Adaptive priority-aware LoRaWAN resource allocation for Internet of Things applications

Abstract

LoRaWAN is the most adopted communication technology for the Internet of Things (IoT) applications for enabling flexible, cost-effective long-range communication with low power consumption. The deployment of a wide range of IoT devices brings many benefits to many applications, but new challenges arise from the network point of view with this deployment. For instance, IoT applications have different requirements in terms of Quality of Service (QoS) and energy efficiency in dense IoT scenarios. In this context, it is crucial to design an efficient resource allocation mechanism that enables real-time adjustments of radio-related parameters to improve the system’s scalability while providing QoS and reducing IoT devices’ energy consumption depending on the application requirements. This article proposes an adaptive priority-aware resource allocation mechanism to improve LoRaWAN scalability and energy consumption in a dense IoT scenario, called APRA. The simulation results show the benefits of APRA in improving the energy consumption by 95% and increasing the battery discharge time of the end device up to 5 years while yielding high packet delivery and low delay to high priority applications.

Introduction

Internet of Things (IoT) is expanding its portfolio to include a wide range of IoT applications, mainly due to the advances in different areas, such as embedded systems, microelectronics, communication, and sensing [1], [2]. As a result, the number of 5G-connected IoT devices might reach 4.1 billion by 2024 [3]. In this context, IoT has been receiving much attention from both academia and industry due to its potential for use in the most diverse areas of human activity, such as smart cities, healthcare, agriculture, environmental monitoring, logistics, home/building automation, smart grid, critical infrastructure monitoring, and several other application areas [4], [5], [6].

IoT applications require low energy consumption (to address 10-years battery discharge time), high coverage, and cost-effectiveness [7], [8], [9]. So the communication technology used to transmit the collected IoT data plays a vital role in the massive adoption and deployment of IoT applications. To satisfy IoT applications’ requirements, the Low-Power Wide-Area Network (LPWAN) emerged as a promising communication technology for supporting a wide variety of IoT applications in rural and urban areas [10]. Over the last few years, LPWAN has been increasingly used on a large scale by the industry. Recent market data shows an increase of 109% per year of connected LPWAN devices and an annual investment of more than US$4.5 billion from 2018 to 2023 [11].

Industry and research communities support different LPWAN technologies, such as LoRaWAN, SigFox, and NB-IoT [12]. Instead of common cellular infrastructures such as 3G and 4G, LPWAN solutions implement a communication technology with lower operating costs, low bit rate, long-range, and low energy consumption [13]. According to Haxhibeqiri et al. [7], the number of publications on Long Range (LoRa) areas has grown tremendously in the past years. In this context, LoRa is a wireless communication technology under patent by Semtech. LoRa refers to the radio modulation and the devices that use this modulation type. LoRa enables the devices to transmit over distances up to hundreds of kilometers. On the other hand, LoRa Alliance [14] considers LoRa radio as the physical layer and defines the upper layers and network architecture of LoRaWAN, i.e., open network protocol and ecosystem [15], [16]. In this way, LoRaWAN offers a cost-effective way to enable a large-scale deployment of End Devices (ED) that require less-complex medium access control mechanisms at the expense of low throughput [17].

Future LoRaWAN-based scenarios, especially in urban areas, will be composed of several thousand IoT devices per square meter. However, the densification of LoRaWAN generates a severe problem when more connected devices coexist in the same area with limited radio resources [18]. This issue significantly impacts the number of packets lost due to collision and interference thereby affecting network scalability and efficiency [19]. For instance, LoRaWAN Gateway (GW) might receive messages from many EDs on the same channel, leading to interference and poor application performance. In this context, the LoRaWAN physical layer considers a set of radio parameters that can be adjusted on-the-fly to provide a trade-off among transmission range, bit rate, airtime, energy consumption, and interference [18], [20]. Existing works have demonstrated that an efficient combination of these radio-related parameters configured by a resource allocation mechanism has a significant impact on the IoT applications, resulting in better coverage, data delivery, and robustness with lower energy consumption [21], [22].

The LoRaWAN architecture allows the deployment of many IoT devices and applications to coexist in the same physical space. In this context, there is a high heterogeneity of IoT applications in terms of Quality of Service (QoS) requirements, such as delay and reliability [23]. This scenario leads to new challenges in the design an efficient resource allocation mechanisms, which must consider the application priority to configure radio-related parameters to efficiently satisfy the required QoS for each IoT application cost-effectively. Hence, it is essential to consider key application requirements individually rather than a one-size-fits-all solution. It is important to advance the state-of-the-art with a resource allocation mechanism that can dynamically adjust LoRaWAN radio-related parameters based on the application requirements and radio conditions. As a result, the system will be able to use the available channels efficiently while minimizing both the number of collisions and the energy consumption.

This article proposes an Adaptive Priority-aware Resource Allocation mechanism called APRA. It aims to improve LoRaWAN scalability, efficiency, and energy consumption by selecting radio-related parameters on-the-fly while performing a priority-aware resource control for each type of IoT application. Initially, EDs have distributed among the Spreading Factor (SF) channels according to the application priority, favoring the allocation of EDs in the lowest SFs to reduce collisions. Besides, the highest priority EDs are configured with a higher Bandwidth (BW), if possible. Finally, to reduce energy consumption, the algorithm performs an optimal Transmission Power (TP) selection. Simulation results obtained demonstrate APRA’s efficiency in Packet Delivery Ratio (PDR), Packet Error Ratio (PER), Time on Air (ToA), and energy consumption compared to recently proposed state-of-the-art algorithms. Specifically, APRA optimizes up to 71% of the network energy consumption, improves the ED’s battery discharge time by 3.5 years, delivers above 99% of packets, and reduces the high priority EDs ToA.

APRA aims at advancing the state-of-the-art in improving LoRaWAN scalability and energy consumption in dense IoT scenarios. In this context, we summarize the main contributions of this work as follows:

• We propose an optimal SF allocation scheme by considering the requirements of IoT applications to deliver smaller ToA and reduce collisions.

• We present a BW selection mechanism to enable a higher bit rate for high-priority EDs.

• We introduce a TP allocation scheme to improve the energy efficiency of the IoT system.

• We also perform simulation experiments to evaluate the impact and benefits of APRA. The results show that APRA can effectively mitigate the challenges related to specific radio-related parameters on-the-fly to improve LoRaWAN scalability, efficiency, and energy consumption.

We organize the rest of the article as follows. Section 2 outlines the state-of-the-art about LoRaWAN resource allocation for IoT applications. Section 3 describes the scenario overview and APRA algorithm. Section 4 discusses the simulation description, methodology, and results. Finally, Section 5 concludes the work and outlines future directions.