starlink added that since the beginning of the year, its teams have deployed and tested 193 different satellite software builds, 75 gateway software builds, 222 Starlink software builds and 57 Wi-Fi software builds. As a result, it says that already in the US alone, it has reduced median latency by more than 30%, from 48.5ms to 33ms during hours of peak usage.

Worst-case peak hour latency (p99) has dropped by over 60%, from over 150ms to less than 65ms. Outside of the US, the company says it has also reduced median latency by up to 25%, and worst-case latencies by up to 35%.

Read more on Network monitoring and analysis

Data Center
Data Management