Dongxu Wang

Papers

Total Citations

H-Index

About

Dongxu Wang is a researcher at the forefront of edge computing and adaptive inference systems, with a focus on optimizing real-time, resource-constrained environments. His most notable contribution, detailed in his 2025 paper "Adaptive scheduling of online inference pipelines at the edge: A post-hoc request-oriented approach," introduces a novel framework that dynamically reorders and schedules inference tasks based on incoming request characteristics. This work addresses the critical challenge of balancing latency, accuracy, and throughput in distributed edge networks, where traditional static scheduling often fails under variable workloads. By leveraging a post-hoc, request-oriented strategy, Wang’s approach enables more efficient use of limited computational resources without sacrificing model performance. Although early in its citation trajectory, this paper has already garnered attention for its practical implications in IoT, autonomous systems, and real-time analytics. Wang’s research bridges the gap between theoretical scheduling algorithms and deployable edge solutions, positioning him as an emerging voice in the intersection of machine learning systems and distributed computing. His work promises to shape how future edge platforms handle the growing demand for low-latency, high-accuracy inference.