With the breakthroughs in deep learning, recent years have witnessed a booming of artificial intelligence (AI) applications and services, spanning from personal assistant to recommendation systems video/audio surveillance. More recently, with proliferation mobile computing Internet Things (IoT), billions IoT devices are connected Internet, generating zillions bytes data at network edge. Driving by this trend, there is an urgent need push AI frontiers edge so as fully unleash potential big...

We study the capacity of multiple-input multiple- output (MIMO) relay channels. first consider Gaussian MIMO channel with fixed conditions, and derive upper bounds lower that can be obtained numerically by convex programming. present algorithms to compute bounds. Next, we generalize Rayleigh fading case. find an bound a on ergodic capacity. It is somewhat surprising meet under certain regularity conditions (not necessarily degradedness), therefore characterized exactly; previously this has...

Federated learning is an emerging machine technique that enables distributed model training using local datasets from large-scale nodes, e.g., mobile devices, but shares only updates without uploading the raw data. This provides a promising privacy preservation for devices while simultaneously ensuring high performance. The majority of existing work has focused on designing advanced algorithms with aim to achieve better However, challenges, such as incentive mechanisms participating in and...

Crowdsensing offers an efficient approach to meet the demand in large scale sensing applications. In crowdsensing, it is of great interest find optimal task allocation, which challenging since tasks with different requirements quality are typically associated specific locations and mobile users constrained by time budgets. We show that allocation problem NP hard. then focus on approximation algorithms, devise local ratio based algorithm (LRBA). Our analysis shows aggregate rewards obtained...

Thanks to the convergence of pervasive mobile communications and fast-growing online social networking, networking is penetrating into our everyday life. Aiming develop a systematic understanding networks, in this paper we exploit ties human networks enhance cooperative device-to-device (D2D) communications. Specifically, as handheld devices are carried by beings, leverage two key phenomena, namely trust reciprocity, promote efficient cooperation among devices. With insight, coalitional...

In-home health monitoring has attracted great attention for the ageing population worldwide. With abundant user data accessed by Internet of Things (IoT) devices and recent development in machine learning, smart healthcare seen many successful stories. However, existing approaches in-home do not pay sufficient to privacy thus are far from being ready large-scale practical deployment. In this paper, we propose FedHome, a novel cloud-edge based federated learning framework monitoring, which...

Mobile edge caching aims to enable content delivery within the radio access network, which effectively alleviates backhaul burden and reduces response time. To fully exploit storage resources, most popular contents should be identified cached. Observing that user demands on certain vary greatly at different locations, this paper devises location-customized schemes maximize total hit rate. Specifically, a linear model is used estimate future For case with zero-mean noise, ridge...

Recent advances in artificial intelligence have driven increasing intelligent applications at the network edge, such as smart home, factory, and city. To deploy computationally intensive Deep Neural Networks (DNNs) on resource-constrained edge devices, traditional approaches relied either offloading workload to remote cloud or optimizing computation end device locally. However, cloud-assisted suffer from unreliable delay-significant wide-area network, local computing are limited by...

Metaverse encapsulates our expectations of the next-generation Internet, while bringing new key performance indicators (KPIs). Although conventional ultra-reliable and low-latency communications (URLLC) can satisfy objective KPIs, it is difficult to provide a personalized immersive experience that distinctive feature Metaverse. Since quality (QoE) be regarded as comprehensive KPI, URLLC evolved towards next generation (xURLLC) with resource allocation scheme achieve higher QoE. To deploy...

Distributed learning is envisioned as the bedrock of next-generation intelligent networks, where agents, such mobile devices, robots, and sensors, exchange information with each other or a parameter server to train machine models collaboratively without uploading raw data central entity for centralized processing. By utilizing computation/communication capability individual distributed paradigm can mitigate burden at processors help preserve privacy users. Despite its promising applications,...

We consider a cyber-physical system consisting of two interacting networks, i.e., cyber-network overlaying physical-network. It is envisioned that these systems are more vulnerable to attacks since node failures in one network may result (due the interdependence) other network, causing cascade would potentially lead collapse entire infrastructure. The robustness interdependent against this sort catastrophic failure hinges heavily on allocation (interconnecting) links connect nodes network....

Online dynamic security assessment (DSA) is examined in a data-mining framework by taking into account the operating condition (OC) variations and possible topology changes of power systems during horizon. Specifically, robust scheme proposed based on adaptive ensemble decision tree (DT) learning. In offline training, boosting algorithm employed to build classification model as weighted voting multiple unpruned small-height DTs. Then, DTs are periodically updated incorporating new training...

