Special Sessions

Focused tracks at the frontier of IoT, Data & Cloud Computing — curated by leading researchers to spotlight the most consequential directions shaping 2027 and beyond.

Curated Tracks

Five Frontiers, One Conference

Special Sessions complement the main technical program with deep, theme-driven tracks led by domain experts. Each session welcomes original research papers, position papers, and demonstrations, and follows the same double-blind peer review process as the main track. Selected high-quality papers will be invited for extension and publication in prestigious international journals — with the first MDPI partner journals already confirmed (see below).

🧠
Special Session 01

Generative AI and Large Language Models (LLMs) at the Edge

Generative models are moving off the cloud and onto the devices that surround us. This session explores how foundation models and LLMs can be compressed, distilled, and deployed across resource-constrained edge hardware — enabling private, low-latency intelligence without a round trip to the data center.

Special Issue Partner Journal

Selected papers from this session will be invited for extension and publication in the following MDPI journal:

Topics of Interest

On-device inference & tiny LLMs (quantization, pruning, distillation)
Edge–cloud hybrid and split inference architectures
Retrieval-augmented generation (RAG) for embedded systems
Federated fine-tuning & parameter-efficient adaptation (LoRA, adapters)
Multimodal generative models for sensor & vision streams
Energy-aware & carbon-aware LLM serving at the edge
Privacy-preserving & on-device personalization
NPU/TPU accelerators & hardware co-design for transformers

Why it matters

As LLMs become infrastructure, the bottleneck shifts from model capability to deployment: latency, cost, privacy, and energy. Pushing generative intelligence to the edge unlocks real-time assistants, autonomous agents, and context-aware IoT — while keeping sensitive data on the device.

⚙️
Special Session 02

Edge AI Hardware & Embedded Intelligence for IoT

Intelligence at the edge is ultimately a silicon problem. As models shrink to run on microcontrollers and battery-powered sensors, the decisive gains come from the hardware — accelerators, low-power circuits, and tight hardware–software co-design. This session focuses on the chips, architectures, and embedded systems that make real-time, energy-efficient AI possible on constrained IoT devices.

Special Issue Partner Journal

Selected papers from this session will be invited for extension and publication in the following MDPI journal:

Topics of Interest

TinyML & on-device learning for microcontrollers and constrained nodes
Edge AI accelerators: NPU/TPU, FPGA & ASIC design for inference
Hardware–software co-design & hardware-aware model compression
Low-power & energy-efficient circuits for always-on / battery-free inference
In-memory, analog & neuromorphic computing for edge intelligence
SoC & heterogeneous architectures for IoT edge devices
Embedded vision & sensor-edge signal processing
Reliability, hardware security & trusted execution on edge silicon

Why it matters

Every milliwatt and square millimeter of silicon decides what AI can run untethered from the cloud. Advances in TinyML, edge accelerators, and energy harvesting move inference onto devices that were once too small or too power-starved to be "smart" — turning billions of passive sensors into intelligent endpoints.

🛡️
Special Session 03

Cybersecurity & Zero-Trust Architecture in 6G-Enabled IoT

The perimeter is dead. As 6G promises massive connectivity, ultra-low latency, and AI-native networks, billions of IoT endpoints become both the surface and the target. This session examines how Zero-Trust principles — never trust, always verify — can be engineered into next-generation IoT at scale.

Topics of Interest

Zero-Trust Network Access (ZTNA) for constrained IoT devices
Continuous authentication & micro-segmentation in 6G
AI/ML-driven intrusion & anomaly detection at the edge
Post-quantum & lightweight cryptography for IoT
Identity, attestation & hardware roots of trust
Blockchain & decentralized trust frameworks
Secure network slicing & software-defined perimeters
Threat modeling & resilience for critical IoT infrastructure

Why it matters

6G will connect orders of magnitude more devices than 5G — and traditional castle-and-moat defenses cannot scale to that surface. Zero-Trust reframes security around identity, context, and continuous verification, making it the foundational model for trustworthy hyper-connected IoT.

🫀
Special Session 04

Digital Twins for Smart Healthcare & Wearable IoT

A digital twin is a living, data-driven replica — and in healthcare it can mean a virtual model of a patient, an organ, or an entire hospital workflow. Fed by continuous streams from wearable and implantable IoT, twins enable prediction, personalization, and prevention in ways static records never could.

Topics of Interest

Patient & organ-level digital twin modeling
Real-time physiological monitoring via wearable IoT
Predictive diagnostics & remote patient management
Edge analytics for biosignals (ECG, EEG, PPG, glucose)
Federated & privacy-preserving health data pipelines
Digital twins for hospital operations & resource planning
Interoperability standards (HL7 FHIR, IEEE 11073)
AI-assisted clinical decision support & simulation

Why it matters

Wearables already generate continuous, high-resolution health data. Digital twins turn that stream into foresight — simulating interventions before they reach the patient, personalizing treatment, and shifting medicine from reactive to predictive and preventive care.

🌱
Special Session 05

Sustainable IoT: Green Computing and Energy Harvesting

Trillions of connected devices carry a hidden cost — energy, e-waste, and carbon. This session gathers research on making IoT genuinely sustainable: from battery-free sensors powered by ambient energy to carbon-aware computing and circular hardware design.

Topics of Interest

Energy harvesting (solar, RF, thermal, kinetic, vibration)
Battery-free & intermittent computing
Ultra-low-power protocols & duty-cycling
Carbon-aware & energy-proportional edge/cloud computing
TinyML for energy-constrained inference
Sustainable data centers & green cloud orchestration
Lifecycle assessment, e-waste & circular IoT design
Renewable-powered networks & smart energy grids

Why it matters

The environmental footprint of computing is on track to rival entire industries. Energy harvesting and green computing make it possible to deploy IoT at planetary scale without a proportional rise in carbon and waste — turning sustainability from a constraint into a design principle.

Submitting to a Special Session

📝

Same Review, Same Quality

Special Session papers undergo the identical double-blind peer review as the main track, with 3+ independent expert reviewers.

🏆

Journal Publication Track

Outstanding papers will be selected and their authors invited to extend their work for publication in prestigious international journals — with confirmed MDPI partners (Electronics, Computers) for selected sessions.

🏷️

How to Submit

Select the relevant Special Session on the submission page and follow the standard conference template. Full papers: 6–8 pages.

Distinguished Keynote