Starlink Performance Achieves Improved Vehicular Mobility with Dynamic Beam Switching

Researchers are now intensely focused on understanding the real-world performance of the Starlink satellite network, but a critical gap remains in knowledge regarding how it functions under the challenging conditions of vehicular movement! Jinwei Zhao from the University of Victoria, alongside Jack Baude and Ali Ahangarpour, also of the University of Victoria, with Vaibhava Krishna Devulapalli, Sree Ganesh Lalitaditya Divakarla, and Zhi-Li Zhang et al from the University of Minnesota, Twin Cities, have investigated dynamic beam switching in response to signal degradation during mobility! Their work reveals that Starlink user terminals actively attempt multiple beam switches when experiencing connection issues , caused by obstructions or poor signal , extending handover times beyond previously understood limits! This novel identification method, utilising diagnostic data, crucially demystifies performance degradations in mobile environments, paving the way for optimising transport protocols and enhancing connectivity for a wide range of applications.

Dynamic Beam Switching in Moving Starlink Networks requires

Scientists have recently unveiled a novel method for identifying communicating satellites for Starlink user terminals (UTs) in motion, addressing a critical gap in understanding network performance under realistic conditions. The research team achieved a breakthrough in detecting dynamic beam switching events, a capability crucial for maintaining connectivity when UTs experience signal degradation due to obstructions or mobility. Existing Starlink satellite identification techniques function only in stationary, unobstructed scenarios, failing to account for the complexities introduced by vehicular movement and environmental interference. This study reveals that UTs actively attempt multiple dynamic beam switches when a satellite link weakens, potentially exceeding the standard 15-second handover interval, a phenomenon previously unobserved and unexplained.

The team’s approach leverages diagnostic data from the UTs alongside connected satellite information to plausibly explain network performance degradations, demystifying issues encountered in mobile settings. Experiments show that UTs can initiate several dynamic beam switching attempts within a single 15-second timeslot to overcome obstructions or poor signal-to-noise ratios (SNR). Observations on stationary, obstructed UTs indicate that even with dynamic beam switching, multiple failed attempts can occur, leading to prolonged network interruptions, while proactive switching significantly reduces obstruction time. The research establishes that mobile UTs, due to their increased exposure to transient obstructions, experience more frequent dynamic beam switching events and network interruptions, with SKY_SEARCH events doubling and accounting for up to 45.18% of total outage time compared to stationary, heavily obstructed units.

This work opens new avenues for enhancing the end-to-end performance of transport layer protocols and diverse application scenarios reliant on Starlink connectivity. By accurately identifying communicating satellites and understanding the dynamics of beam switching, the researchers provide crucial insights into optimising network performance in challenging mobile environments. The proposed mobility-aware identification method supports both stationary and mobile scenarios, operating effectively across varying obstruction conditions, and represents a significant advancement over existing techniques. Furthermore, the study’s findings are particularly relevant given the growing deployment of Starlink UTs on vehicles, recreational vehicles, trains, marine vessels, and airplanes, where maintaining a stable connection amidst constant movement and potential obstructions is paramount.

As of December 2025, SpaceX’s Starlink boasts the largest low-earth-orbit (LEO) satellite constellation, comprising over 9,000 operational satellites, bringing high-speed, low-latency internet to previously underserved areas. The research team addressed the limitations of current studies, which primarily focus on stationary UTs and regular 15-second handovers, by investigating the impact of frequent UT orientation changes and transient obstructions, such as bridges, highway signs, and dense foliage, on signal quality and network performance. The study answers key questions regarding the impact of dynamic beam switching on latency and throughput, the UT’s satellite selection strategy under deteriorating SNR, and the influence of varying obstruction conditions on network performance, ultimately contributing to a more robust and reliable mobile Starlink experience. Dynamic.

Experiments meticulously tracked UT behaviour during vehicular movement, identifying instances where the terminal switched between satellites even within a single 15-second timeslot, a phenomenon previously unobserved in stationary, unobstructed scenarios. The team measured the impact of transient obstructions on network performance, observing that a single obstruction event at approximately 12:39.380 lasted for 1,280,066,127 nanoseconds, equivalent to 1.28 seconds! Data shows that despite the UT conducting dynamic beam switching, a regular handover to STARLINK-4809 occurred at 12:42.338, highlighting the complex interplay between dynamic and scheduled handovers. Further analysis revealed another dynamic beam switch to STARLINK-5931 at around 12:44.833, triggered by a subsequent obstruction.

Synchronising ICMP ping latency tests with these events confirmed a direct correlation between obstructions, beam switching, and network performance degradation. Researchers developed a mobility-aware Starlink identification method capable of detecting these dynamic beam switching events using UT diagnostic data and connected information. The algorithm processes obstruction maps, converting pixel coordinates to celestial coordinates to identify communicating satellites. By applying connected component labeling (CCL) to binarized obstruction maps, the team detected spatially distinct line segments indicative of beam switching, with the number of line segments increasing during periods of high mobility.

This allowed for the identification of multiple communicating satellites within the same 15-second timeslot, a feat impossible with existing algorithms. Measurements confirm that the proposed algorithm accurately identifies communicating satellites even amidst beam switching, overcoming the limitations of algorithms that incorrectly utilise pixel coordinates from different satellite trajectories. In one instance, the team identified STARLINK-31253 before the conclusion of an initial outage event, thanks to the presence of obstructed pixels, while STARLINK-5931 was identified only after the outage ended due to gRPC interface limitations. The breakthrough delivers a crucial understanding of mobile Starlink network behaviour, paving the way for enhanced transport layer protocols and improved performance in diverse application scenarios.

Dynamic Beam Switching impacts Mobile Starlink performance significantly

Scientists have developed a new method for identifying communicating satellites for Starlink user terminals (UTs) while in motion! This research reveals that Starlink UTs actively attempt multiple dynamic beam switchings to maintain connectivity when experiencing degraded links due to obstructions or poor signal quality, extending handover events beyond the standard 15-second interval! Researchers demonstrated that these beam switching attempts, alongside regular handovers, significantly impact network performance, often resulting in prolonged connection outages and latency spikes. The findings clarify the causes of performance degradation in mobile Starlink networks, which is vital for optimising transport layer protocols and diverse applications.

The authors acknowledge a limitation in that their analysis relies on diagnostic data from the UT, potentially introducing biases related to device-specific behaviours! Future work should focus on developing realistic LEO constellation simulations incorporating both satellite identification results and network performance traces to create more accurate, mobility-aware scenarios. This work establishes a crucial step towards a deeper understanding of Starlink network behaviour under vehicular mobility! By identifying communicating satellites, this research opens avenues for further investigation into end-to-end performance and satellite selection strategies, ultimately contributing to enhanced network efficiency and reliability.