Optimizing LoRaWAN With Adaptive Receive Window Adjustments Considering Network and Application Requirements

Long-Range Wide-Area Network (LoRaWAN) is gaining widespread adoption due to its ability to provide long-range, low-power, and cost-effective communication, making it ideal for transmitting sensor data to remote servers. In LoRaWAN, end devices initiate two receive windows to accept downlink after transmitting data. The optimization of the delay in opening the first receive window, referred to as RX1Delay, is crucial for seamless uplink and downlink communication. However, fixed RX1Delay settings pose challenges in adapting to dynamic application and network requirements. The need for flexibility in RX1Delay stems from both application-driven demands and network-centric considerations. For example, in alert-based applications, the requirement is to reduce RX1Delay to transmit critical data promptly triggered by specific events. Conversely, from a network standpoint, there is a need to elevate RX1Delay when server loads hinder the timely scheduling of downlink packets. This paper proposes a novel approach for estimating and configuring optimal RX1Delay for end devices, leveraging the age of information at the network server. The approach dynamically adjusts RX1Delay to accommodate varying demands across different scenarios, ensuring timely data transmission. Through extensive real-world experimentation within a LoRaWAN deployment, our approach demonstrates significant enhancements in communication efficiency and network performance. It outperforms existing methods by improving packet reception rate and energy efficiency consistently across diverse network conditions, while maintaining balanced throughput for both confirmed and unconfirmed messages.