Figure 1 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
Large scale deployments of Internet of Things (IoT) networks are becoming reality. From a technology perspective, a lot of information related to device parameters, channel states, network and application data are stored in databases and can be used for an extensive analysis to improve the functionality of IoT systems in terms of network performanc...
Context in source publication
Context 1
... promising Low Power Wide Area Network technology is LoRaWAN [1], which, by providing a robust modulation scheme and a simple network architecture, is able to connect thousands of End Devices (EDs) to gateways in an area of several kilometers. Figure 1 sketches the LoRaWAN network architecture. The transmission by an ED is made very simple, with an adaptive tuning of the modulation parameters and without an explicit association to a specific gateway. ...
Similar publications
![Figure 2.2.: Visualisation of two position uncertainty ellipsoids [35]](profile/Sascha-Metz-3/publication/344268107/figure/fig5/AS:936377990524929@1600261318680/Visualisation-of-two-position-uncertainty-ellipsoids-35_Q320.jpg)
This work presents the application of Machine Learning to collision risk assessment for satellites. In an explainable approach, the collision risk is retrieved from predicting
the uncertainties in the estimated positions of Space Debris objects at the time of a close approach. For the first time, a dataset of collision warning messages collected by...
Citations
... In [41], the authors suggested a device profiling model based on unsupervised learning method (K-means). They grouped the packets that have the similar transmissions features in clusters. ...
... like in [2] where prediction process depends on i/p parameters of LoRa which lead to very computation overhead. While in [4], [5], [41] determining the number of clusters is very critical and depend on the application parameters used in the clustering procedure. Work in [42] share the channel information with the end node which is very critical. ...
... Because the EKF updates the parameters mapping hidden states to forecast results and freezes the weight of hidden layers, this model is substantially more stable than the online model and can adapt to new stable changes. In [41], the authors used MAE and RMSE to quantify the performance of proposed pipeline. DT outperforms LSTM in terms of IT prediction. ...
Low-power wide-area networks (LPWANs) are a new
class of wireless communications technology aimed at the Internet
of Things [29]. Long-range wide-area network (LoRa) is a promising LPWAN network standard that allows for long-distance wireless communication while conserving energy. LoRa has been used
in many applications such as healthcare, smart meters and traffic
analysis. However, LoRaWAN still faces several challenges in terms
of reliability and scalability because it has limited number of choices
for power and SF(spreading factor). These challenges such as source
allocation, collision behavior, energy consumption, communication
range and security have attracted several researchers’ attention to
focus on them. Now, machine learning has great contributions in
many fields like manufacturing, healthcare and life science that encourage the researchers to use machine learning techniques to enhance the performance of LoRa networks by proposing optimum
solutions to these challenges. This paper provides an overall survey
on using machine learning in LoRa networks and demonstrates the
recent solutions to enhance the performance of LoRa network. The
purpose of this survey is to encourage greater efforts to improve
LoRa’s performance
... The gateway carries out two key tasks; it looks into the spectrum and receives LoRa packets from end devices and forwards to the network server (NS) which determines if they are valid [55]. These network servers process the received LoRa packets further by removing those that are similar and forwards their payloads to application servers [22].Jamming is mostly done at the LoRa gateway, not at the LoRa node because the end node does not transmit signals, it just receives. In both the deployments above, LoRaWAN is able to connect to other devices within a network in two ways: Over The Air Activation (OTAA) or Activation By Personalization (ABP). ...
LoRaWAN technology has revolutionized Internet of Things (IoT) devices with
its outstanding long range, low power capabilities. However, security is still a
major concern as observed in emerging research, as inaccuracies are still evident
in LoRa’s on-air transmissions. We therefore explored LoRaWAN protocols’ susceptibility to packet collision and reactive jamming using collected datasets from
an experiment. We analyzed LoRa metadata through empirical research and
machine learning methods to investigate whether a single signal transmission is
normal or has been tampered with. Different metrics for detecting attacks, such
as Packet Delivery Ratio (PDR), Packet Inter-Arrival Time (IAT) and, comparing
modem settings at the transceiver and at the gateway were employed to give an
overview of how these attacks affected the system. Additionally, binary classifiers
were implemented using selected metadata to detect anomalous changes and distinguish collision attacks from jamming attacks. We obtained an accuracy score
of 93% using eXtreme Gradient Boosting (XGBoost) algorithm from the prediction model and found that the Signal-To-Noise ratio yielded better results in
the analysis than other metrics and parameters. Overall, it was concluded that
increased profiling of LoRa metadata was paramount in understanding LoRa’s
security overview and gaining knowledge on possible countermeasures for Denial-of-Service (DoS) attacks to adopt in this ever-growing, ideal IoT platform.
