SD-CPS

Software-Defined Cyber-Physical Systems

SD-CPS is a research framework for orchestrating and managing Industrial Edge Cyber-Physical Systems (CPS) using a software-defined approach. It implements advanced features like SMART adaptation, control plane high-availability, chaos engineering, and privacy-by-design.

Architecture

The framework is structured into a modular hierarchy:

• org.sdcps: System Bootstrap (SDCPSMain).
• org.sdcps.core: Orchestration and Topology Management.
• org.sdcps.knowledge: SMART Adaptation Logic and Empirical Validation.
• org.sdcps.data: Digital Twin Simulation and Privacy Guards.
• org.sdcps.network: Messaging and Transport Layer (M4T).
• org.sdcps.workflow: Automation and Edge Workflow Management.

Quick Start

Prerequisites

• Java 21+
• Maven 3.8+

1. Build the System

mvn clean install

2. Run the Research Suite

mvn exec:java -Dexec.mainClass="org.sdcps.SDCPSMain"

3. Automated Parity Verification

/bin/bash ./verify_parity.sh

Verified Research Case Studies

The project includes an automated suite to verify 10 critical research case studies:

1. Static Placement: Greedy placement of services on edge nodes (SDS 2017).
2. SMART Adaptation: Dynamic subflow cloning for elephant flows (Cluster Computing 2019).
3. Link Fluctuations: Real-time response to transient network jitter.
4. Multi-tenancy: Secure namespace isolation between tenants (SDS 2017).
5. High Availability: ODL-style orchestrator clustering and warm failover.
6. Thermal Constraints: Energy-aware placement avoiding thermal breaches.
7. Empirical Validation: 1000+ iteration simulation with 95% Confidence Intervals.
8. Chaos Engineering: Automated self-healing after simulated node crashes.
9. Privacy-by-Design: Automated PII masking and telemetry anonymization.
10. Unit Testing: 100% logic coverage for core adaptation decision paths.

Citing SD-CPS

If you use SD-CPS in your research, please cite the following papers:

• Kathiravelu, P., Van Roy, P. and Veiga, L., 2019. SD-CPS: software-defined cyber-physical systems. Taming the challenges of CPS with workflows at the edge. Cluster Computing, 22(3), pp.661-677.

• Kathiravelu, P. and Veiga, L., 2017, May. SD-CPS: taming the challenges of cyber-physical systems with a software-defined approach. In 2017 Fourth International Conference on Software Defined Systems (SDS) (pp. 6-13). IEEE.