To enable the large scale and efficient deployment of Artificial Intelligence (AI), confluence AI Edge Computing has given rise to Intelligence, which leverages on computation communication capabilities end devices edge servers process data closer where it is produced. One enabling technologies privacy preserving machine learning paradigm known as Federated Learning (FL), enables owners conduct model training without having transmit their raw third-party servers. However, FL network...

Federated Learning (FL) is a promising privacy-preserving distributed machine learning paradigm. However, communication inefficiency remains the key bottleneck that impedes its large-scale implementation. Recently, hierarchical FL (HFL) has been proposed in which data owners, i.e., workers, can first transmit their updated model parameters to edge servers for intermediate aggregation. This reduces instances of global and straggling workers. To enable efficient HFL, it important address...

Dubbed as the next-generation Internet, meta-verse is a virtual world that allows users to interact with each other or objects in real-time using their avatars. The metaverse envisioned support novel ecosystems of service provision an immersive environment brought about by intersection and physical worlds. native AI systems will personalized user experience over time shape scalable, seamless, synchronous way. However, characterized diverse resource types amid highly dynamic demand...

Due to the advanced capabilities of Internet Vehicles (IoV) components such as vehicles, Roadside Units (RSUs) and smart devices well increasing amount data generated, Federated Learning (FL) becomes a promising tool given that it enables privacy-preserving machine learning can be implemented in IoV. However, performance FL suffers from failure communication links missing nodes, especially when continuous exchanges model parameters are required. Therefore, we propose use Unmanned Aerial...

Distributed computing has become a common approach for large-scale computation tasks due to benefits such as high reliability, scalability, speed, and cost-effectiveness. However, distributed faces critical issues related communication load straggler effects. In particular, nodes need exchange intermediate results with each other in order calculate the final result, this significantly increases overheads. Furthermore, network may include straggling that run intermittently slower. This longer...

Federated Learning (FL) is a privacy-preserving collaborative learning approach that trains artificial intelligence (AI) models without revealing local datasets of the FL workers. While ensures privacy workers, its performance limited by several bottlenecks, which become significant given increasing amounts data generated and size network. One main challenges straggler effects where computation delays are caused slow As such, Coded (CFL), leverages coding techniques to introduce redundant...

Amid growing concerns on data privacy, Federated Learning (FL) has emerged as a promising privacy preserving distributed machine learning paradigm. Given that the FL network is expected to be implemented at scale, several studies have proposed system architectures towards improving scalability and efficiency. Specifically, Hierarchical (HFL) utilizes cluster heads, e.g., base stations, for intermediate aggregation relay of model parameters. Serverless also recently, in which owners, i.e.,...

In 5G and Beyond networks, Artificial Intelligence applications are expected to be increasingly ubiquitous. This necessitates a paradigm shift from the current cloud-centric model training approach Edge Computing based collaborative learning scheme known as edge learning, in which is executed at of network. this article, we first introduce principles technologies learning. Then, establish that successful, scalable implementation requires communication, caching, computation, resources (3C-L)...

Distributed computing has become a common approach for large-scale computation of tasks due to benefits such as high reliability, scalability, speed, and costeffectiveness. However, distributed faces critical issues related communication load straggler effects. In particular, nodes need exchange intermediate results with each other in order calculate the final result, this significantly increases overheads. Furthermore, network may include straggling that run intermittently slower. This...

Federated Edge Learning (FEL) is a distributed Machine (ML) framework for collaborative training on edge devices. FEL improves data privacy over traditional centralized ML model by keeping the devices and only sending local updates to central coordinator aggregation. However, challenges still remain in existing architectures where there high communication overhead between coordinator. In this paper, we present working prototype of blockchain-empowered communication-efficient framework, which...

Due to the advanced capabilities of Internet Vehicles (IoV) components such as vehicles, Roadside Units (RSUs) and smart devices well increasing amount data generated, Federated Learning (FL) becomes a promising tool given that it enables privacy-preserving machine learning. However, performance FL suffers from failure communication links missing nodes. Therefore, we propose use Unmanned Aerial (UAVs) wireless relays facilitate communications between IoV server thus improving accuracy FL....

One of the enabling technologies Edge Intelligence is privacy preserving machine learning paradigm called Federated Learning (FL). However, communication inefficiency remains a key bottleneck in FL. To reduce node failures and device dropouts, Hierarchical (HFL) framework has been proposed whereby cluster heads are designated to support data owners through intermediate model aggregation. This decentralized approach reduces reliance on central controller, e.g., owner. issues resource...

