arXiv:2505.10946v3 Announce Type: replace-cross Abstract: Token communications (TokenCom) is an emerging generative semantic communication paradigm, where tokens serve as compact representation units across modalities. Their contextual dependencies can be exploited by pretrained large models for semantic recovery. In this paper, we propose token-domain multiple access (ToDMA), a large-model-driven semantic multiple access scheme for massive token communications. ToDMA integrates unsourced random access with context-aware token processing. It enables massive uncoordinated devices to transmit to
The proliferation of generative AI models and the increasing demand for efficient data communication for AI applications are driving innovations in semantic communication paradigms.
This development proposes a new method for massive, uncoordinated device communication crucial for next-generation IoT and AI-driven networks, potentially enhancing data efficiency and network capacity significantly.
The proposed ToDMA scheme redefines how devices communicate by leveraging large models for semantic rather than purely syntactic data transfer, moving beyond traditional communication bottlenecks.
- · Telecommunications infrastructure providers
- · AI model developers
- · IoT device manufacturers
- · Semiconductor companies
- · Legacy communication protocol developers (if slow to adapt)
- · Cloud providers reliant on traditional bandwidth scaling
- · Companies with inefficient data transfer needs
Implementation of ToDMA could lead to a substantial increase in the number of connected devices and the volume of AI-driven data exchange.
This efficiency gain might accelerate the deployment of decentralized AI and edge computing solutions, reducing reliance on central cloud infrastructure for certain tasks.
The enhanced communication capability could enable new categories of AI applications requiring real-time, massive sensory data integration, impacting fields like autonomous systems and smart cities.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG