Showing 1–5 of 5 results for author: Duran, K

Q-CSM: Q-Learning-based Cognitive Service Management in Heterogeneous IoT Networks

Authors: Kubra Duran, Mehmet Ozdem, Kerem Gursu, Berk Canberk

Abstract: The dramatic increase in the number of smart services and their diversity poses a significant challenge in Internet of Things (IoT) networks: heterogeneity. This causes significant quality of service (QoS) degradation in IoT networks. In addition, the constraints of IoT devices in terms of computational capability and energy resources add extra complexity to this. However, the current studies rema… ▽ More The dramatic increase in the number of smart services and their diversity poses a significant challenge in Internet of Things (IoT) networks: heterogeneity. This causes significant quality of service (QoS) degradation in IoT networks. In addition, the constraints of IoT devices in terms of computational capability and energy resources add extra complexity to this. However, the current studies remain insufficient to solve this problem due to the lack of cognitive action recommendations. Therefore, we propose a Q-learning-based Cognitive Service Management framework called Q-CSM. In this framework, we first design an IoT Agent Manager to handle the heterogeneity in data formats. After that, we design a Q-learning-based recommendation engine to optimize the devices' lifetime according to the predicted QoS behaviour of the changing IoT network scenarios. We apply the proposed cognitive management to a smart city scenario consisting of three specific services: wind turbines, solar panels, and transportation systems. We note that our proposed cognitive method achieves 38.7% faster response time to the dynamical IoT changes in topology. Furthermore, the proposed framework achieves 19.8% longer lifetime on average for constrained IoT devices thanks to its Q-learning-based cognitive decision capability. In addition, we explore the most successive learning rate value in the Q-learning run through the exploration and exploitation phases. △ Less

Submitted 21 November, 2024; originally announced November 2024.

Generative AI-enabled Digital Twins for 6G-enhanced Smart Cities

Authors: Kubra Duran, Lal Verda Cakir, Mehmet Ozdem, Kerem Gursu, Berk Canberk

Abstract: 6G networks are envisioned to enable a wide range of applications, such as autonomous vehicles and smart cities. However, this rapid expansion of network topologies makes the management of 6G wireless networks more complex and leads to performance degradation. Even though state-of-the-art applications on network services are providing promising results, they also risk disrupting the network's perf… ▽ More 6G networks are envisioned to enable a wide range of applications, such as autonomous vehicles and smart cities. However, this rapid expansion of network topologies makes the management of 6G wireless networks more complex and leads to performance degradation. Even though state-of-the-art applications on network services are providing promising results, they also risk disrupting the network's performance. To overcome this, the services have to leverage what-if implementations covering a variety of scenarios. At this point, traditional simulations fall short of encompassing the dynamism and complexity of a physical network. To overcome these challenges, we propose the Generative AI-based Digital Twins. For this, we derive an optimization formula to differentiate different network scenarios by considering the specific key performance indicators (KPIs) for wireless networks. Then, we fed this formula to the generative AI with the historical twins and real-time twins to start generating the desired topologies. To evaluate the performance, we implement network and smart-city-oriented services, namely massive connectivity, tiny instant communication, right-time synchronization, and truck path routes. The simulation results reveal that our approach can achieve 38% more stable network throughput in high device density scenarios. Furthermore, the generated scenario accuracy is able to reach up to 98% level, surpassing the baselines. △ Less

Submitted 21 November, 2024; originally announced November 2024.

