Open-source edge stack

OpenISAC-Edge

A research platform for building, testing, and sharing integrated sensing and communication workflows across edge devices, wireless hardware, and chip-aware AI systems.

GitHub Paper Docs

Wireless Edge Experiments

Prototype ISAC pipelines close to sensing, radio, and compute resources.

Chip and Hardware Awareness

Bridge software workflows with edge boards, radio front ends, and accelerator constraints.

AI-Native Integration

Connect edge intelligence with O-RAN/6G experimentation and system validation.

Why OpenISAC-Edge

A shared edge stack for sensing-communication research

OpenISAC-Edge is being developed to make ISAC experiments easier to reproduce across labs, devices, and wireless testbeds. The platform brings together sensing data capture, radio-side context, edge inference, and experiment orchestration so researchers can evaluate communication and sensing behavior as one closed-loop system.

The project emphasizes practical edge deployment: lightweight services, hardware-aware configuration, repeatable experiment recipes, and interfaces that can connect AI-native control logic with O-RAN and 6G prototyping environments.

Platform stack

Modular layers for end-to-end ISAC workflows

01

Sensing and Radio Interfaces

Adapters for collecting sensing streams, radio measurements, beam or channel context, and device-side metadata from edge test environments.

02

Edge Runtime

Lightweight runtime components for coordinating data movement, local services, model execution, logging, and experiment state close to the hardware.

03

AI-Native Control

Hooks for learning-based decision modules, including adaptive sensing-communication modes, edge intelligence, and closed-loop wireless control.

04

Evaluation and Sharing

Reusable experiment descriptions, metrics, and documentation patterns to help teams compare results and contribute reproducible workflows.

Experiment workflow

From testbed setup to reusable results

  1. Configure

    Describe the edge device, radio or sensing hardware, software services, model assets, and experiment parameters in a repeatable setup.

  2. Run

    Launch coordinated sensing, communication, inference, and logging tasks on local or testbed-connected edge nodes.

  3. Observe

    Track system behavior through telemetry, logs, performance counters, sensing outputs, and communication-side measurements.

  4. Compare

    Package results with configuration context so ISAC algorithms, hardware settings, and AI control policies can be evaluated side by side.

Research use cases

Designed for collaborative FutureG experimentation

ISAC Prototyping

Build integrated sensing and communication experiments where waveform, beam, sensing, and inference choices affect one another.

Adaptive Mode Selection

Evaluate policies that switch between sensing, communication, and composite operating modes based on environment and utility.

Chip-Aware Edge AI

Study model placement, latency, memory, accelerator use, and power-aware deployment constraints on edge hardware.

O-RAN and 6G Testbeds

Connect edge services with open RAN experimentation, xApp/rApp-style control, and AI-native network validation workflows.

Community

Open collaboration across wireless, AI, and hardware communities

We welcome collaborators interested in edge deployment, ISAC algorithms, O-RAN integration, experiment automation, documentation, and reproducible testbed recipes.

View GitHub

Founders

Xiaochan Xue, PhD

Xiaochan Xue, Ph.D.

Founder; Assistant Professor, University of Hawaii at Manoa

Yao Zheng, PhD

Yao Zheng, Ph.D.

Co-Founder; Associate Professor, University of Hawaii at Manoa