... The latest and trending approach for suitable spreading factors allocation is Machine Learning (ML) Algorithm based SFs Allocation [26][27][28][29][30][31]. In this approach of SF allocation, researchers use different Machine Learning algorithms (like K -mean, tree-based, reinforcement, deep-reinforcement, long-short term memory neural networks, decision tree, and many more) to assign SFs to the end users optimally. ...
... 1. K-mean clustering K -mean is unsupervised learning. To enhance the system performance, in reported work [26,27,31], the authors adopted the K-mean machine learning algorithm to allocate the suitable SFs to the end devices. 2. Reinforcement learning The reinforcement learning approach is preferable for SF allocation in a dynamic network (where the system parameters change rapidly). ...
... In [31], the authors have discussed K-mean, Long-Short term Memory Neural Networks, and decision tree to improve the LoRaWAN network performance. One of the challenges to having a Machine Learning based analysis study is the data collection. ...
LoRaWAN is one of the most suitable communication protocols for the IoT applications that require low power over long-range communication. However, the LoRa network suffers from scalability, low data rate, and other performance issues that significantly affect the network performance. The study of the optimal spreading factor allocation can overcome these issues and help to improve the network performance. Hence, this article puts forward the state-of-the-art literature review on the Spreading Factors Allocation schemes for the LoRaWAN. Industry and academia have done an extensive research to address the issues related to optimal resource allocation, like spreading factor allocation to the spatially distributed end-devices of the network. Most of the problems concerning spreading factor allocation are being explored and resolved. Therefore, this paper reviews and compares various spreading factor allocation schemes proposed by the researchers. Furthermore, we provide a summary of the different review studies of the LoRaWAN. The literature presented in this paper motivates researchers to examine other aspects of spreading factor allocation schemes to improve the LoRa network performance.
... The most pressing implementation problems are the effective integration of various sensing technologies, the creation of an appropriate network infrastructure, population education, and the examination of sustainability issues such as carbon footprint, among others. [1] [14] [17] Currently, governmental and public agencies are taking more measures in connection to trash management in order to enhance the collection and sensible disposal of garbage created by a city. These are already taking into account the rapid speed of global urbanization and the growth of the industrial sector, as well as the healthcare and manufacturing industries, which are anticipated to create substantial amounts of trash and may be efficiently managed by smart management. ...
Urbanization reached severe dimensions as a result of rising population density and rural flight to cities, posing a massive waste generating problem in cities. Increased garbage creation has long been seen as a serious burden for large metropolitan centers across the world, and it is a vital issue for nations experiencing rapid urbanization. The Internet of Things (IoT) and machine learning (ML) offer an automotive possibility through cyber-bodily systems, which revolutionized solid waste management. This study conducts a comprehensive analysis of different framework for waste management models available in the existing works based on infrastructure of IoT for effective treatment of trash generated in municipal situations, taking into account IoT needs. The relationship between concessionaires and trash generators (citizens) is being studied in order to minimize collection time and costs while also promoting citizenship. The objective of this paper is to emphasize about the most important methods and categorize outstanding research challenges on the topic by describing an IoT-based reference model and analyzing the available solutions.
... This implies that devices may not occupy the ISM band for more than 36 seconds per hour, forbidding the transmission of new packets when this limit is attained [123]. Machine learning can be applied to model and analyze technical problems, improving the scalability of LoRa networks and predicting network congestion [124]. Further developments could include enhanced ADR mechanisms, optimization of GW locations, and interference cancellation techniques [125]. ...
This paper provides a survey on the adoption of LoRa in the agricultural field, and reviews state-of-the-art solutions for Smart Agriculture, analyzing the potential of this technology in different infield applications. In particular, we consider four reference scenarios, namely irrigation systems, plantation and crop monitoring, tree monitoring, and livestock monitoring, which exhibit heterogeneous requirements in terms of network bandwidth, density, sensors’ complexity, and energy demand, as well as latency in the decision process. We discuss how LoRa-based solutions can work in these scenarios, analyzing their scalability, interoperability, network architecture, and energy-efficiency. Finally, we present possible future research directions and point out some open issues which might become the main research trends for the next years.
