Design and implementation of a LoRa mesh network with an optimized routing protocol

By: Call Number: AIT ISPR CS no.25-01 Contributor(s): Material type: SeriesSeries: Asian Institute of Technology. Internship Report ; no. CS-25-01Publication details: Pathum Thani, Thailand : Asian Institute of Technology, 2025Description: 108 leaves : ill.+ 1 online resourceSubject(s): Online resources: Dissertation note: Internship Report (M. Sc.) - Asian Institute of Technology, 2025 Summary: Low-Power Wide-Area Networks using standard LoRaWAN struggle with coverage gaps due to single-hop topology limitations. While mesh networking addresses this through multi-hop relay, existing LoRa mesh protocols face scalability barriers from broadcast-based routing and fixed-interval control overhead. Flooding protocols create exponential tra!c violating duty cycle constraints, while table-driven protocols waste airtime with unnecessary periodic control packets regardless of network stability. This research implements a gateway-aware cost routing protocol combining three mechanisms: (1) Trickle adaptive scheduling reducing HELLO overhead through exponential backo" and redundancy suppression, (2) multi-metric cost function integrating signal quality and gateway load for path selection, and (3) proactive fault detection with safety mechanisms preventing over-suppression while enabling rapid convergence.The research makes six novel contributions: (1) first complete Trickle adaptive scheduler integrated with LoRaMesher firmware achieving 85-90% suppression e!ciency, (2) discovery that Trickle operates as local per-node decisions rather than network-wide cascades, limiting fault impact regionally, (3) zero-overhead ETX tracking via sequence-gap detection eliminating ACKoverhead, (4) active gateway load sharing with real-time load encoding enabling dynamic traffic distribution, (5) safety HELLO mechanism preventing over-suppression while enabling rapid fault detection, and (6) proactive health monitoring reducing fault detection time versus library baseline.Hardware validation on ESP32-S3 nodes demonstrates approximately 30% HELLO overhead reduction, 96-100% packet delivery ratio in indoor scenarios, and successful dual-gateway load distribution. Multi-hop routing capability is validated with relay nodes forwarding traffic and cost-based routing selecting quality-aware paths unavailable in hop-count protocols. The adaptive overhead reduction and fault-tolerant design enable scalable LoRa mesh deploy ments for resource-constrained applications including agricultural monitoring, industrial IoT, and environmental sensing requiring duty cycle compliance and network resilience. The local fault isolation discovery and zero-overhead ETX tracking represent fundamental contributions to LPWAN mesh protocol research.
Tags from this library: No tags from this library for this title. Log in to add tags.
Star ratings
    Average rating: 0.0 (0 votes)
Holdings
Cover image Item type Current library Home library Collection Shelving location Call number Materials specified Vol info URL Copy number Status Notes Date due Barcode Item holds Item hold queue priority Course reserves
67-Electronic Resource Asian Institute of Technology Library Archives AIT ISPR CS no.25-01 (Browse shelf(Opens below)) 1 Not for loan

An Internship Study submitted in partial fulfillment of the requirements for the degree of Master of Science in Computer Science

Internship Report (M. Sc.) - Asian Institute of Technology, 2025

Low-Power Wide-Area Networks using standard LoRaWAN struggle with coverage gaps due to single-hop topology limitations. While mesh networking addresses this through multi-hop relay, existing LoRa mesh protocols face scalability barriers from broadcast-based routing and fixed-interval control overhead. Flooding protocols create exponential tra!c violating duty cycle constraints, while table-driven protocols waste airtime with unnecessary periodic control packets regardless of network stability. This research implements a gateway-aware cost routing protocol combining three mechanisms: (1) Trickle adaptive scheduling reducing HELLO overhead through exponential backo" and redundancy suppression, (2) multi-metric cost function integrating signal quality and gateway load for path selection, and (3) proactive fault detection with safety mechanisms preventing over-suppression while enabling rapid convergence.The research makes six novel contributions: (1) first complete Trickle adaptive scheduler integrated with LoRaMesher firmware achieving 85-90% suppression e!ciency, (2) discovery that Trickle operates as local per-node decisions rather than network-wide cascades, limiting fault impact regionally, (3) zero-overhead ETX tracking via sequence-gap detection eliminating ACKoverhead, (4) active gateway load sharing with real-time load encoding enabling dynamic traffic distribution, (5) safety HELLO mechanism preventing over-suppression while enabling rapid fault detection, and (6) proactive health monitoring reducing fault detection time versus library baseline.Hardware validation on ESP32-S3 nodes demonstrates approximately 30% HELLO overhead reduction, 96-100% packet delivery ratio in indoor scenarios, and successful dual-gateway load distribution. Multi-hop routing capability is validated with relay nodes forwarding traffic and cost-based routing selecting quality-aware paths unavailable in hop-count protocols. The adaptive overhead reduction and fault-tolerant design enable scalable LoRa mesh deploy ments for resource-constrained applications including agricultural monitoring, industrial IoT, and environmental sensing requiring duty cycle compliance and network resilience. The local fault isolation discovery and zero-overhead ETX tracking represent fundamental contributions to LPWAN mesh protocol research.

There are no comments on this title.

to post a comment.
คัดลอกแล้ว!