LoRa-PDMM: Partially Distributed Mobility Management for Reduced-Latency Active Handover in LoRaWAN Networks

Abstract

Handover management in LoRaWAN networks is a critical challenge for dynamic IoT applications such as connected vehicles and drones. Existing approaches fail to simultaneously satisfy low latency, active handover with full context transfer, and scalability without modifying end devices. This paper proposes LoRa-PDMM, a hybrid architecture combining local Distribution Servers for fast intra-zone handover and a Central Server for inter-zone coordination, without any modification to existing end devices. Evaluation uses discrete-event simulation (Python/SimPy) with explicit LoRaWAN constraints including 1% duty cycle, Class A RX1/RX2 windows, ALOHA-based channel access with capture effect, and configurable context size, over scenarios from 10 to 100 mobile nodes with Random Waypoint and Gauss-Markov mobility. LoRa-PDMM reduces server-side handover latency by 54.7% and total latency by 23.2% over a centralized active handover baseline, with packet delivery ratio between 86.1% and 98.6% under nominal conditions. The latency reduction targets the server-side processing component, which is the architectural variable LoRa-PDMM directly controls; the Class A downlink window wait is a LoRaWAN protocol constraint common to all approaches.