The virtual multiple input output (MIMO) technique can dramatically improve the performance of a multi-cell distributed antenna system (DAS), thanks to its great potentials for inter-cell interference mitigation. One most challenging issues MIMO is acquisition channel state information at transmitter (CSIT), which usually leads an overwhelming amount overhead. In this work, we focus on case that only slowly-varying large-scale required transmitter, and explore gain be achieved by coordinated...

This paper investigates the relation between three different notions of privacy: identifiability, differential privacy, and mutual-information privacy. Under a unified privacydistortion framework, where distortion is defined to be expected Hamming distance input output databases, we establish some fundamental connections these privacy notions. Given maximum allowable D, define privacy-distortion functions ∈ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML"...

We study the problem of minimum-number full-view area coverage in camera sensor networks, i.e., how to select minimum number sensors guarantee a given region. Full-view is challenging because 2-D continuous domain has be considered. To tackle this challenge, we first intrinsic geometric relationship between and point prove that can guaranteed, as long selected ensuring set points covered. This leads significant dimension reduction for problem. Next, NP-hard propose two approximation...

We study the diffusion of information in an overlaying social-physical network. Specifically, we consider following set-up: There is a physical network where spreads amongst people through conventional communication media (e.g., face-to-face communication, phone calls), and conjoint to this network, there are online social networks via web sites such as Facebook, Twitter, FriendFeed, YouTube, etc. quantify size critical threshold epidemics by assuming that diffuses according SIR epidemic...

A data mining approach using ensemble decision trees (DTs) learning is proposed for online dynamic security assessment (DSA), with the objective of mitigating impact possibly missing PMU data. Specifically, multiple small DTs are first trained offline a random subspace method. In particular, developed method exploits hierarchy wide-area monitoring system (WAMS), locational information attributes, and availability measurements, so as to improve overall robustness Then, performance re-checked...

Wind ramps introduce significant uncertainty into wind power generation. Reliable system operation, however, requires accurate detection and forecast of ramps, especially at high penetration levels. In this paper, to deal with the ramp dynamics, a support vector machine (SVM)-enhanced Markov model is developed for short-term forecast, based on one key observation from measurement data that often occur specific patterns. Specifically, using historical turbine outputs recorded an actual farm,...

Fault diagnosis in power grids is known to be challenging, due the massive scale and spatial coupling therein. In this study, we explore multiscale network inference for fault detection localization. Specifically, model phasor angles across buses as a Markov random field (MRF), where conditional correlation coefficients of MRF are quantified terms physical parameters systems. Based on model, then study decentralized diagnosis, through change localization matrix MRF. Particularly, based...

Multicast communications is an efficient mechanism for one-to-many transmissions over a broadcast wireless channel, and considered as key technology supporting emerging broadband multimedia services in the next generation networks, such Internet Protocol Television (IPTV), mobile TV, etc. Therefore, it critical to design multicast scheduling schemes support these services. In this paper, we propose cooperative scheme achieving reliable transmission IEEE 802.16 based metropolitan area...

Smart meters, designed for information collection and system monitoring in smart grid, report fine-grained power consumption to utility providers. With these highly accurate profiles of energy usage, however, it is possible identify consumers' specific activities or behavior patterns, thereby giving rise serious privacy concerns. This paper addresses concerns by designing a cost-effective privacy-preserving management technique that uses rechargeable battery. From holistic perspective,...

In this paper, we study distributed opportunistic scheduling (DOS) in an ad hoc network, where many links contend for the same channel using random access. such a DOS involves process of joint probing and scheduling. Due to fading, link condition corresponding successful could be either good or poor. latter case, further probing, although at cost additional delay, may lead better conditions hence yield higher throughput. The desired tradeoff boils down judiciously choosing optimal stopping...

Integrating volatile renewable energy resources into the bulk power grid is challenging, due to reliability requirement that at each instant load and generation in system remain balanced. In this study, we tackle challenge for smart with integrated wind generation, by leveraging multi-timescale dispatch scheduling. Specifically, consider grids two classes of users - traditional opportunistic (e.g., meters or appliances), investigate pricing timescales, via day-ahead scheduling realtime...

In wireless sensor networks, relay node placement has been proposed to improve energy efficiency. this paper, we study two-tiered constrained problems, where the nodes can be placed only at some prespecified candidate locations. To meet connectivity requirement, connected single-cover problem each is covered by a base station or (to which transmit data), and form network with stations. survivability 2-connected double-cover two stations nodes, We these problems under assumption that R \ge 2r...

In this paper, short-term forecast of wind farm generation is investigated by applying spatio-temporal analysis to extensive measurement data collected from a large where multiple classes turbines are installed. Specifically, using the turbines' power outputs recorded across two consecutive years, graph-learning based carried out characterize statistical distribution and quantify level crossing rate farm's aggregate output. Built on these characterizations, finite-state Markov chains...