Drift-Plus-Penalty Based Queue Management for Edge LLM Inference with Repeated Sampling.

Hongtao Li, Fengxian Guo, Ruihong Jiang, Mugen Peng

Browse the full WCNC paper archive.