Metraipolis › archive › dc_49fe43beaa0e

ImpactHO: Importance-Aware KV Cache Transfer for Multi-User Edge LLM Handover — arXiv:2608.10545

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.10545
License (as archived)
CC0-1.0
Archived
2026-10-19T06:09:20.423Z
SHA-256 of the archived text
81facb49ba9320693e6a70d06a8d249a0c245c04dca30da478c78650eb32d6ad
Stable URL of this page
https://metraipolis.com/documento/dc_49fe43beaa0e

How to cite

ImpactHO: Importance-Aware KV Cache Transfer for Multi-User Edge LLM Handover — arXiv:2608.10545 — arXiv (en), https://arxiv.org/abs/2608.10545 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_49fe43beaa0e (Metraipolis archive dc_49fe43beaa0e, sha256 81facb49ba9320693e6a70d06a8d249a0c245c04dca30da478c78650eb32d6ad)
BibTeX
@misc{dc49fe43beaa0e,
  title = {ImpactHO: Importance-Aware KV Cache Transfer for Multi-User Edge LLM Handover — arXiv:2608.10545},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.10545},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_49fe43beaa0e; sha256 81facb49ba9320693e6a70d06a8d249a0c245c04dca30da478c78650eb32d6ad},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.10545. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Edge LLMs must preserve inference continuity when a user hands over between edge nodes, requiring key-value (KV) cache transfer to the target node. However, simultaneous handovers saturate the backhaul, preventing full cache delivery within the mobility-imposed transfer window. Rather than allocating bandwidth as if all cache entries were equally valuable, we order each user's KV cache by importance and transmit only its most informative fraction, turning token-level sparsity into communication savings. We cast the transfer as a multi-user backhaul allocation problem that maximizes average accuracy across users. Each user's partial-cache accuracy serves as its utility: a sigmoid that fits measurements on the RULER benchmark with $R^2>0.99$ across models and context lengths. Because importance ordering front-loads the high-value entries, the concave region of the accuracy curve spans nearly the entire cache. Our proposed allocator keeps served users within this region, making each per-slot allocation problem convex. The optimum is derived via a closed-form weighted water-filling solution that generalizes information-theoretic water-filling and enables online scheduling. The proposed allocator attains over 93.7% average accuracy in a 500ms transfer window, within 0.5pp of the full-cache ceiling, and reaches 98.2-99.5% of a clairvoyant upper bound.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-12 in cs.AI).
Categorie arXiv: Networking and Internet Architecture (cs.NI) ; Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC).
Pagina del lavoro: https://arxiv.org/abs/2608.10545 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T05:57:28.383Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-12?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).