Coded distributed computing (CDC) has emerged as a promising approach because it enables computation tasks to be carried out in manner while mitigating straggler effects, which often account for the long overall completion times. Specifically, by using polynomial codes, computed results from only subset of edge servers can used reconstruct final result. However, incentive issues have not been studied systematically complete CDC tasks. In this paper, we propose tractable two-level...

As the amount of data collected for crowdsensing applications increases rapidly due to improved sensing capabilities and increasing number Internet Things (IoT) devices, cloud server is no longer able handle large-scale datasets individually. Given computational edge coded distributed computing has become a promising approach given that it allows computation tasks be carried out in manner while mitigating straggler effects, which often account long overall completion times. Specifically, by...

Today, modern unmanned aerial vehicles (UAVs) are equipped with increasingly advanced capabilities that can run applications enabled by machine learning techniques, which require computationally intensive operations such as matrix multiplications. Due to computation constraints, the UAVscan offload their tasks edge servers. To mitigate stragglers, coded distributed computing (CDC) based offloading be adopted. In this paper, we propose an Optimal Task Allocation Scheme (OTAS) on Stochastic...

Amid data privacy concerns, Federated Learning (FL) has emerged as a promising machine learning paradigm that enables privacy-preserving collaborative model training. However, there exists need for platform matches owners (supply) with requesters (demand). In this paper, we present CrowdFL, to facilitate the crowdsourcing of FL It coordinates client selection, training, and reputation management, which are essential steps operations. By implementing training on actual mobile devices,...

With the ubiquitous sensing enabled by Internet-of-Things (IoT), massive amount of data is generated every second, transforming way we interact with world. To manage big and enable analytics at edge network, large computation power required to perform intensive tasks. However, energy-constrained IoT devices are not able tasks without compromising quality-of-service applications. In this paper, propose a hybrid network in which users can offload their servers through coded offloading or local...

Parkinson's disease (PD) is a chronic with high risk of incidence after the age 60 and problem for many countries facing an aging population. Current works have mainly focused on supervised learning using data collected from various sensors to differentiate between PD healthy subjects. However, such methods are not ideal prognosis where there no labels (i.e., we do know in advance which subjects will develop future). We propose tackle as semi-supervised anomaly detection task, model...

The COVID-19 pandemic is one of the most severe challenges world faces today. In order to contain transmission COVID-19, people around have been advised practise social distancing. However, maintaining distance a challenging problem, as we often do not know beforehand how crowded places intend visit are. this paper, demonstrate crowded.sg, an AI-empowered platform that leverages on Unmanned Aerial Vehicles (UAVs), crowdsourced images, and computer vision techniques provide distancing...

Federated Learning (FL) is a promising privacy-preserving distributed machine learning paradigm. However, communication inefficiency remains the key bottleneck that impedes its large-scale implementation. Recently, hierarchical FL (HFL) has been proposed in which data owners, i.e., workers, can first transmit their updated model parameters to edge servers for intermediate aggregation. This reduces instances of global and straggling workers. To enable efficient HFL, it important address...

Edge computing has been a promising approach that allows resource-constrained Internet-of-Things (IoT) devices to offload their computation-intensive tasks the edge of networks. Instead offloading entire computation task single device, IoT can leverage on resources multiple devices. In order improve performance distributed tasks, coded is proposed mitigate straggler effects. However, may not have sufficient bandwidth transmit computed results reliably. Hence, we present deep-learning based...

The COVID-19 pandemic has disrupted the lives of millions across globe. In Singapore, promoting safe distancing by managing crowds in public areas have been cornerstone containing community spread virus. One most important solutions to maintain social is monitor crowdedness indoor and outdoor points interest. Using Nanyang Technological University (NTU) as a testbed, we develop deploy platform that provides live predicted crowd counts for key locations on campus help users plan their trips...

Dubbed as the next-generation Internet, metaverse is a virtual world that allows users to interact with each other or objects in real-time using their avatars. The envisioned support novel ecosystems of service provision an immersive environment brought about by intersection and physical worlds. native AI systems will personalized user experience over time shape scalable, seamless, synchronous way. However, characterized diverse resource types amid highly dynamic demand environment. In this...

Leveraging on the wealth of data and advancements in Artificial Intelligence, smart cities have demonstrated their great potential providing solutions to challenges that urban population faces today. However, as becomes more privacy sensitive with introduction stringent regulations, differential-private FL (DPFL) is a promising technology can enable privacy-preserving collaborative model training. In this paper, we consider an network owners heterogeneous budgets preferences respectively....