DTRAN: A Special Use Case of RAN Optimization using Digital Twin

Authors: Caglar Tunc, Kubra Duran, Buse Bilgin, Gokhan Kalem, Berk Canberk

Abstract: The emergence of beyond 5G (B5G) and 6G networks underscores the critical role of advanced computer-aided tools, such as network digital twins (DTs), in fostering autonomous networks and ubiquitous intelligence. Existing solutions in the DT domain primarily aim to model and automate specific tasks within the network lifecycle, which lack flexibility and adaptability for fully autonomous design and… ▽ More The emergence of beyond 5G (B5G) and 6G networks underscores the critical role of advanced computer-aided tools, such as network digital twins (DTs), in fostering autonomous networks and ubiquitous intelligence. Existing solutions in the DT domain primarily aim to model and automate specific tasks within the network lifecycle, which lack flexibility and adaptability for fully autonomous design and management. Unlike the existing DT approaches, we propose RAN optimization using the Digital Twin (DTRAN) framework that follows a holistic approach from core to edge networks. The proposed DTRAN framework enables real-time data management and communication with the physical network, which provides a more accurate and detailed digital replica than the existing approaches. We outline the main building blocks of the DTRAN and describe the details of our specific use case, which is RAN configuration optimization, to demonstrate the applicability of the proposed framework for a real-world scenario. △ Less

Submitted 2 September, 2024; originally announced September 2024.

AI in Energy Digital Twining: A Reinforcement Learning-based Adaptive Digital Twin Model for Green Cities

Authors: Lal Verda Cakir, Kubra Duran, Craig Thomson, Matthew Broadbent, Berk Canberk

Abstract: Digital Twins (DT) have become crucial to achieve sustainable and effective smart urban solutions. However, current DT modelling techniques cannot support the dynamicity of these smart city environments. This is caused by the lack of right-time data capturing in traditional approaches, resulting in inaccurate modelling and high resource and energy consumption challenges. To fill this gap, we explo… ▽ More Digital Twins (DT) have become crucial to achieve sustainable and effective smart urban solutions. However, current DT modelling techniques cannot support the dynamicity of these smart city environments. This is caused by the lack of right-time data capturing in traditional approaches, resulting in inaccurate modelling and high resource and energy consumption challenges. To fill this gap, we explore spatiotemporal graphs and propose the Reinforcement Learning-based Adaptive Twining (RL-AT) mechanism with Deep Q Networks (DQN). By doing so, our study contributes to advancing Green Cities and showcases tangible benefits in accuracy, synchronisation, resource optimization, and energy efficiency. As a result, we note the spatiotemporal graphs are able to offer a consistent accuracy and 55% higher querying performance when implemented using graph databases. In addition, our model demonstrates right-time data capturing with 20% lower overhead and 25% lower energy consumption. △ Less

Submitted 28 January, 2024; originally announced January 2024.

Digital Twin-Native AI-Driven Service Architecture for Industrial Networks

Authors: Kubra Duran, Matthew Broadbent, Gokhan Yurdakul, Berk Canberk

Abstract: The dramatic increase in the connectivity demand results in an excessive amount of Internet of Things (IoT) sensors. To meet the management needs of these large-scale networks, such as accurate monitoring and learning capabilities, Digital Twin (DT) is the key enabler. However, current attempts regarding DT implementations remain insufficient due to the perpetual connectivity requirements of IoT n… ▽ More The dramatic increase in the connectivity demand results in an excessive amount of Internet of Things (IoT) sensors. To meet the management needs of these large-scale networks, such as accurate monitoring and learning capabilities, Digital Twin (DT) is the key enabler. However, current attempts regarding DT implementations remain insufficient due to the perpetual connectivity requirements of IoT networks. Furthermore, the sensor data streaming in IoT networks cause higher processing time than traditional methods. In addition to these, the current intelligent mechanisms cannot perform well due to the spatiotemporal changes in the implemented IoT network scenario. To handle these challenges, we propose a DT-native AI-driven service architecture in support of the concept of IoT networks. Within the proposed DT-native architecture, we implement a TCP-based data flow pipeline and a Reinforcement Learning (RL)-based learner model. We apply the proposed architecture to one of the broad concepts of IoT networks, the Internet of Vehicles (IoV). We measure the efficiency of our proposed architecture and note ~30% processing time-saving thanks to the TCP-based data flow pipeline. Moreover, we test the performance of the learner model by applying several learning rate combinations for actor and critic networks and highlight the most successive model. △ Less

Submitted 24 November, 2023; originally announced November 2023.