Wide-Area Signalling Enabled by the High-Throughput-Satellite Communication: A Machine Learning Approach

The robots are considered to perform search, rescue and environment mothering in the ocean, desert and rainforest, where no terrestrial networks are available. To enable such scenarios, extensive signalling and information exchange are envisioned to be supported by satellite communication networks. To provide both massive access and high throughput, code-domain non-orthogonal multiple access (NOMA) is expected to be exploited by satellite communication to reduce inter-robots interference. However, the code alAlocation schemes are of high complexity, which limits its usability in satellite communication. To this end, we propose a machine learning-enabled satellite communication scheme using code-domain NOMA. Extensive simulation results demonstrate the superiority of the proposed scheme in terms of both the network capacity and fairness between robots.