Kyunghan Lee
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Kyunghan Lee has a total of 87 co-authors. Here are the top co-authors by number of shared publications:
- Sangtae Ha
University of Colorado Boulder, Boulder, USA
10 shared publications - Chanjeong Park
Seoul National University, Seoul, Republic of Korea
5 shared publications - Gibum Park
Seoul National University, Seoul, Republic of Korea
5 shared publications - Wooseung Nam
Department of Electrical and Computer Engineering, Seoul National University, Seoul, South Korea
5 shared publications - Donggyu Yang
Seoul National University, Seoul, Republic of Korea
4 shared publications - Jinsung Lee
Qualcomm Technologies Inc., San Jose, CA, USA
4 shared publications
Latest Publications
- Open Access
- 19 November 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Networking
- Vol. 4 (CoNEXT4) , 1-24
Unlocking Opportunistic Transmission in Datacenter Flow Control
As datacenter link bandwidth grows, hop-by-hop flow control (FC) has become essential for rapid congestion suppression by propagating backpressure to pause upstream transmission. However, since this backpressure applies at a per-queue granularity, FC suffers from Head-of-Line (HoL) blocking, where uncongested traffic is throttled unnecessarily when sharing queues with congested flows. The state-of-the-art FC schemes mitigate this by isolating congested flows into separate queues, yet such isolation is fundamentally limited by queue scarcity in commodity switches. This paper rethinks how FC reacts to congestion. Our key observation is that paused queues exhibit highly heterogeneous congestion levels, with many of them still carrying substantial uncongested traffic. Building on this insight, we present G ambit , an FC augmentation that enables opportunistic transmission under HoL blocking. By selectively relaxing pause on less congested queues based on per-queue congestion levels, G ambit allows uncongested flows to make progress by exploiting unused bandwidth, while bounding congestion risks. We verify the hardware feasibility of \gambit by prototyping it on the Intel Tofino switch with P4, and further conduct large-scale trace-driven simulations with production workloads. Our evaluation shows that G ambit effectively recovers the performance of FC lost due to practical queue constraints, reducing average FCT slowdown by up to 49.7% and 33.1%, and short-flow 99th-percentile slowdown by up to 50.6% and 55.6%, for BFC and Pyrrha, respectively.
- 19 November 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Networking
- Vol. 4 (CoNEXT4) , 1-24
Open AccessUnlocking Opportunistic Transmission in Datacenter Flow Control
As datacenter link bandwidth grows, hop-by-hop flow control (FC) has become essential for rapid congestion suppression by propagating backpressure to pause upstream transmission. However, since this backpressure applies at a per-queue granularity, FC suffers from Head-of-Line (HoL) blocking, where uncongested traffic is throttled unnecessarily when sharing queues with congested flows. The state-of-the-art FC schemes mitigate this by isolating congested flows into separate queues, yet such isolation is fundamentally limited by queue scarcity in commodity switches. This paper rethinks how FC reacts to congestion. Our key observation is that paused queues exhibit highly heterogeneous congestion levels, with many of them still carrying substantial uncongested traffic. Building on this insight, we present G ambit , an FC augmentation that enables opportunistic transmission under HoL blocking. By selectively relaxing pause on less congested queues based on per-queue congestion levels, G ambit allows uncongested flows to make progress by exploiting unused bandwidth, while bounding congestion risks. We verify the hardware feasibility of \gambit by prototyping it on the Intel Tofino switch with P4, and further conduct large-scale trace-driven simulations with production workloads. Our evaluation shows that G ambit effectively recovers the performance of FC lost due to practical queue constraints, reducing average FCT slowdown by up to 49.7% and 33.1%, and short-flow 99th-percentile slowdown by up to 50.6% and 55.6%, for BFC and Pyrrha, respectively.
- Open Access
- 26 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 282-296
TACToC: Enabling Task-Adaptive Communication System with a Task-Describable Encoder
- 26 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 282-296
Open AccessTACToC: Enabling Task-Adaptive Communication System with a Task-Describable Encoder
- Open Access
- 26 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 1801-1803
Demo: Responsive On-Device LLM Inference under Multitasking
- 26 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 1801-1803
Open AccessDemo: Responsive On-Device LLM Inference under Multitasking
- Open Access
- 26 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 1494-1508
mzCache: On-Device LLM Memory Management under Multitasking
- 26 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 1494-1508
Open AccessmzCache: On-Device LLM Memory Management under Multitasking
- 7 October 2026
- Published by Elsevier in Neurocomputing
- Vol. 709
BVSR: BitVector-sum representation for distribution-aware LLM weight quantization
- 7 October 2026
- Published by Elsevier in Neurocomputing
- Vol. 709
BVSR: BitVector-sum representation for distribution-aware LLM weight quantization
- Open Access
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1315-1329
DeepSFU: Scalable Deepfake Detection for Video Conferencing
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1315-1329
Open AccessDeepSFU: Scalable Deepfake Detection for Video Conferencing
- Open Access
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 13-14
Poster: Accelerating Offloaded ViT Inference via Progressive Refinement
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 13-14
Open AccessPoster: Accelerating Offloaded ViT Inference via Progressive Refinement
- Open Access
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 405-417
Ouroboros: Instilling Motion Awareness in ViTs for Efficient Video Analytics on the Edge
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 405-417
Open AccessOuroboros: Instilling Motion Awareness in ViTs for Efficient Video Analytics on the Edge
- Open Access
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 580-592
SAIL: Redesigning Collaborative Language Inference with a Single Server-to-Mobile Handoff
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 580-592
Open AccessSAIL: Redesigning Collaborative Language Inference with a Single Server-to-Mobile Handoff
- Open Access
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 23-24
Poster: Selective Semantic Cache Transfer for Mobile-Cloud LLM Inference
Direct key-value (KV) cache transfer between large language models (LLMs) is emerging as a richer alternative to token exchange, but the KV cache is orders of magnitude larger than tokens, and mobile-cloud inference systems are not yet prepared for this shift. We present CAIL, a system that reframes semantic cache delivery as a selection problem: rather than compressing or scheduling the entire cache, CAIL identifies which cache positions genuinely help the edge small language model (SLM) using an uncertainty-aware selection policy. Building on recent advances in cross-model cache projection (C2C) and server-assisted edge inference (SAIL), we ask which positions to send rather than how to compress them. Preliminary experiments on the WikiText-103 dataset show that selective transfer at 25% of positions outperforms full-cache transfer in perplexity, closing 28.7% of the SLM-to-LLM gap.
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 23-24
Open AccessPoster: Selective Semantic Cache Transfer for Mobile-Cloud LLM Inference
Direct key-value (KV) cache transfer between large language models (LLMs) is emerging as a richer alternative to token exchange, but the KV cache is orders of magnitude larger than tokens, and mobile-cloud inference systems are not yet prepared for this shift. We present CAIL, a system that reframes semantic cache delivery as a selection problem: rather than compressing or scheduling the entire cache, CAIL identifies which cache positions genuinely help the edge small language model (SLM) using an uncertainty-aware selection policy. Building on recent advances in cross-model cache projection (C2C) and server-assisted edge inference (SAIL), we ask which positions to send rather than how to compress them. Preliminary experiments on the WikiText-103 dataset show that selective transfer at 25% of positions outperforms full-cache transfer in perplexity, closing 28.7% of the SLM-to-LLM gap.