Jingbin Zhou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jingbin Zhou has a total of 45 co-authors. Here are the top co-authors by number of shared publications:
- Kun Tan
HUAWEI, Shenzhen, China
4 shared publications - Shuihai Hu
HUAWEI, Shenzhen, China
4 shared publications - Hao Wu
Huawei Technologies Co Ltd
3 shared publications - Xinle Du
Tsinghua University, Beijing, China
3 shared publications - Jiaxin Liang
Huawei Technologies Co Ltd, Shenzhen, China
2 shared publications - Jie Li
Huawei Technologies Co Ltd, Shenzhen, China
2 shared publications
Latest Publications
- Open Access
- 5 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 130-136
Shielding Mobile Battery from Middlebox Timeouts via Server-Driven Zero-Wakeup Keep-Alives
Mobile applications rely on persistent connections for push notifications, but network middleboxes silently drop idle connections to reclaim resources. To keep these connections alive, clients transmit periodic heartbeats, which in turn trigger frequent radio wake-ups and drain battery. To address the problem, we present Strand, a server-driven zero-wakeup keep-alive architecture that crafts server-side heartbeats with a precisely calibrated Time-To-Live. These packets intentionally “strand” (expire) at the final middlebox, successfully refreshing the mapping state without traversing the energy-expensive last-hop wireless link, keeping the mobile radio completely asleep. Furthermore, to minimize probing overhead, Strand employs a cost-aware weighted Dynamic Programming (DP) algorithm that incorporates timeout distribution priors to find search boundaries and overcome the severe temporal latency penalties of the traditional exponential approach. Real-world experiments show that Strand reduces active push energy overhead by ∼ 40%. Additionally, our weighted DP-based probing algorithm accelerates timeout convergence by ∼ 28% for TCP and ∼ 26% for UDP over state-of-the-art baselines.
- 5 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 130-136
Open AccessShielding Mobile Battery from Middlebox Timeouts via Server-Driven Zero-Wakeup Keep-Alives
Mobile applications rely on persistent connections for push notifications, but network middleboxes silently drop idle connections to reclaim resources. To keep these connections alive, clients transmit periodic heartbeats, which in turn trigger frequent radio wake-ups and drain battery. To address the problem, we present Strand, a server-driven zero-wakeup keep-alive architecture that crafts server-side heartbeats with a precisely calibrated Time-To-Live. These packets intentionally “strand” (expire) at the final middlebox, successfully refreshing the mapping state without traversing the energy-expensive last-hop wireless link, keeping the mobile radio completely asleep. Furthermore, to minimize probing overhead, Strand employs a cost-aware weighted Dynamic Programming (DP) algorithm that incorporates timeout distribution priors to find search boundaries and overcome the severe temporal latency penalties of the traditional exponential approach. Real-world experiments show that Strand reduces active push energy overhead by ∼ 40%. Additionally, our weighted DP-based probing algorithm accelerates timeout convergence by ∼ 28% for TCP and ∼ 26% for UDP over state-of-the-art baselines.
- Open Access
- 29 January 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Internet of Things
- Vol. 7 (1) , 1-24
Unveiling SparkLink Low Energy: A Comparative Measurement Study
SparkLink Low Energy (SLE) is a new short-range wireless communication technology designed to satisfy the increasing demands of future short-range scenarios. While SLE promises fundamental improvements over traditional Bluetooth, a thorough real-world evaluation is lacking. This article provides a detailed evaluation framework that includes access layer benchmarking and end-to-end application validation, presenting the first comprehensive comparison between SLE and traditional Bluetooth technologies. Benchmarking reveals SLE offers a 4× increase in data rate, a 3.84× reduction in latency, and a 10 dB improvement in sensitivity. To evaluate SLE against actual application QoS requirements, we implement an end-to-end evaluation system for validation across three practical application scenarios. Based on experimentation and parameter analysis, we propose a design for QoS-centered optimization. These results underscore SLE’s fundamental advancements over Bluetooth and its potential in real-world applications, offering insights for further optimization of this communication technology.
- 29 January 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Internet of Things
- Vol. 7 (1) , 1-24
Open AccessUnveiling SparkLink Low Energy: A Comparative Measurement Study
SparkLink Low Energy (SLE) is a new short-range wireless communication technology designed to satisfy the increasing demands of future short-range scenarios. While SLE promises fundamental improvements over traditional Bluetooth, a thorough real-world evaluation is lacking. This article provides a detailed evaluation framework that includes access layer benchmarking and end-to-end application validation, presenting the first comprehensive comparison between SLE and traditional Bluetooth technologies. Benchmarking reveals SLE offers a 4× increase in data rate, a 3.84× reduction in latency, and a 10 dB improvement in sensitivity. To evaluate SLE against actual application QoS requirements, we implement an end-to-end evaluation system for validation across three practical application scenarios. Based on experimentation and parameter analysis, we propose a design for QoS-centered optimization. These results underscore SLE’s fundamental advancements over Bluetooth and its potential in real-world applications, offering insights for further optimization of this communication technology.
