Research on Hybrid Data Flow Scheduling Algorithm for Mobile Robot Edge-End Collaborative Inference

DING Nan; FANG Xi-qi; HAO Yun-tao; HU Chuang-ye; XU Li

doi:10.12263/DZXB.20250674

您当前的位置：

首页 >

文章列表页 >

Research on Hybrid Data Flow Scheduling Algorithm for Mobile Robot Edge-End Collaborative Inference

PAPERS | 更新时间：2026-04-24

- Research on Hybrid Data Flow Scheduling Algorithm for Mobile Robot Edge-End Collaborative Inference
- ACTA ELECTRONICA SINICA Vol. 53, Issue 12, Pages: 4575-4591(2025)
- 作者机构：
  
  1.大连理工大学计算机科学与技术学院，辽宁大连 116024
  2.大连理工大学工业装备智能控制与优化教育部重点实验室，辽宁大连 116024
  3.北京字节跳动科技有限公司，北京 100101
- 作者简介：
- 基金信息：
  
  National Natural Science Foundation of China(62262066;62473071)
- DOI：10.12263/DZXB.20250674
  CLC： TP393;
- Received：04 August 2025，
  
  Accepted：12 December 2025，
  
  Published：25 December 2025
- 稿件说明：
移动端阅览
丁男, 方玺淇, 郝云涛, 等. 面向移动机器人端边协同推理的混合数据流调度算法研究[J]. 电子学报, 2025, 53(12): 4575-4591.

DING Nan, FANG Xi-qi, HAO Yun-tao, et al. Research on Hybrid Data Flow Scheduling Algorithm for Mobile Robot Edge-End Collaborative Inference[J]. Acta Electronica Sinica, 2025, 53(12): 4575-4591.
丁男, 方玺淇, 郝云涛, 等. 面向移动机器人端边协同推理的混合数据流调度算法研究[J]. 电子学报, 2025, 53(12): 4575-4591. DOI：10.12263/DZXB.20250674

DING Nan, FANG Xi-qi, HAO Yun-tao, et al. Research on Hybrid Data Flow Scheduling Algorithm for Mobile Robot Edge-End Collaborative Inference[J]. Acta Electronica Sinica, 2025, 53(12): 4575-4591. DOI：10.12263/DZXB.20250674

摘要

基于分区的深度神经网络（Deep Neural Network，DNN）端边协同推理技术通过将模型拆分并分别在移动机器人终端和边缘服务器上部署，能够有效缓解端设备的资源受限以及现有模型轻量级化技术导致的推理精度降低等问题.然而，该技术也为机器人操作系统（Robot Operating System2，ROS2）的通信调度提出了新的挑战：现有的通信策略难以在保障协同推理关键数据流有效传输的同时，兼顾其他应用数据流的传输需求.针对这一问题，本研究提出了机器人操作系统中面向移动机器人深度神经网络端边协同推理的混合数据流动态调度算法（Hybrid Data Flow Dynamic Scheduling Algorithm for Mobile Robot Deep Neural Network Edge-End Collaborative Inference in the Robot Operating System2，DRECHS）.首先，基于端边协同推理的机理分析，定义了深度神经网络中间数据的最大允许传输时间边界条件，为传输优化提供了理论基础.结合边界条件，设计了一种基于混杂切换系统理论的调度模型，将流调度过程建模为包含优先级优先子系统和时间优先子系统的动态切换模型.在此基础上，提出了具体的混合数据流调度算法.该算法集成在机器人操作系统的数据分发服务（Data Distribution Service，DDS）流控制器中，能够依据计算出的队列状态指标动态生成输出队列，实现对底层数据传输顺序的细粒度控制，从而在满足推理任务数据传输要求的基础上，实现对不同优先级数据流的差异化服务质量（Quality of Service，QoS）优化，有效平衡了系统的整体传输性能.针对所采用的动态分区方法，设计不同带宽条件下的仿真实验，对比分析了所提算法与系统内置调度算法等在传输延迟和丢包率方面的性能差异.实验结果表明，本研究提出的调度算法通过混杂切换系统模型和动态调度策略，在满足高优先级数据传输需求的同时，成功实现了对不同优先级数据流的差异化服务质量优化.此外，本研究提出了相应的部署方案，并在真实设备上部署了该调度算法及深度神经网络端边协同推理框架，完成了系统验证.该部署方案为本研究所提算法及框架在真实场景中的部署提供了参考.

