Loading...
16 results
Search Results
Now showing 1 - 10 of 16
- Spectrum sharing for LTE and 5G-NR coexistencePublication . Busari, Sherif Adeshina; Correia, NoƩlia; Saghezchi, Firooz B.; Mumtaz, Shahid; Rodriguez, JonathanSpectrum sharing provides a rapid migration pathway toward 5G by enabling the coexistence of 4G LTE and 5G new radio (NR) that share the same spectrum. Due to significant differences in the LTE and 5G-NR air interfaces, several enablers are required to facilitate the spectrum sharing. In this study, we explore the coexistence features and investigate their impacts on network performance. For static and dynamic spectrum sharing scenarios, we assess the impacts of different spectrum sharing ratios, user ratios, MIMO configurations, mixed numerology profiles and traffic patterns on the user throughput and network capacities of spectrum sharing networks, compared with the LTE only and 5G-NR only networks with exclusive spectrum access. The key results show that spectrum sharing leads to a marginal capacity gain over LTE only network and achieves considerably lower capacity than the 5G-NR only network. Also, the results show that mixed numerology profiles between the LTE and 5G-NR lead to capacity losses due to inter-numerology interference. In addition, user and spectrum sharing ratios between LTE and 5G-NR have critical impacts on performance. Reduced spectrum per device as the number of 5G devices increases, higher signaling overhead and higher scheduling complexity are other limiting factors for spectrum sharing networks. The results show limited capacity benefits and reinforce spectrum sharing between LTE and 5G-NR as mainly an evolutionary path to accommodate 5G users in the same LTE spectrum while migrating to the fully-fledged 5G networks. For significant capacity increase, other features such as carrier aggregation, overlay of small cells and higher order MIMO would need to be incorporated into the network.
- Performance evaluation of radio resource schedulers in LTE and 5G NR two-tier HetNetsPublication . Busari, Sherif Adeshina; Correia, NoƩlia; Mumtaz, Shahid; Rodriguez, Jonathan; Saghezchi, Firooz B.Network performance is critically dependent on the employed radio resource scheduler (RRS). The impact becomes even more significant in 5G ultra-dense networks due to the challenges of complicated base station distribution, user association, load balancing and inter-cell interference, among others. Using a combination of three popular schedulers (i.e., round robin (RR), proportional fairness (PF) and best channel quality indicator (BCQI)), we evaluate, in this work, the performance of two-tier heterogeneous networks where the different tiers employ the same or different RRSs. Using user throughput, cell capacity and system fairness as metrics, the results show that, on one hand, the average user throughput-system fairness tradeoff favours the use of the RR-PF combination (where the macrocell tier employs RR while the small cell tier uses PF). On the other hand, the BCQI-BCQI combination produces the highest network capacity, principally from about 5-10% of the total users, thereby sacrificing fair allocation of resources among the users. The results show that there is no globally optimal RRS combination across the metrics. As the mobile network operators have the freedom to deploy schedulers as they deem fit, the RRS combination can be selected to satisfy the performance targets of the respective use cases and deployment scenarios.
- Mobility planning of LoRa gateways for edge storage of IoT dataPublication . Carvalho, Rodrigo; Correia, NoƩlia; Al-Tam, FaroqLoRaWAN is now a leading technology in IoT developments due to its low power consumption and simple deployment features. Despite being termed a long-distance technology, its coverage is often hampered by the harsh propagation environment. In such cases, a very cost-effective solution is to use mobile gateways for both data collection and delivery at specific places of the edge, avoiding the need for cellular communication modules. This article addresses the optimization problem of deciding how mobile gateways should move, and which edge devices should be visited (and when) for collected data to be fed into the edge distributed storage, while ensuring data usefulness. This optimization problem of gateway mobility with edge storage is formulated mathematically, and a deep reinforcement learning framework is proposed to solve it. Results show that the proposed framework can achieve near optimal results, in particular for large-dense deployments and not very strict time windows. The proposed deep reinforcement learning solution is able to capture underlying patterns, like device coverage and traffic, without prior knowledge of such information.
- Optimization of mixed numerology profiles for 5G wireless communication scenariosPublication . Correia, NoĆ©lia; Al-Tam, Faroq; Rodriguez, JonathanThe management of 5G resources is a demanding task, requiring proper planning of operating numerology indexes and spectrum allocation according to current traffic needs. In addition, any reconfigurations to adapt to the current traffic pattern should be minimized to reduce signaling overhead. In this article, the pre-planning of numerology profiles is proposed to address this problem, and a mathematical optimization model for their planning is developed. The idea is to explore requirements and impairments usually present in a given wireless communication scenario to build numerology profiles and then adopt one of the profiles according to the current users/traffic pattern. The model allows the optimization of mixed numerologies in future 5G systems under any wireless communication scenario, with specific service requirements and impairments, and under any traffic scenario. Results show that, depending on the granularity of the profiles, the proposed optimization model is able to provide satisfaction levels of 60ā100%, whereas a non-optimized approach provides 40ā65%, while minimizing the total number of numerology indexes in operation.
