Self-Healing Network of Interconnected Edge Devices Empowered by Infrastructure-as-Code and LoRa Communication

2025-08-22

AI Summary

To address low throughput, packet collisions, and single-point-of-failure vulnerabilities in LoRa ad hoc networks under traditional network failure scenarios, this paper proposes a self-healing edge network architecture based on Raspberry Pi. Methodologically, it integrates Infrastructure-as-Code (IaC) principles with the LoRa physical layer to design a containerized lightweight deployment framework; incorporates time-slot scheduling, adaptive fragmentation-based retransmission, and distributed health monitoring; and eliminates TCP/IP dependency to enable sub-second service auto-migration. The primary contribution is the first application of the IaC paradigm at the LoRa link layer, enabling decentralized, network-stack-agnostic autonomous recovery. Experiments demonstrate that fragmentation-based retransmission improves effective throughput by 37%; service migration latency upon node failure is ≤1 second; and system service availability remains at 99.2% even under high packet loss (≥40%) and long propagation delays (>5 s).

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Application Domains: Internet of Things, Sensor Networks & Smart Cities

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

This Paper proposes a self-healing, automated network of Raspberry Pi devices designed for deployment in scenarios where traditional networking is unavailable. Leveraging the low-power, long-range capabilities of the LoRa (Long Range) protocol alongside Infrastructure as Code (IaC) methodologies, the research addresses challenges such as limited bandwidth, data collisions, and node failures. Given that LoRa's packet-based system is incompatible with conventional IaC tools like Ansible and Terraform, which rely on TCP/IP networking, the research adapts IaC principles within a containerised architecture deployed across a Raspberry Pi cluster. Evaluation experiments indicate that fragmenting data packets and retransmitting any missed fragments can mitigate LoRa's inherent throughput and packet size limitations, although issues such as collisions and line-of-sight interference persist. An automated failover mechanism was integrated into the architecture, enabling unresponsive services to be redeployed to alternative nodes within one second, demonstrating the system's resilience in maintaining operational continuity despite node or service failures. The paper also identifies practical challenges, including the necessity for time-slotting transmissions to prevent data packet overlap and collisions. Future research should explore the integration of mesh networking to enhance range, develop more advanced scheduling algorithms, and adopt cutting-edge low-power wide-area network (LPWAN) techniques.

Problem

• Creating self-healing networks where traditional networking is unavailable
• Adapting Infrastructure-as-Code for LoRa's packet-based non-TCP/IP system
• Addressing LoRa limitations like bandwidth constraints and node failures

Innovation

• LoRa protocol with Infrastructure-as-Code adaptation
• Containerized architecture on Raspberry Pi cluster
• Automated failover mechanism for service redeployment