- Open Access
- 24 November 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Networking
- Vol. 3 (CoNEXT4) , 1-20
Harmonics: Scalable Collective Scheduling in Multi-Tenant GPU Clusters
Distributed machine learning (DML), such as large language model (LLM) training, has become one of the most critical services in multi-tenant cloud computing. However, communication contention among concurrent DML jobs significantly degrades overall GPU utilization, leading to inefficient training cluster performance. Existing approaches either achieve high performance at the cost of long scheduling runtime or reduce scheduling time at the expense of poor performance. We present Harmonics, a novel two-tier scheduling framework that strikes a balance between scheduling latency and performance. It coordinates decisions between Local Schedulers and a lightweight Global Coordinator to enable scalable and adaptive scheduling. By combining rack-level epoch-based optimization with global coordination, Harmonics alleviates communication contention and improves resource efficiency. We implement and evaluate Harmonics on real distributed ML workloads running on a GPU testbed. Compared to state-of-the-art methods such as fair sharing, optimal scheduling, Crux, and Cassini, Harmonics reduces training time by up to 33% and communication time by up to 48%. Large-scale simulations show that it reduces scheduling time by up to 91× while improving training time by 26% in large-cluster settings.
- 24 November 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Networking
- Vol. 3 (CoNEXT4) , 1-20
Open AccessHarmonics: Scalable Collective Scheduling in Multi-Tenant GPU Clusters
Distributed machine learning (DML), such as large language model (LLM) training, has become one of the most critical services in multi-tenant cloud computing. However, communication contention among concurrent DML jobs significantly degrades overall GPU utilization, leading to inefficient training cluster performance. Existing approaches either achieve high performance at the cost of long scheduling runtime or reduce scheduling time at the expense of poor performance. We present Harmonics, a novel two-tier scheduling framework that strikes a balance between scheduling latency and performance. It coordinates decisions between Local Schedulers and a lightweight Global Coordinator to enable scalable and adaptive scheduling. By combining rack-level epoch-based optimization with global coordination, Harmonics alleviates communication contention and improves resource efficiency. We implement and evaluate Harmonics on real distributed ML workloads running on a GPU testbed. Compared to state-of-the-art methods such as fair sharing, optimal scheduling, Crux, and Cassini, Harmonics reduces training time by up to 33% and communication time by up to 48%. Large-scale simulations show that it reduces scheduling time by up to 91× while improving training time by 26% in large-cluster settings.
- Open Access
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 135-141
OmniDMA: Scalable RDMA Transport over WAN
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 135-141
Open AccessOmniDMA: Scalable RDMA Transport over WAN
- Open Access
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 172-177
FlexSpark: Robust and Efficient Multi-Device Collaborative Inference over Wireless Network
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 172-177
Open AccessFlexSpark: Robust and Efficient Multi-Device Collaborative Inference over Wireless Network
- Open Access
- 3 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 88-94
Revisiting Congestion Control for WiFi Networks
WiFi networks, widely utilized by wireless devices, have become increasingly complex and congested environments, leading to noticeable delays, jitter, and throughput degradation for end-to-end network flows in today’s Internet. Through detailed experimental observations, we identified the TCP victim problem in WiFi, where TCP erroneously detects congestion and interacts with WiFi routers, resulting in a significant throughput decrease for certain hosts. In this paper, we introduce Cupid, a novel congestion control algorithm that relies on receiver-side WiFi physical layer measurements. Cupid accurately assesses congestion by measuring parameters such as airtime utilization, concurrency, and rates, allowing for precise rate adjustments. Our results demonstrate that whether employed alone or alongside other congestion controls, Cupid effectively mitigates the TCP victim problem. Furthermore, when used independently, Cupid can also reduce latency.
- 3 August 2024
- Published by Association for Computing Machinery (ACM)
- p. 88-94
Open AccessRevisiting Congestion Control for WiFi Networks
WiFi networks, widely utilized by wireless devices, have become increasingly complex and congested environments, leading to noticeable delays, jitter, and throughput degradation for end-to-end network flows in today’s Internet. Through detailed experimental observations, we identified the TCP victim problem in WiFi, where TCP erroneously detects congestion and interacts with WiFi routers, resulting in a significant throughput decrease for certain hosts. In this paper, we introduce Cupid, a novel congestion control algorithm that relies on receiver-side WiFi physical layer measurements. Cupid accurately assesses congestion by measuring parameters such as airtime utilization, concurrency, and rates, allowing for precise rate adjustments. Our results demonstrate that whether employed alone or alongside other congestion controls, Cupid effectively mitigates the TCP victim problem. Furthermore, when used independently, Cupid can also reduce latency.
- Open Access
- 29 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 95-101
Amphis: Rearchitecturing Congestion Control for Capturing Internet Application Variety
- 29 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 95-101
Open AccessAmphis: Rearchitecturing Congestion Control for Capturing Internet Application Variety
- Open Access
- 29 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-7
FastWake: Revisiting Host Network Stack for Interrupt-mode RDMA
- 29 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessFastWake: Revisiting Host Network Stack for Interrupt-mode RDMA