- Deep PC-MAC: a deep reinforcement learning pointer-critic media access protocolPublication . Al-Tam, Faroq; Mazayev, Andriy; Correia, NoƩlia; Rodriguez, J.Developing artificial intelligence (AI) solutions for communication problems is one of the hottest topics nowadays. This article presents Deep PC-MAC, a novel deep reinforcement learning (DRL) solution to solve the fair coexistence problem (FCP) between heterogeneous nodes in the unlicensed bands. It is based on a hybrid architecture between pointer networks (Ptr-nets) and advantage actor-critic (A2C), i.e., pointer-critic architecture. The proposed model allows base stations to fairly share unlicensed bands with incumbent nodes. It jointly protects the incumbent nodes from spectrum starvation and improves key-performance indicators (KPIs). Deep PC-MAC is trained from scratch with zero-knowledge about FCP and experimental results demonstrate its efficiency when compared to a baseline method.
- Attention-based model and deep reinforcement learning for distribution of event processing tasksPublication . Mazayev, Andriy; Al-Tam, Faroq; Correia, NoƩliaEvent processing is the cornerstone of the dynamic and responsive Internet of Things (IoT). Recent approaches in this area are based on representational state transfer (REST) principles, which allow event processing tasks to be placed at any device that follows the same principles. However, the tasks should be properly distributed among edge devices to ensure fair resources utilization and guarantee seamless execution. This article investigates the use of deep learning to fairly distribute the tasks. An attention-based neural network model is proposed to generate efficient load balancing solutions under different scenarios. The proposed model is based on the Transformer and Pointer Network architectures, and is trained by an advantage actorcritic reinforcement learning algorithm. The model is designed to scale to the number of event processing tasks and the number of edge devices, with no need for hyperparameters re-tuning or even retraining. Extensive experimental results show that the proposed model outperforms conventional heuristics in many key performance indicators. The generic design and the obtained results show that the proposed model can potentially be applied to several other load balancing problem variations, which makes the proposal an attractive option to be used in real-world scenarios due to its scalability and efficiency.
- Resource design in federated sensor networks using RELOAD/CoAP overlay architecturesPublication . Rodrigues, Luis; Guerreiro, Joel; Correia, NoƩliaSensor networks can be federated for wide-area geographical coverage using RELOAD/CoAP architectures. In this case, proxy nodes of constrained environments form a P2P overlay to announce device resources or sensor data. Although this is a standard-based solution, consistency problems may arise because P2P resources (data objects stored at the overlay network) may end up including similar device resource entries. This is so because device resource entries, or sensor data, can be announced under different P2P resource umbrellas, meaning that any update to them will require changing multiple P2P resources. Here in this article, a multi-layer approach is proposed to solve this issue, allowing for a more efficient storage/retrieval of IoT data. Information at the overlay network is kept consistent, although additional P2P anonymous resources must be created. A mathematical model is proposed for the planning of such multi-layer organization of P2P resources, together with a heuristic algorithm. A required overlay service is also discussed.
- Radio Resource Scheduling with Deep Pointer Networks and Reinforcement LearningPublication . Al-Tam, Faroq; Mazayev, Andriy; Correia, NoƩlia; Rodriguez, J.This article presents an artificial intelligence (AI) adaptable solution to handle the radio resource scheduling (RRS) task in 5G networks. RRS is one of the core tasks in radio resource management (RRM) and aims to efficiently allocate frequency domain resources to users. The proposed solution is an advantage pointer critic (APC) deep reinforcement learning (DRL) agent. It is built with a deep pointer network architecture and trained by the policy gradient algorithm. The proposed agent is deployed in a system level simulator and the experimental results demonstrate its adaptability to network dynamics and efficiency when compared to baseline algorithms.
- Adaptive spectrum allocation for 5G wireless communication scenariosPublication . Correia, NoƩlia; Al-Tam, Faroq; Rodriguez, J.5G resources should be properly planned for users to have a good quality of service. This planning includes defining the most suitable numerology indexes and best spectrum allocation considering the requirements of current traffic, which may change over time. Furthermore, when accounting for changes in traffic pattern, any necessary reconfigurations should be minimized. Here in this article, an optimization model is developed for the planning of spectrum allocation to the best mix numerology. The model considers adapting the operating numerology mix according to the current presence (or not) of traffic requirements. The model also works under any wireless communication scenario in 5G, and under any traffic pattern.
- RELOAD/CoAP architecture for the federation of wireless sensor networksPublication . Pisco, LuĆs; Guerreiro, Joel; Correia, NoĆ©liaSensing devices are expected to interconnect over large geographical areas and federations of wireless sensor networks are expected in a near future. In such environments a critical issue is how to discover the resources available at devices in a scalable manner. For this purpose, a Constrained Application Protocol (CoAP) Usage for REsource LOcation And Discovery (RELOAD), a generic self-organizing Peer-to-Peer (P2P) overlay network service, has been defined to be used as a lookup service, to store available device resources and as a cache for sensor data. Each P2P resource, at the RELOAD/CoAP overlay, includes references to device resources, hosted at one or more constrained nodes, but no provision is made for the insertion of bindings/references to P2P resources already available at the P2P overlay network. Such feature would increase the efficiency and consistency of storage, avoiding duplicate references, a very relevant issue for future IoT applications relying on the federation of wireless sensor networks. In this article an extension to the service provided by CoAP Usage is proposed so that resource bindings can be managed. Two binding models, and a heuristic algorithm for their implementation, are proposed. Results show that such models lead to a better resource organization, reducing the number of sensor resource entries and/or fetches to the P2P overlay.