Mobile social networks (MSNs) are a kind of delay tolerant network that consists lots mobile nodes with characteristics. Recently, many social-aware algorithms have been proposed to address routing problems in MSNs. However, these tend forward messages the locally optimal characteristics, and thus cannot achieve performance. In this paper, we propose distributed Community-Aware Opportunistic Routing (CAOR) algorithm. Our main contributions home-aware community model, whereby turn an MSN into...

Real-time analytics on video data demands intensive computation resources and high energy consumption. Traditional cloud-based relies large centralized clusters to ingest streams. With edge computing, we can offload compute-intensive analysis tasks the nearby server, thus mitigating long latency incurred by transmission via wide area networks. When offloading frames from front-end device application configuration (frame sampling rate frame resolution) will impact several metrics, such as...

A mobile social network (MSN) is a special delay tolerant (DTN) composed of nodes with characteristics. Mobile in MSNs generally visit community homes frequently, while other locations are visited less frequently. We propose novel zero-knowledge MSN routing algorithm, homing spread (HS). The have higher priority to messages into the network. Theoretical analysis shows that proposed algorithm can given number message copies an optimal way when inter-meeting times between any two and node home...

Mobile CrowdSensing (MCS), through which a requester can coordinate crowd of workers to accomplish some data collection tasks, has been recognized as promising paradigm for large-scale acquisition in recent years. Many researches focus on the worker recruitment problem MCS, but most them either have assumption that workers' qualities are known ahead time or cannot ensure report costs honestly. In this paper, we propose an incentive mechanism based Combinatorial Multi-Armed Bandit and reverse...

Major cities worldwide have millions of cameras deployed for surveillance, business intelligence, traffic control, crime prevention, etc. Real-time analytics on video data demands intensive computation resources and high energy consumption. Traditional cloud-based relies large centralized clusters to ingest streams. With edge computing, we can offload compute-intensive analysis tasks nearby servers, thus mitigating long latency incurred by transmission via wide area networks. When offloading...

In this paper, we investigate the incentive problem in Spatial Crowdsourcing (SC), where mobile social-aware workers have unknown qualities and can share their answers to tasks via social networks. The objectives are recruit high-quality maximize all parties' utilities simultaneously. However, most existing works assume that of known advance or cannot take into account together, especially having not considered impact Thus, propose an mechanism based on multi-armed bandit three-stage...

With the explosive spread of smart mobile devices, Mobile CrowdSensing (MCS) has been becoming a promising paradigm, by which platform can coordinate group workers to complete large-scale data collection tasks using their devices. In this paper, we investigate incentive mechanism design in MCS systems, taking freshness collected and social benefits into consideration. First, Age Information (AoI) metric is introduced measure data. Then, model with AoI guarantees as an incomplete information...

In this paper, we propose a time-sensitive utility model for delay tolerant networks (DTNs), in which each message has an attached benefit that decays over time. The of is the minus transmission cost incurred by delivering message. This analogous to postal service real world, inherently provides good balance between and cost. Under model, Time-sensitive Opportunistic Utility-based Routing (TOUR) algorithm. TOUR single-copy opportunistic routing algorithm, forwarding set maintained node...

Due to the explosive proliferation of mobile cloud computing applications, much data needs be transmitted between users and clouds, incurring a huge traffic demand on cellular networks. Mobile offloading is promising approach address this challenge. In paper, we focus problem many deadline-sensitive items some WiFi networks with capacity constraints; that is, how schedule each item networks, so can offload as before their deadlines possible, while taking constraints transmission into...

In the multi-access edge computing environment, app vendors deploy their services and applications at network edges, users offload computation tasks to servers. We study user-perceived delay-aware service placement user-allocation problem in environment. model MEC-enabled network, where delay consists of transmission delay. The total cost offloading system is defined as sum placement, server usage energy consumption cost, we need minimize by determining overall service-placing decision...

A mobile social network (MSN) is a special kind of delay tolerant (DTN) composed nodes that move around and share information with each other through their carried short-distance wireless communication devices. main characteristic MSNs in the networks generally visit some locations (namely, community homes) frequently, while visiting less frequently. In this paper, we propose novel zero-knowledge multi-copy routing algorithm, homing spread (HS), for homogeneous MSNs, which all homes. HS...

We study a spectrum auction problem where each request from new users has spatial, temporal, and spectral features. Our goal is to design truthful mechanisms that maximize either the overall social efficiency of (a.k.a buyers) or revenue owner seller). Given optimal conflict-free allocation NP-hard, this paper proposes series near-optimal based on following approximation techniques: linear programming (LP) relaxation, randomized rounding, derandomized monotone derandomization, Lavi-Swamy...

