Long Range Wide Area Network (LoRaWAN) is a promising communication technology for environmental monitoring due to its low power consumption and long-range capabilities. Despite its advantages, several challenges are associated with LoRaWAN due to technical limitations, environmental factors, and operational complexities. Continued advancements in adaptive algorithms and AI-based optimization are essential for overcoming these challenges and fully realizing the potential of LoRaWAN in diverse IoT applications. Transmission parameter allocation is one of the most studied aspects of LoRaWAN, typically required to reduce energy consumption and improve the signal quality in dense LoRaWAN. Evaluations of the optimization algorithms for parameter allocation are usually done using simulators. However, they do not imitate the dynamic nature of the network environment and other signal characteristics. Thus, it becomes difficult to understand the performance of these algorithms when deployed on real devices. This paper introduces transmission parameter allocation strategies using a State–Action–Reward–State–Action (SARSA) and Deep Q-Learning Network (DQN) based Reinforcement Learning (RL)-based scheduling algorithm for allocating transmission parameters in LoRaWAN communication. We experimentally evaluate this algorithm in a water quality monitoring system using actual LoRaWAN devices to assess Signal to Noise ratio (SNR), Received Signal Strength Indicator (RSSI), Time on Air (ToA), and power consumption of our RL-based algorithms with the default and Adaptive Data Rate (ADR) in LoRaWAN communication.

Beyond 5G and future 6G aim to address rising energy demands as the world becomes more interconnected. LoRaWAN network is an energy-efficient IoT solution, but frequent retransmissions can quickly deplete sensor batteries. Efficient traffic management and collision avoidance are crucial. Initially, LoRaWAN used ALOHA for multi-access, causing increased network collisions. The recent innovation of Channel Activity Detection (CAD) has emerged to tackle these issues. CAD enhances multi-access by sensing channel activity before transmission. Although an improvement, CAD is not foolproof. Our paper introduces enhancements to CAD through the Deep Deterministic Policy Gradient-based Algorithm (DDPG_CAD). To assess CAD functionality, we develop LoRaCAD, a dedicated simulator. We also conduct a thorough comparative analysis of scheduling strategies, considering energy efficiency, latency, and packet delivery ratio.

The Internet of Things (IoT) is a burgeoning field for study and experimentation. It allows users to create and receive a wide bevy of information from a massive array of devices. But as the IoT network gets denser, a load-balancing algorithm is required to keep itself running smoothly. Load Balancing is all but required in large IoT networks to avoid a part of servers getting overloaded and others being free. Existing solutions show both heuristic and machine learning algorithms designed for load balancing. Static algorithms go well with traditional IoT networks but are not built to scale dynamically and respond to loads. Dynamic heterogeneity and massive IoT disrupt load balancing. Machine learning-based algorithms have proven to give better scheduling solutions and improve performance in such networks. To analyze the performance of machine learning algorithms over heuristic ones, we designed an experimental testbed using a POX SDN controller and Mininet. We also show results confirming that the machine learning-based algorithms are better in terms of packet loss and response time.

Many tech stars like Netflix, Amazon, PayPal, eBay, and Twitter are evolving from monolithic to a microservice (containerization) architecture due to the benefits for Agile and DevOps teams. Microservices architecture can be applied to multiple industries, like IoT, using containerization. Since the IoT industry has exponential growth, universities' responsibility is to teach IoT with hands-on labs to minimize the gap between what the students learn and what is on-demand in the job market. There are many approaches in the containerization field, but they can be challenging to use without depth knowledge in virtualization and code encapsulation. After a deep analysis of the containerization challenges, in this paper, we present a cyberinfrastructure based on containers to solve the virtualization and code-encapsulation problems. The cyberinfrastructure will provide the necessary tools for data collection and code development and testing using an IoT Cluster. It is a web-based platform that allows users to securely go into containerization without spending time learning virtualization. Results show that our proposed cyberinfrastructure allows the creation and deployment of microservices in multiple IoT devices and ensures easy data collection for posterior cyberanalysis.

Quantum communications are gaining momentum in finding applications in a wide range of domains, especially those require high-security data transmissions. On the other hand, machine learning has achieved numerous breakthrough successes in various application domains including networking. However, currently, machine learning is not as much utilized in quantum networking as in other areas. With such motivation, we propose a machine-learning-powered entanglement routing scheme for quantum networks that aims to accommodate maximum numbers of demands (source-destination pairs) within a time window. More specifically, we present a deep reinforcement routing scheme that is called Deep Quantum Routing Agent (DQRA). In short, DQRA utilizes an empirically designed deep neural network that observes the current network states to schedule the network’s demands which are then routed by a qubit-preserved shortest path algorithm. DQRA is trained towards the goal of maximizing the number of resolved requests in each routing window by using an explicitly designed reward function. Our experiment study shows that, on averse, DQRA is able to maintain a rate of successfully routed requests above 80% in a qubit-limited grid network, and about 60% in extreme conditions i.e. each node can act as a repeater exactly once within a window. Furthermore, we show that the complexity and the computational time of DQRA are polynomial in terms of the sizes of the quantum networks.

Long Range Wide Area Network (LoRaWAN) is suitable for wide area sensor networks due to its low cost, long range, and low energy consumption. A device can transmit without interference if it chooses a unique channel, spread factor, transmission power different than any other transmitting device in network. However, in a dense network, the probability of interference increases because number of devices exceeds the total number of unique choices thus mandating retransmission after collision until successfully transmitted. Eventually, energy consumption of devices increases. In this poster, we present a Deep deterministic policy gradient reinforcement learning-based scheduling algorithm to improve energy efficiency by collision avoidance in a dense LoRaWAN network. We support our proposition with evaluation results for reducing energy consumption.

Nowadays, in almost every computer system, log files are used to keep records of occurring events. Those log files are then used for analyzing and debugging system failures. Due to this important utility, researchers have worked on finding fast and efficient ways to detect anomalies in a computer system by analyzing its log records. Research in log-based anomaly detection can be divided into two main categories: batch log-based anomaly detection and streaming log- based anomaly detection. Batch log-based anomaly detection is computationally heavy and does not allow us to instantaneously detect anomalies. On the other hand, streaming anomaly detection allows for immediate alert. However, current streaming approaches are mainly supervised. In this work, we propose a fully unsupervised framework which can detect anomalies in real time. We test our framework on hdfs log files and successfully detect anomalies with an F-1 score of 83%.

Energy efficiency is the essential consideration in advances of wireless sensor networks (WSNs) usefully designed for low data rate, low power consumption, and low-cost networking. Mindful of the above constraints, selecting an appropriate routing protocol can significantly improve overall performance with limited sensor network resources, especially energy awareness in WSNs. We propose an energy-aware routing protocol improved by our adaptive-gossip algorithm that will reduce the redundant routing messages so that it can minimize the overall energy consumption of a WSN. To analyze the energy efficiency, we introduce two energy performance metrics: Delay*Energy and Normalized-RoutingLoad*Energy. These metrics suggest that energy consumption heavily depends on both packet losses and packet delivery successes that affect delay and routing overhead respectively. We present both analytical and experimental results thoroughly to evaluate our adaptive-gossip proposal, and demonstrate its advantages over flooding and static-gossiping based protocols for densely deployed networks and different types of network such as peer-to-peer and multi-to-one.