Intelligent Resource Allocation in LoRaWAN Using Machine Learning Techniques

Abstract

With the ubiquitous growth of Internet-of-things (IoT) devices, current low-power widearea network (LPWAN) technologies will inevitably face performance degradation due to congestion and interference. The rule-based approaches to assign and adapt the device parameters are insufficient in dynamic massive IoT scenarios. For example, the adaptive data rate (ADR) algorithm in LoRaWAN has been proven inefficient and outdated for large-scale IoT networks. Meanwhile, new solutions involving machine learning (ML) and reinforcement learning (RL) techniques are shown to be very effective in solving resource allocation in dense IoT networks. In this article, we propose a new concept of using two independent learning approaches for allocating spreading factor (SF) and transmission power to the devices using a combination of a decentralized and centralized approach. SF is allocated to the devices using RL for contextual bandit problem, while transmission power is assigned centrally by treating it as a supervised ML problem. We compare our approach with existing state-of-the-art algorithms, showing a significant improvement in both network level goodput and energy consumption, especially for large and highly congested networks.

Enhancement in Optimal Resource-Based Data Transmission Over LPWAN Using a Deep Adaptive Reinforcement Learning Model Aided by Novel Remora With Lotus Effect Optimization Algorithm

“…In 2023, Minhaj et al [22]. implemented a novel way of distributing the spreading factor (SF) and transmission power to the devices by combining a decentralized and centralized method with two independent learning methodologies.…”

“…However, it suffers from contextual bandit problems, and it is hard to maintain the network's stability using various network conditions. Combining reinforcement learning and supervised machine learning [22] has improved the energy efficiency, output quality, and Packet Reception Rate (PRR) of dense LoRa networks, reducing the required processing time. However, its operation necessitates a feedback system, potentially leading to uplink and downlink interference.…”

LoRaWAN Meets ML: A Survey on Enhancing Performance with Machine Learning

“…The authors in [ 129 ] solve the resource classification problem (e.g., TP) for static EDs in LoRaWAN through various ML techniques, such as RF, SVM, logistic regression (LR), K-nearest neighbor (KNN), LDA, and GNB. The authors used LoRaSim [ 30 , 130 , 131 ] network simulator for dataset collection, which is designed for LoRaWAN IoT networks based on Python.…”

“…RL approach has several advantages over traditional methods of optimizing LoRaWAN networks [ 129 , 152 , 153 , 154 , 155 ].…”

“…The allocation of SF is dependent on few parameters, such as GW sensitivity ( ) [ 27 ] and ED sensitivity ( ) [ 105 ] thresholds (illustrated in Table 9 ), the distance between the GW and ED [ 7 ], path loss [ 129 , 146 ], ToA [ 133 ], interference/collision [ 201 ], ED positions with SNR [ 147 , 148 ], retransmission [ 10 ], the ratio of UL and downlink ACK [ 105 ], and packet drop ratio [ 258 ]. However, to apply ML/DL for SF allocation to EDs, multiple parameters have been adopted in the current literature, for example, the received power (i.e., RSSI), SNR, the distance between ED and GW, and ED position [ 114 ].…”