Priorityfec prioritizes key cache data for faster edge lLM collaboration
PriorityFEC: Deadline-Aware Streaming FEC for KV-Cache Handoff in Edge Multi-Agent Systems
Networking and Internet ArchitectureArtificial Intelligence
Summary
When multiple AI agents work together near users, they share information stored in a special cache to avoid repeating work. Sending this cache over slow or unreliable networks can cause delays, making teamwork slower. The authors propose PriorityFEC, a method that sends the most important parts of this cache first and uses smart error correction to protect the data before a tight deadline. Their tests show it improves task quality, especially when time is very limited or networks are unreliable.
What this means in practice
- •For edge system engineers: Improve collaborative LLM tasks at network edges by efficiently transferring key cached states under strict deadlines and unreliable connections.
- •For distributed ai platform operators: Enhance multi-agent LLM system responsiveness by prioritizing and protecting crucial data during state handoffs across unreliable edge networks.
Authors
Han Shen, Qifu Tyler Sun, Fanyang Meng, Ye Wang, Yongsheng Liang
Abstract
Multi-agent large language model (LLM) systems are increasingly deployed at the edge to support complex tasks that require collaboration and information exchange among multiple agents. However, when task contexts are passed across agents, shared context that has already been processed by upstream agents often needs to be prefilled again, resulting in substantial redundant computation. Reusing an already prefilled key--value (KV) cache can avoid this recomputation, but requires the reusable KV state to be transferred across the network before the receiving agent can begin generation. Under strict time-to-first-token (TTFT) deadlines and unreliable edge links, delivering the complete KV cache may be infeasible, making it necessary to prioritize and protect the most valuable KV information within a limited transmission budget. This paper proposes PriorityFEC, a utility-aware KV-cache handoff scheme that jointly exploits model-side KV importance and transport-side reliability. PriorityFEC first constructs a utility-aware source stream by placing high-utility KV blocks at earlier transmission positions. It then protects the ordered stream using a feedback-adaptive Streaming Forward Erasure Correction (FEC) mechanism, which schedules repair packets according to receiver feedback before the TTFT deadline. We implement PriorityFEC on a distributed edge LLM serving prototype and evaluate it with Qwen2.5-7B-Instruct and Llama-3.1-8B-Instruct across multiple datasets. Results show that PriorityFEC achieves superior task quality over the compared transport schemes, with its advantages becoming particularly pronounced under stringent deadline constraints. These results demonstrate the importance of jointly optimizing KV-cache utility and deadline-aware transport reliability when reusable LLM state must be handed off across unreliable edge networks.