Communication-native inference over complex topologies

Determine how wireless superposition and broadcast properties can be exploited for efficient embedding and model aggregation over complex topological structures beyond simple graphs.

Background

Wireless communication can perform many-to-one and one-to-many aggregation through superposition and broadcast. Initial work has explored these capabilities, but the survey states that their use with complex topological structures remains unresolved.

References

While initial results exist in~\citep{fiorellino2024topological}, further research is still needed to fully exploit these network environment properties through complex topological structures beyond simple graphs.

— From Euclidean to Graph-Structured Data: A Survey of Collaborative Learning  (2609.02984 - Bourgerie et al., 2 Sep 2026) in Section 6.3, paragraph “Open challenges”