Abstract

Partition-based edge-end collaborative inference technology for deep neural networks (DNN)

by splitting models and deploying them on mobile robot terminals and edge servers respectively

can effectively alleviate resource constraints on terminal devices and address the issue of reduced inference accuracy caused by existing model lightweighting techniques. However

this technology also poses new challenges for communication scheduling in the robot operating system2 (ROS2): existing communication strategies struggle to ensure the effective transmission of critical collaborative inference data flows while simultaneously accommodating the transmission needs of other application data flows. To address this problem

this study proposes a hybrid data flow dynamic scheduling algorithm for mobile robot deep neural network edge-end collaborative inference in the robot operating system2 (DRECHS). First

based on the mechanism analysis of edge-end collaborative inference

we define the maximum allowable transmission time boundaries for DNN intermediate data to provide a theoretical basis for transmission optimization. Combining these boundary conditions

we design a scheduling model based on hybrid switching system theory

modeling the flow scheduling process as a dynamic switching model containing a priority-first subsystem and a time-first subsystem. On this basis

the specific hybrid data flow scheduling algorithm is proposed. Integrated into the data distribution service (DDS) flow controller of the robot operating system2

this algorithm is capable of dynamically generating output queues based on calculated queue status metrics

realizing fine-grained control of the underlying data transmission order. Thus

while meeting the transmission requirements of inference tasks

it achieves differentiated quality of service (QoS) optimization for data flows with different priorities

effectively balancing overall system transmission performance. Targeting the adopted dynamic partitioning method

we design simulation experiments under different bandwidth conditions to compare and analyze the performance differences between the proposed algorithm and the system's built-in scheduling algorithms and others in terms of transmission delay and packet loss rate. Experimental results show that the proposed scheduling algorithm

through the hybrid switching system model and dynamic scheduling strategy

successfully achieves differentiated quality of service optimization for data flows with different priorities while meeting high-priority data transmission requirements. Furthermore

this study proposes a corresponding deployment scheme

and deploys the scheduling algorithm and the deep neural network edge-end collaborative inference framework on real devices

completing system verification. This deployment scheme provides a reference for the deployment of the proposed algorithm and framework in real-world scenarios.

关键词

Keywords

references

李智勇 , 王琦 , 陈一凡 , 等 . 车辆边缘计算环境下任务卸载研究综述 [J ] . 计算机学报 , 2021 , 44 ( 5 ): 963 - 982 .

LI Z Y , WANG Q , CHEN Y F , et al . A survey on task offloading research in vehicular edge computing [J ] . Chinese Journal of Computers , 2021 , 44 ( 5 ): 963 - 982 . (in Chinese)

周悦芝 , 张迪 . 近端云计算: 后云计算时代的机遇与挑战 [J ] . 计算机学报 , 2019 , 42 ( 4 ): 677 - 700 .

ZHOU Y Z , ZHANG D . Near-end cloud computing: Opportunities and challenges in the post-cloud computing era [J ] . Chinese Journal of Computers , 2019 , 42 ( 4 ): 677 - 700 . (in Chinese)

GROSHEV M , BALDONI G , COMINARDI L , et al . Edge robotics: Are we ready? An experimental evaluation of current vision and future directions [J ] . Digital Communications and Networks , 2023 , 9 ( 1 ): 166 - 174 .

CHOI H S , ENRIGHT D , SOBHANI H , et al . Priority-driven real-time scheduling in ROS 2: Potential and challenges [C ] // 1st International Workshop on Real-time And intelliGent Edge computing (RAGE) . Piscataway : IEEE , 2022 .

芦倩 , 李晓娟 , 关永 , 等 . 面向数据流的ROS2数据分发服务形式建模与分析 [J ] . 软件学报 , 2021 , 32 ( 6 ): 1818 - 1829 .

LU Q , LI X J , GUAN Y , et al . Modeling and analysis of ROS2 data distribution service for data flow [J ] . Journal of Software , 2021 , 32 ( 6 ): 1818 - 1829 . (in Chinese)

鲁敬敬 , 秦云川 , 刘志中 , 等 . 机器人操作系统ROS安全性研究综述 [J ] . 软件学报 , 2024 , 35 ( 2 ): 1010 - 1027 .

LU J J , QIN Y C , LIU Z Z , et al . Survey on security of robot operating system ROS [J ] . Journal of Software , 2024 , 35 ( 2 ): 1010 - 1027 . (in Chinese)

SABOIA M , CLARK L , THANGAVELU V , et al . ACHORD: Communication-aware multi-robot coordination with intermittent connectivity [J ] . IEEE Robotics and Automation Letters , 2022 , 7 ( 4 ): 10184 - 10191 .

MAHEMUTI P , YANG L , LUO L L . Modeling, analysis and synthesis of hybrid system: A review [J ] . Applied Mechanics and Materials , 2014 , 615 : 36 - 43 .

陈阳 , 皮德常 , 代成龙 , 等 . 多无人机协同陆地设施辅助移动边缘计算的系统能耗最小化方法 [J ] . 电子学报 , 2023 , 51 ( 4 ): 984 - 992 .

CHEN Y , PI D C , DAI C L , et al . Energy minimization for multi-UAVs cooperative ground access points assisted mobile edge computing [J ] . Acta Electronica Sinica , 2023 , 51 ( 4 ): 984 - 992 . (in Chinese)

许小龙 , 杨威 , 杨辰翊 , 等 . 车联网边缘计算环境下基于流量预测的高效任务卸载策略研究 [J ] . 电子学报 , 2025 , 53 ( 2 ): 329 - 343 .