Indoor air quality (IAQ) is significantly impacted by ambient air pollution, and assessing and understanding pollution causes is crucial for developing mitigation plans. The prospect of a "smart environment" using IoT is being explored to improve indoor living conditions. Technological advancements have simplified the design of ambient assisted living systems. This review explores how IoT can regulate indoor air quality, focusing on models of dispersion, land use, sensors, and their coupling with cellphones, as well as air pollution simulations. The review discusses smart ventilation techniques like RIVEC and exhaust-only DCV systems, as well as computational tools for accurate health risk assessment. It covers challenges like data coding and safety concerns, and future prospects like deep learning models and mobile app-based systems for predictions. The report also includes literature indexed using top databases, concentrating on replicating actual circumstances and real-time IAQ monitoring devices.
Keywords: Internet of Things, technological advancements, smart ventilation techniques, Indoor air quality, deep learning models.
Maize flour obtained from the dried corn is one of the most consumed foods in Rwanda. It is imperative that this should be healthy and risk-free for a safe consumption. Therefore, it is vital to keep track of the environmental conditions during the drying process and the characteristics that exist inside maize storage containers. In Rwanda, traditional methods are most commonly used by maize farmers for drying and storage purposes, where no smart system is being used to monitor the environmental conditions under which the maize grains are dried and stored. This mostly affects the quality of maize and flour being produced which will finally affect food security. In this research, temperature, humidity, and light sensors are deployed in the grain storage containers for environmental parameter detection purposes to achieve the primary goal of providing practical, secure, and easily accessible storage in inclement weather. Temperature and humidity are two factors that have an impact on grain quality while in storage. The ThingSpeak platform has been used to help farmers monitor the drying and storing conditions of the maize on a real-time basis. A global system for mobile (GSM) communication module is used to notify farmers by sending a short message in case of critical drying or storing environmental parameters under which the maize grains are stored. The result is shown in the form of humidity, temperature, and light graphs which are displayed on the ThingSpeak platform in real-time mode.
The Internet of Things is rapidly growing with the demand for low-power, long-range wireless communication technologies. Long Range Wide Area Network (LoRaWAN) is one such technology that has gained significant attention in recent years due to its ability to provide long-range communication with low power consumption. One of the main issues in LoRaWAN is the efficient utilization of radio resources (e.g., spreading factor and transmission power) by the end devices. To solve the resource allocation issue, machine learning (ML) methods have been used to improve the LoRaWAN network performance. The primary aim of this survey paper is to study and examine the issue of resource management in LoRaWAN that has been resolved through state-of-the-art ML methods. Further, this survey presents the publicly available LoRaWAN frameworks that could be utilized for dataset collection, discusses the required features for efficient resource management with suggested ML methods, and highlights the existing publicly available datasets. The survey also explores and evaluates the Network Simulator-3-based ML frameworks that can be leveraged for efficient resource management. Finally, future recommendations regarding the applicability of the ML applications for resource management in LoRaWAN are illustrated, providing a comprehensive guide for researchers and practitioners interested in applying ML to improve the performance of the LoRaWAN network.
Adaptive data rate (ADR) is a widely adopted resource assignment approach in long-range wide-area networks (LoRaWANs) for static Internet of Things (IoT) applications such as smart grids and metering. Blind ADR (BADR) has been recommended for mobile IoT applications such as pet and industrial asset tracking. However, ADR and BADR cannot provide appropriate measures to alleviate the massive packet loss problem caused by the unsuitable spreading factors (SFs) assigned to end devices when they are mobile. This paper proposes a novel proactive approach---``artificial intelligence-empowered resource allocation’’ (AI-ERA)---to address the resource assignment issue in static and mobile IoT applications. The AI-ERA approach consists of two modes, namely offline and online modes. First, a deep neural network (DNN) model is trained with a dataset generated at ns-3 in the offline mode. Second, the proposed AI-ERA approach utilizes the pre-trained DNN model in the online mode to proactively assign an efficient SF for the end device before each uplink packet transmission. The proactive behavior of the AI-ERA improved the packet success ratio by an average of 32% and 28% in static and mobility scenarios compared with the typical LoRaWAN ADR, respectively.
