DetSpace: Distributed congestion free routing for global deterministic network in mega-constellations

Emerging long-range industrial IoT applications (e.g., remote patient monitoring) have increasingly higher requirements for global deterministic delay. Although many existing methods have built deterministic networks in small-scale networks through centralized computing and resource reservation, they cannot be applied on a global scale. The emerging mega-constellations enable new opportunities for realizing deterministic delay globally. As one constellation (e.g., Starlink) might be managed by a single operator (e.g., SpaceX), packets can be routed within deterministic number of hops. Moreover, the path diversity brought by the highly symmetrical network structure in mega-constellations can help to construct a congestion free network by routing. This paper leverages these unique characteristics of mega-constellations to avoid the traditional network congestion caused by multiple inputs and single output, and to determine the routing hops, and thus realizing a global deterministic network (DetSpace). The model based on the 2D Markov chain theoretically verifies the correctness of DetSpace. The effectiveness of DetSpace in different traffic load conditions is also verified by extensive simulations.