Syntropy & Starlink: Testing Results

Our team has completed the first round of testing, verifying several ways Syntropy and Starlink can work together.

Our tech team has been testing Syntropy technology with Starlink’s low-earth orbit satellites. We wanted to answer a simple question: could Syntropy’s smart-routing technology reduce latency and improve performance using Starlink’s network?

To answer this question, we needed to run thousands of tests, cross-checking results across multiple variables, including time of day, route path, intermediary provider, and ground station. While these results are still preliminary, we are excited to share that we verified several ways Syntropy and Starlink could work together to improve global internet access and performance.

Syntropy’s use case for Starlink

Syntropy adds smart-routing capabilities to your data, reducing latency, avoiding outages, and lowering packet loss. Our data suggest that up to 70% of internet routes can be improved using our drop-in software solution.

Smart-routing is like having your internet connection ride on Google Maps versus paper maps. Using Syntropy, your data can avoid congestion, outages, and even censorship by rerouting in real-time.

This smart-routing system is backed by our Distributed Autonomous Routing Protocol (DARP) technology. DARP nodes constantly share latency information, creating a global intelligence layer for internet pathways. DARP nodes relay network traffic when they provide a superior route versus the default public internet path, earning NOIA tokens that facilitate the value-based economy.

Our test results verified the hypothesis of whether Starlink nodes could benefit from our smart-routing network. With DARP, Starlink nodes could enhance the rest of the network by providing pathways with improved latency, reduced jitter, and minimized packet loss.

Testing approach

We connected the Syntropy Windows App to a Starlink-enabled device and routed traffic through hundreds of servers across dozens of data centers. We began testing only a slice of the available options before scaling operations further.

The preliminary results were encouraging enough to ramp testing further, and our team continues to conduct even more testing to complement these results with additional real-world data.

Testing results

The combination of Syntropy and Starlink is promising. As you will see from the testing data below, there are thousands of opportunities for our technology to improve performance on the Starlink network. And because our technology pivots to the public internet when no better pathways are available, Starlink users are always ensured public internet speeds or better.

Our first tests focused on long-haul routes. We wanted to quickly check the performance of connections when traffic travels across the globe.

Our first long-haul test delivered very encouraging latency improvements.

The results were impressive. The example above shows the latency metrics from a Starlink connection on the U.S. west coast to Indonesia. Our network identified multiple better-performing relays, with the most optimal being located in Singapore. Optimizations consistently reached 10ms to 20ms, with some spikes up to 30ms. In the cases where Syntropy couldn’t optimize the connection, Starlink users would get their default internet speeds.

During testing, we repeated these results across thousands of other long haul routes. We selected a small sample below representing a range of providers and locations.

Note the immediate optimization opportunity in the Seattle->Seattle->Lansing route.

It’s not surprising to see Syntropy technology improve latency across long haul routes, as we’ve heavily validated this use case in the past. Long haul routes typically have the longest distance to travel and the most “hops” until the end destination, resulting in increased optimization opportunities.

The ability to produce short haul optimizations is more impressive. Yet again, our testing showed that Syntropy could optimize Starlink’s network even when data travels within the same geographical region.

Optimization opportunities were promising, both within the same country and on international routes.

Even more impressive, our technology improved the performance of connections where the relay point was in close proximity to the destination. Below are a few examples of this in three major cities around the world.

Thousands of similar optimizations were discovered on both domestic and international routes.

Our testing also identified times in which Syntropy may prevent outages while using Starlink. The example below shows how part of the Starlink network (or at least a segment of the total route path) experienced intermittent connection downtime. Using a Syntropy relay, Starlink could have avoided this downtime.

Testing data suggests Syntropy and DARP can help Starlink users avoid intermittent outages.

The examples above only scratch the surface of our testing efforts. Digging deeper, we wanted to see whether Syntropy could stabilize Starlink connections even if the optimization opportunities weren’t abundant.

Connection stability is provided even when latency reductions are minimal.

For some destinations, latency improvements using Syntropy were pretty minimal. However, our network would be able to stabilize latency considerably. Our team conducted additional tests by adding new relays in Japan, Indonesia, and more — all with similar results. The takeaway is that Syntropy can both optimize and stabilize Starlink connections.

Starlink jitter (fluctuations in latency) can be reduced with Syntropy, stabilizing connections on top of raw latency improvements.

One challenge we encountered during testing was Starlink’s volatile connection latency, also known as high jitter. Starlink connections appear to have significantly more jitter than traditional internet connections. To highlight this, compare the charts below, which visualize jitter for a Starlink ground station (left) and a Starlink-connected device (right).

As expected, jitter was higher for end users than when testing directly in the nearest Starlink ground station.

To mitigate the impact of jitter on test results, we identified the exact ground station used by our original Starlink node and placed our node directly inside this ground station. These more granular tests confirmed our initial results.

While our complete test results span thousands of routes, you can click below to go deeper into a small fraction of those results. In total, DARP testing indicates that 42.6% of Starlink routes can be improved in some way. The data below details 149 unique relays. Given a well-distributed network topology, roughly 40% of DARP nodes could expect to earn tokens from cases like Starlink once our SDN is integrated with our blockchain launch.

Bottom line

After researching and testing Starlink across thousands of paths, the results were more promising than expected. Latency, jitter, and packet loss can all be improved by combining Syntropy and Starlink. Our community-run DARP network can consistently provide performance improvements. At the same time, our built-in end-to-end encryption can give every Starlink user increased privacy and peace of mind by default, with no performance degradation.

With an increasing amount of validating data, this could become a solid real-world use case for our technology. We are pursuing this opportunity aggressively, and will continue to generate more testing data, updating you as this opportunity advances.