In this paper, we focus on the privacy preserving mechanism design for crowdsourced Federated Learning (FL), where a requester can outsource its model training task to some workers via an FL platform. A potential way preserve of workers' local data is leverage Differential Privacy (DP) mechanisms models. However, most these studies cannot allow dominate their own protection levels by themselves. Thus, propose Personalized Preserving Mechanism, called P3M, satisfy heterogeneous needs workers,...

Currently, BPEL is the de-facto standard for web service composition. Because services are autonomous and loosely coupled, processes susceptible to a wide variety of faults. However, only provides limited constructs handling faults, which makes fault time-consuming error-prone task. In this paper, we propose declarative approach enhancing reliability processes. Our solution specifies logic through set Event- Condition-Action (ECA) rules build on an extensible fault-tolerant patterns. These...

A mobile opportunistic social network (MOSN) is a new type of delay tolerant (DTN), in which the users contact each other opportunistically. While cooperative caching Internet has been studied extensively, MOSNs considerably different and challenging problem due to probabilistic nature among MOSNs. In order reduce total access delay, we let cooperatively cache these data items their limited buffer space. We balance between selfishness (caching according its own preference) unselfishness...

Mobile CrowdSensing (MCS) is a mobile computing paradigm, through which platform can coordinate crowd of workers to accomplish large-scale data collection tasks using their devices. Information freshness has attracted much focus on MCS research worldwide. In this paper, we investigate the incentive mechanism design in systems that take collected and social benefits into concerns. First, introduce Age (AoI) metric measure data. Then, model with AoI guarantees as novel incomplete information...

In recent years, deep neural networks have been applied in machine translation systems, resulting the so-called (NMT) models that can improve quality significantly. However, due to brittleness of network, systems could return erroneous translations lead misunderstandings or even cause serious losses. To detect errors, various testing techniques proposed. As a popularly used technique, metamorphic mainly relies on text syntactic structure while ignoring meaning sentences (i.e., semantic...

Federated learning (FL) is a newly emerging distributed machine paradigm, whereby server can coordinate multiple clients to jointly train model by using their private datasets. Many researches focus on designing incentive mechanisms in FL, but most of them cannot allow that flexibly determine privacy budgets themselves. In this article, we propose privacy-preserving mechanism (NICE) based differential (DP) and Stackelberg game for FL systems industrial Internet Things. First, design flexible...

Crowdsensing Data Trading (CDT), through which a platform can aggregate some data collected by group of mobile users with sensing devices (a.k.a., sellers) and sell the corresponding statistics to consumers, has been recognized as promising paradigm for large-scale trading in recent years. It is critical select sellers high qualities maximize all participants' profits simultaneously. However, most existing CDT systems either assume that sellers' are known advance or cannot realize concurrent...

Mobile crowdsensing (MC), an excellent solution to large-scale spatiotemporal data sensing problems, has recently received lots of attention from both industry and academia. In the MC system, any requester can acquire for his points interest (PoIs) by offering some payments attract a group mobile users capable completing these PoI-related tasks. However, current work neglected three vital factors, more or less. First, they assume that distributed are mutually independent in MC, ignoring...

With the development of deep learning, artificial intelligence applications and services have boomed in recent years, including recommendation systems, personal assistant video analytics. Similar to other edge computing environment, tasks are pushed network edge. In this paper, we consider multi-user edge-assisted analytics task offloading (MEVAO) problem, where users with various accuracy requirements. All independently choose their decisions, satisfying requirement, offload data server....

Federated Learning (FL) is a newly-emerging distributed ML model, where server can coordinate multiple workers to cooperatively train learning model by using their private datasets, while ensuring these datasets not be revealed others. In this paper, we focus on the incentive mechanism design for FL systems. Taking incentives into consideration, first two utility functions and workers, respectively. Then, corresponding optimization problem as two-stage Stackelberg game seeing leader some...

With the explosive proliferation of smartphones, many mobile cloud computing applications have emerged in recent years. These generally involve data transmissions between users and side. In order to reduce monetary cost these transmissions, an effective approach is offload partial traffic from cellular networks WiFi networks, when pass by some Access Points (APs). this paper, we focus on problem offloading deadline-sensitive items APs with capacity constraints; that is, how schedule each...

Traditional spatial crowdsourcing (SC) systems employ a centralized server platform to provide services for requesters. Such design requires powerful resource capacity and often cannot accomplish the urgent demands due unpredictable network latency. In order ensure scalability of quality services, we study distributed SC (DSC), where diversity location-relative provided by various service providers (SPs) can deploy on edge clouds (ECs) with low time Since resources are limited, SPs need...