Networking for a New Era of Global Satellite Connectivity

Large constellations of satellites in low Earth orbit (LEO) allow for unparalleled global coverage but require new networking approaches.
A schematic of the Earth with many satellite links.
Data from a ship are routed along an efficient path through inter-satellite links to a ground site. After a satellite along this path fails, data are rerouted along a longer path until a new link is added to provide a shorter, more optimal path.

Satellite constellations containing hundreds to thousands of satellites orbiting the Earth at altitudes of less than 2,000 kilometers will provide global data distribution services of unprecedented scale and reach. Commercial examples of such proliferated LEO (pLEO) satellite networks — some currently under design and others already in the process of being deployed — are SpaceX's Starlink, Amazon's Kuiper, Eutelsat's OneWeb, and Telesat's Lightspeed. Government-owned pLEO systems are also in the works, including the Space Development Agency's Proliferated Warfighter Space Architecture and SpaceX's Starshield for the U.S. Space Force.

Emerging pLEO networks offer several benefits, including low-latency data transmission; global coverage, even in remote or inaccessible regions; and rapid replenishment and upgrading of satellites. However, they present new technical challenges, especially in network architecture, design, and protocols. Over the last few years, Lincoln Laboratory has been exploring solutions to challenges in several areas:

  • Space network topology design: Network topology is the pattern of satellite interconnections — which satellites connect to which other satellites, how many neighbors each satellite connects to, and how frequently these links are repointed or reconfigured. Most emerging pLEO constellations join satellites via inter-satellite links (ISLs), typically using laser communications technology. Space networks have constraints that most terrestrial networks do not; each satellite can carry only a small number of ISL terminals because of size, weight, and power restrictions, and the range of ISLs is limited. Our team is investigating topology designs for several different constellations, each with hundreds of satellites; we have examined network performance characteristics such as resilience to failures, time delays from data sources to destinations, and overall system data capacity.
  • Constellation resilience: Resilience is the ability of a system architecture to continue providing required capabilities during system failures, environmental effects, or adversary actions. To increase the resilience of pLEO network designs for U.S. government systems, we are evaluating techniques such as adding more ISL connections and dynamically reconfiguring ISL connections after failures.
  • Space network routing: Routing is the process of selecting the best path for data to travel through a network. In terrestrial networks like the Internet, significant processing power is typically required to calculate optimum routes. Satellites can theoretically do these calculations, but their processing power is very limited compared to that of most terrestrial nodes. Routes can instead be calculated on the ground and uploaded to each satellite, but a backup scheme is needed in case the ground-based route calculation engine fails or gets disconnected from any satellite. Our team is studying how simple backup routing methods can be used if terrestrial routing fails, and how primary and backup routing methods can work together as satellites move through their orbits and failure scenarios change.