Edge Intelligence in the Making: Optimization, Deep Learning, and Applications

When AI moves from centralized servers toward phones, sensors, and other network-edge devices, the technical questions change: how should models be trained, how can communication be kept efficient, and where should inference happen? Edge Intelligence in the Making examines these challenges at the intersection of deep learning, optimization, and networked systems.

Sen Lin, Zhi Zhou, Zhaofeng Zhang, Xu Chen, and Junshan Zhang combine a survey of the field with research projects that illuminate both theoretical questions and implementation concerns. The result is a focused technical account of how learning can operate across edge devices and the cloud.

From edge computing to edge intelligence

The book establishes the motivation for bringing AI closer to data generated by mobile computing and Internet of Things applications. It introduces relevant deep-learning concepts and considers why model training and inference at the network edge call for approaches suited to distributed systems.

Training models across distributed devices

Model training receives substantial attention, including federated learning and techniques related to communication cost, aggregation, gradient compression, hardware acceleration, knowledge transfer, and distillation. These themes show how learning methods must account for the practical limits and coordination demands of edge environments.

Federated learning, optimization, and collaboration

Further chapters explore federated meta-learning and collaborative learning between edge and cloud resources, including distributionally robust optimization. The discussion connects algorithmic analysis with questions about adaptation, coordination, and how learning workloads can be shared across a network.

Inference and applications at the edge

Beyond training, the volume considers model inference and the broader applications, marketplace, and research directions associated with edge intelligence. Its scope makes the book relevant to readers studying AI systems as well as the communications and computing infrastructure that supports them.

For readers working across AI and networks

This research-oriented survey is suited to advanced students, researchers, and technical professionals interested in edge computing, deep learning, IoT, federated learning, and distributed optimization. It offers a way to see these subjects not as isolated topics, but as interlocking parts of the same engineering problem.