XU X L , YANG W , YANG C Y , et al . Efficient task offloading based on traffic prediction in IoV-enabled edge computing [J ] . Acta Electronica Sinica , 2025 , 53 ( 2 ): 329 - 343 . (in Chinese)

QUERALTA J P , LI Q Q , ZHUO Z , et al . Enhancing autonomy with blockchain and multi-access edge computing in distributed robotic systems [C ] // 2020 Fifth International Conference on Fog and Mobile Edge Computing . Piscataway : IEEE , 2020 : 180 - 187 .

WAN S H , GU Z H , NI Q . Cognitive computing and wireless communications on the edge for healthcare service robots [J ] . Computer Communications , 2020 , 149 : 99 - 106 .

MATSUBARA Y , LEVORATO M , RESTUCCIA F . Split computing and early exiting for deep learning applications: Survey and research challenges [J ] . ACM Computing Surveys , 2023 , 55 ( 5 ): 1 - 30 .

KANG Y P , HAUSWALD J , GAO C , et al . Neurosurgeon: Collaborative intelligence between the cloud and mobile edge [J ] . ACM SIGARCH Computer Architecture News , 2017 , 45 ( 1 ): 615 - 629 .

HU C , BAO W , WANG D , et al . Dynamic adaptive DNN surgery for inference acceleration on the edge [C ] // IEEE INFOCOM 2019 - IEEE Conference on Computer Communications . Piscataway : IEEE , 2019 : 1423 - 1431 .

JAHIER PAGLIARI D , CHIARO R , MACII E , et al . CRIME: Input-dependent collaborative inference for recurrent neural networks [J ] . IEEE Transactions on Computers , 2021 , 70 ( 10 ): 1626 - 1639 .

JALIL A , KOBAYASHI J , SAITOH T . Performance improvement of multi-robot data transmission in aggregated robot processing architecture with caches and QoS balancing optimization [J ] . Robotics , 2023 , 12 ( 3 ): 87 .

MARUYAMA Y , KATO S , AZUMI T . Exploring the performance of ROS2 [C ] // 2016 International Conference on Embedded Software . Piscataway : IEEE , 2016 : 1 - 10 .

KRONAUER T , POHLMANN J , MATTHÉ M , et al . Latency analysis of ROS2 multi-node systems [C ] // 2021 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems . Piscataway : IEEE , 2021 : 1 - 7 .

JIANG Z Y , GONG Y F , ZHAI J D , et al . Message passing optimization in robot operating system [J ] . International Journal of Parallel Programming , 2020 , 48 ( 1 ): 119 - 136 .

WANG Y P , TAN W D , HU X Q , et al . TZC: Efficient inter-process communication for robotics middleware with partial serialization [C ] // 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems . Piscataway : IEEE , 2020 : 7805 - 7812 .

LUO X T , JIANG X , TANG Y , et al . Analysis and optimization of communication delay in multi-subscriber environments of ROS 2 [J ] . Journal of Systems Architecture , 2025 , 164 : 103428 .

TEPER H , BETZ T , GÜNZEL M , et al . End-to-end timing analysis and optimization of multi-executor ROS 2 systems [C ] // 2024 IEEE 30th Real-Time and Embedded Technology and Applications Symposium . Piscataway : IEEE , 2024 : 212 - 224 .

CASTILLO-SÁNCHEZ J B , GONZÁLEZ-PARADA E , CANO-GARCÍA J M . Swarm robot communications in ROS 2: An experimental study [J ] . IEEE Access , 2024 , 12 : 142930 - 142943 .

AL ARAFAT A , WILSON K , YANG K C , et al . Dynamic priority scheduling of multithreaded ROS 2 executor with shared resources [J ] . IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems , 2024 , 43 ( 11 ): 3732 - 3743 .

TEPER H , BELL O , GÜNZEL M , et al . Reconciling ROS 2 with classical real-time scheduling of periodic tasks [C ] // 2025 IEEE 31st Real-Time and Embedded Technology and Applications Symposium . Piscataway : IEEE , 2025 : 177 - 189 .

TAO X D , QIAN X Y , HAN L , et al . Key flow first prioritized flow scheduling strategy in multi-tenant data centers [J ] . IEEE Transactions on Network and Service Management , 2024 , 21 ( 3 ): 3264 - 3277 .

DING N , WANG Y H , ZHAO Y W , et al . Packet scheduling algorithm based on hybrid system theory in the industrial Internet of Things [C ] // 2021 IEEE International Conference on Smart Internet of Things . Piscataway : IEEE , 2021 : 33 - 39 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Static Facial Expression Recognition System Using ROI Deep Neural Networks

Related Author

SUN Xiao

PAN Ting

Related Institution

School of Computer and Information, Hefei University of Technology.

⁰