Mobile cloud computing is emerging as a main ubiquitous platform to provide rich resources for various applications of mobile devices. Although most existing studies in focus on energy savings devices by offloading computing-intensive jobs from remote clouds, the access delays between users and clouds usually are long sometimes unbearable. Cloudlet new technology capable bridge this gap, can enhance performance significantly while meeting crisp response time requirements users. In paper, we...

With advances in wireless communication technology, more and people depend heavily on portable mobile devices for businesses, entertainments social interactions. Although such can offer various promising applications, their computing resources remain limited due to size. This however be overcome by remotely executing computation-intensive tasks clusters of near computers known as cloudlets. As increasing numbers access the Internet via devices, it is reasonable envision future that cloudlet...

This paper addresses the problem of temporal sentence grounding (TSG), which aims to identify boundary a specific segment from an untrimmed video by query. Previous works either compare pre-defined candidate segments with query and select best one ranking, or directly regress timestamps target segment. In this paper, we propose novel localization framework that scores all pairs start end indices within simultaneously biaffine mechanism. particular, present Context-aware Biaffine Localizing...

With the development of cloud computing, virtual machine migration is emerging as a promising technique to save energy, enhance resource utilizations, and guarantee Quality Service (QoS) in datacenters. Most existing studies on migration, however are based single migration. Although there some researches multiple machines author usually does not consider correlation among these machines. In practice, order energy maintain system performance, providers need migrate correlated or entire...

Wireless energy transfer has emerged as a promising technology for wireless sensor networks to power sensors with controllable yet perpetual energy. In this paper, we study replenishment by employing mobile charger (charging vehicle) charge wirelessly in rechargeable network, so that the sum of charging rewards collected from all charged per tour is maximized, subject capacity charger, where amount reward received proportional sensor. The will be spent on both its mechanical movement and...

Wireless energy transfer technology based on magnetic resonant coupling has emerged as a promising for wireless sensor networks, by providing controllable yet continual to sensors. In this paper, we study the use of mobile charger wirelessly charge sensors in rechargeable network so that sum lifetimes is maximized while travel distance minimized. Unlike existing studies assumed must its full capacity before moving next sensor, here assume each can be partially charged more their depletions....

Mobile edge-cloud (MEC) aims to support low latency mobile services by bringing remote cloud nearer users. However, in order deal with dynamic workloads, MEC is deployed a large number of fixed-location micro-clouds, leading resource wastage during stable/low workload periods. Limiting the micro-clouds improves utilization and saves operational costs, but faces service performance degradations due insufficient physical capacity peak time from nearby micro-clouds. To efficiently requirement...

Query-based moment localization is a new task that localizes the best matched segment in an untrimmed video according to given sentence query. In this task, one should pay more attention thoroughly mine visual and linguistic information. To end, we propose novel Cross- Self-Modal Graph Attention Network (CSMGAN) recasts as process of iterative messages passing over joint graph. Specifically, graph consists Cross-Modal interaction (CMG) relation (SMG), where frames words are represented...

Pushing the cloud frontier to network edge close mobile users has attracted tremendous interest not only from operators but also service providers. In particular, deployment of cloudlets in metropolitan area networks enables providers provide low-latency services through implementing their specified virtualized functions (VNFs) while meeting Quality-of-Service (QoS) requirements. this paper, we formulate a novel task offloading problem edge-cloud network, where each requests function with...

With increasing focus on Artificial Intelligence (AI) applications, Deep Neural Networks (DNNs) have been successfully used in a number of application areas. As the layers and neurons DNNs increases rapidly, significant computational resources are needed to execute learned DNN model. This ever-increasing resource demand is currently met by large-scale data centers with state-of-the-art GPUs. However, availability mobile edge computing 5G technologies provide new possibilities for DNN-driven...

Temporal language localization in videos aims to ground one video segment an untrimmed based on a given sentence query. To tackle this task, designing effective model extract ground-ing information from both visual and textual modalities is crucial. However, most previous attempts field only focus unidirectional interactions query, which emphasizes words listen attends via vanilla soft attention, but clues query-by-video implying where look are not taken into consideration. In paper, we...

With the development of 5G technology, mobile edge computing is emerging as an enabling technique to promote Quality Service (QoS) network services. In particular, response latency services can be significantly reduced by deploying cloudlets at base stations in clouds. Network service providers that usually deploy their remote clouds now shift from proximity users. However, permanent placement into may not economic, since and bandwidth resources are limited relatively expensive. A smart way...

In this paper, we study sensing data collection of IoT devices in a sparse IoT-sensor network, using an energy-constrained Unmanned Aerial Vehicle (UAV), where the sensory is stored while may or not be within transmission range each other. We formulate two novel problems to fully partially collect from UAV, by finding closed tour for UAV that consists hovering locations and sojourn duration at such accumulative volume collected maximized, subject energy capacity on consumes both flying one...

