Juan Luo

Hunan University
1 previous affiliation
4Machine learning papers tracked by Lacuna
Is this you?

Research Overview

Juan Luo focuses on cloud computing resource management, edge computing architectures, and large-scale radio-frequency identification (RFID) systems. Early work addressed virtual machine placement and load balancing in cloud environments to optimize energy consumption and service level agreements. More recently, Luo's research has shifted toward energy-aware computational offloading and dynamic tag estimation in mobile Internet of Things (IoT) networks.

Research Trajectory

Luo’s early research focused on the infrastructure management of centralized cloud computing environments, specifically addressing the operational challenges of resource allocation, system overloads, and energy efficiency. In 2016, Luo contributed as a second author to An Anti-Overload Model for OpenStack Based on an Effective Dynamic Migration. This work investigated methods for maintaining system stability in OpenStack cloud deployments, utilizing dynamic migration techniques to prevent server overloads and ensure consistent performance.

Building on this foundation in cloud resource management, Luo continued to explore the optimization of virtualized environments. In 2019, Luo was the second author of Intra-Balance Virtual Machine Placement for Effective Reduction in Energy Consumption and SLA Violation. This research addressed a critical tension in data center management: the trade-off between aggressive virtual machine (VM) consolidation—which reduces power consumption—and the resulting degradation of performance that leads to Service Level Agreement (SLA) violations. The paper highlighted how unbalanced resource consumption exacerbates the problem of unavailable or wasted resources, proposing an intra-balance VM placement strategy to simultaneously mitigate energy costs and SLA breaches.

Around the same time, Luo’s research expanded beyond centralized data centers to address the computational constraints of distributed IoT devices. As a middle author, Luo participated in the development of An Energy-Aware Offloading Framework for Edge-Augmented Mobile RFID Systems (2019). Because IoT end devices often lack the computation capability and power supply required for demanding tasks, this work analyzed the energy consumption characteristics of mobile RFID readers and proposed offloading computational tasks to edge servers. This transition reflects an engagement with the broader research direction of Intelligent Data Filtering at the Network Edge, an area concerned with alleviating the burden on centralized cloud architectures by processing data closer to its source.

Luo’s involvement in RFID system optimization continued into 2021. Participating again as a middle author, Luo contributed to Fast and Reliable Dynamic Tag Estimation in Large-Scale RFID Systems. This research tackled the inherent volatility of practical RFID applications, such as supply chain and stock management, where tags frequently move in and out of the system's read range. The paper proposed time-efficient solutions for simultaneously estimating the number of missing tags (those that have left the system) and unknown tags (those that have newly entered) within a specified time interval, ensuring high estimation accuracy in highly dynamic environments.

Current Directions

Based on the most recent papers in the evidence, which date to 2021, Luo’s research interests have increasingly centered on the operational challenges of large-scale, dynamic IoT and RFID networks. While early work focused heavily on static resource allocation in cloud data centers, recent efforts investigate how to manage computation and track physical assets in decentralized environments where resources are constrained and system states change rapidly.

The primary open problems addressed in this recent work include minimizing energy consumption on low-power end devices through edge-assisted offloading, and developing fast, reliable algorithms for state estimation in sensor networks with high churn rates. These themes suggest a continuing focus on bridging the gap between hardware limitations at the network edge and the demand for accurate, real-time data processing in large-scale deployments.

Profile updated

See something wrong? Report a mistake