Call for Papers– IEEE IoT Magazine - Vertical Area 5: Edge Intelligence IoT

[Marica Amadeo]

*** Apologies for cross-postings ***
============================================
CALL FOR PAPERS

IEEE Internet of Things Magazine
Vertical Area 5: Edge Intelligence IoT
============================================
Dear Colleagues,

We are pleased to invite submissions to the newly launched vertical
area on "Edge Intelligence for IoT" in IEEE IoT Magazine:
https://www.comsoc.org/publications/magazines/ieee-internet-things-magazine/cfp/vertical-area-5-edge-intelligence-iot

============================================

The IEEE Internet of Things Magazine (IEEE IoTM) is currently
soliciting articles for its vertical area on Edge Intelligence IoT.
The proliferation of IoT devices operating in resource-constrained,
latency-sensitive, and connectivity-limited environments has made it
increasingly inadequate to rely solely on centralized cloud
infrastructures for data processing and decision-making. Edge
computing has emerged as a natural response to these limitations,
bringing computational resources closer to data sources and reducing
the burden on backbone networks. However, proximity alone is no longer
sufficient. As IoT systems grow in complexity and scale, there is a
pressing need to move beyond simple data forwarding or pre-processing
at the edge, and to embed genuine intelligence directly into the
network periphery.
Edge Intelligence represents this next step: a paradigm in which AI
models, inference pipelines, and adaptive learning capabilities are
deployed directly at or near the data sources, on embedded devices,
edge nodes, and edge computing servers, enabling autonomous,
real-time, and context-aware IoT systems. This paradigm is
particularly relevant for advanced IoT applications such as digital
twins and other data-intensive, real-time systems, where continuous
interaction between physical and digital entities must be supported
efficiently and with low latency.

This vertical area focuses on the foundational challenges and
opportunities of bringing intelligence to the edge of the network to
support IoT systems and applications, spanning hardware-aware model
design, edge-native architectures, distributed inference and training
algorithms, and the seamless orchestration of resources across the
edge–cloud continuum.

Articles revolving around this vertical area should discuss:

Edge-native architectures and computing paradigms for IoT:
- How can IoT systems be designed such that intelligence is inherently
embedded at the edge?
- What architectural patterns enable low-latency, autonomous
decision-making on resource-constrained devices?
- How do fog computing, multi-access edge computing (MEC), and
hierarchical edge architectures complement each other in intelligent
IoT deployments

Resource-efficient Edge Intelligence for IoT
- How can machine learning models be designed, trained, and deployed
under the strict memory, compute, and energy constraints of embedded
IoT hardware?
- What are the trade-offs between model compression techniques (e.g.,
quantization, pruning, knowledge distillation) and inference accuracy
in real-world deployments?
- How can on-device learning and model adaptation be achieved with
limited or no reliance on cloud connectivity?

Edge–cloud continuum and workload orchestration
- How can inference and learning tasks be dynamically allocated across
the edge–cloud continuum considering latency, bandwidth, and energy
constraints?
- What mechanisms enable seamless and intelligent offloading decisions
between IoT devices, edge, and cloud tiers?
- How can the lifecycle of AI models, from training to deployment and
continuous updates, be managed across heterogeneous and distributed
IoT edge infrastructures?
- How can federated learning and other distributed learning techniques
be made effective under the constraints of IoT environments?

Dependable and secure edge intelligence for IoT
- How can deterministic response times be guaranteed for AI-driven
decisions in safety-critical or time-sensitive IoT applications?
- What are the trade-offs between inference latency, model complexity,
and hardware capabilities in real-world deployments?
- How can edge intelligence support privacy-preserving and secure data
processing, enabling sensitive data to remain on-device or within
local infrastructures while still benefiting from advanced AI
capabilities?
- What are the specific threats and vulnerabilities introduced by
deploying AI models on exposed and resource-constrained edge nodes,
and how can they be mitigated?

============================================
Important Dates
Publication Date: Open
Manuscript Submission Date: Open

--
Dr. Marica Amadeo
Tenure Track Assistant Professor
Department of Engineering
University of Messina (Italy)
E-mail: marica...@unime.it; marica...@gmail.com