Internet-of-Things-enabled E-health system, which could monitor and collect the personal health information (PHI), has gradually transformed clinical treatment to a more personalized way with in-home monitoring smart devices. Then, collected PHI, decision support systems (CDSSs), are based on data mining techniques historical electronic medical records (EMRs) help clinicians make proper decisions, have attracted considerable attention. To address issues, such as network congestion low rate...

Temporal sentence grounding (TSG) is crucial and fundamental for video understanding. Although existing methods train well-designed deep networks with large amount of data, we find that they can easily forget the rarely appeared cases during training due to off-balance data distribution, which influences model generalization leads unsatisfactory performance. To tackle this issue, propose a memory-augmented network, called Memory-Guided Semantic Learning Network (MGSL-Net), learns memorizes...

Mobile-edge computing (MEC) is a new paradigm that enables cloud and information technology (IT) services to be delivered at the network's edge. By shifting load of individual local servers, MEC helps meet requirements ultralow latency, localized data processing, extends potential Internet Things (IoT) for end-users. However, crosscutting nature multidisciplinary components necessary its deployment have presented additional security privacy concerns. Fortunately, artificial intelligence (AI)...

In this paper we study the cloudlet placement problem in a large-scale Wireless Metropolitan Area Network (WMAN) that consists of many wireless Access Points (APs). Although most existing studies mobile cloud computing mainly focus on energy savings devices by offloading computing-intensive jobs from them to remote clouds, access delay between users and clouds usually is large sometimes unbearable. Cloudlet as new technology capable bridge gap, has been demonstrated enhance performance...

Stringent delay requirements of many mobile applications have led to the development edge clouds, offer low latency network services at edges. Most conventional are implemented via hardware-based functions, including firewalls and load balancers, guarantee service security performance. However, implementing functions usually incurs both a high capital expenditure (CAPEX) operating (OPEX). Network Function Virtualization (NFV) exhibits potential reduce CAPEX OPEX significantly, by deploying...

Multicasting is a fundamental functionality of many network applications, including online conferencing, event monitoring, video streaming, and so on. To ensure reliable, secure, scalable multicasting, service chain that consists functions (e.g., firewalls, intrusion detection systems, transcoders) usually associated with each multicast request. We refer to such request requirement as an function virtualization (NFV)-enabled In this paper, we study NFV-enabled multicasting in...

The Internet of Things (IoT) technology provisions unprecedented opportunities to evolve the interconnection among human beings. However, latency brought by unstable wireless networks and computation failures caused limited resources on IoT devices prevents users from experiencing high efficiency seamless user experience. To address these shortcomings, integrated Mobile Edge Computing (MEC) with remote clouds is a promising platform enable delay-sensitive service provisioning for...

Mobile Edge Computing (MEC) has emerged as a promising paradigm catering to overwhelming explosions of mobile applications, by offloading compute-intensive tasks MEC networks for processing. The surging deep learning brings new vigor and vitality shape the prospect intelligent Internet Things (IoT), edge intelligence arises provision real-time neural network (DNN) inference services users. To accelerate processing DNN user request in an network, model usually can be partitioned into two...

In this paper, we study the deployment of Unmanned Aerial Vehicles (UAVs) to collect data from IoT devices, by finding a collection tour for each UAV. To ensure &#x2018;freshness&#x2019; collected data, total time spent in UAV that consists flying and must be no greater than given delay <inline-formula> <tex-math notation="LaTeX">$B$ </tex-math></inline-formula>, e.g., 20 minutes. consider problem deploying minimum number UAVs their tours, subject constraint any is...

Mobile Edge Computing (MEC) has been identified as a desirable computing paradigm that provides efficient and effective services for various applications, while meeting stringent service delay requirements. Orthogonal to the MEC paradigm, Network Function Virtualization (NFV) technology is another enabling network resource management with great flexibility scalability, where instances of Virtual Functions (VNFs) are deployed in edge servers Service Chains (SFCs) SFC-enabled services....

Temporal video grounding (TVG) aims to localize a target segment in according given sentence query. Though respectable works have made decent achievements this task, they severely rely on abundant video-query paired data, which is expensive collect real-world scenarios. In paper, we explore whether model can be learned without any annotations. To the best of our knowledge, paper first work trying address TVG an unsupervised setting. Considering there no supervision, propose novel Deep...

Serverless computing is emerging as an enabling technology for elastic and low-cost AI applications in the edge of core networks. It allows developers to decompose a complex training time-sensitive inference task into multiple functions with dependency, upload Multi-access Edge Computing platform (MEC) execution. adopts popular design principle: disaggregation storage computation, making 'stateless'. However, most are 'stateful' rely on external service manage their states (ephemeral data)....