Search

With the privacy protection increasingly being concerned, Data centralization often heavily causes a big risk of privacy protection, gradually, there is a prevailing trend to enhance the security performance by means of data decentralization, above all, for health care internet of things (IoT) data. Meanwhile, Federated learning has obvious privacy advantages compared to data center training on protecting privacy data. For this reason, a novel framework based on federated learning is presented in this paper, which is suitable for private and decentralized data sets, such as big data in healthy Internet of Things. Specifically, the main work of the puts forward framework includes: (1) Multi-center data collection of healthy Internet of Things. (2) healthy data analysis of Internet of Things. (3) privacy protection method for data of healthy Internet of Things. Finally, related experiments show that the proposed method is feasible, and compared with the traditional methods, it has significantly improved the performance in Quality of Service (QoS) and IoUs indicator. [ABSTRACT FROM AUTHOR]

The Internet of Things (IoT) is commonly deployed with devices of limited power and computation capability. A centralized IoT architecture provides a simplified management for IoT system but brings redundancy by the unnecessary data traffic with a data center. A decentralized IoT reduces the cost on data traffic and is resilient to the single-point-of-failure. The blockchain technique has attracted a large amount of research, which is redeemed as a perspective of decentralized IoT system infrastructure. It also brings new privacy challenges for that the blockchain is a public ledger of all digital events executed and shared among all participants. The decentralized IoT system relies on the leader election deeply to implement the decentralized communications among the distributed nodes. The conventional leader election must have a centralized authority, contrasting to the decentralization. As an alternative, self-tallying type schemes have been proposed in the literature for decentralized systems. These schemes suffer from adaptive and abortive issues. Also, some additional factors should be considered, such as the availability of candidate nodes. If the candidate node is unavailable after the voting phase due to being offline or ongoing tasks, the next available candidate should be elected. To accommodate such a need, in this paper, we propose a new leader election paradigm called voluntary-tallying leader election, which achieves the core requirements such as ballet secrecy, voter privacy and the additional feature of voluntary-tallying. We formalize the system and security models for this new election paradigm and present a secure and practical construction. [ABSTRACT FROM AUTHOR]

• A new power-exponent function model based on the CPU utilization rate is put forward. • A hardware-aware CPU power measurement (HCPM) is designed and implemented. • Experimental results show the proposed HCPM is effective and efficient. The energy consumption of cloud servers accounts for about 25% of the total energy of cloud data centers. Reducing and optimizing this energy consumption is thus extremely important in energy saving in cloud data centers. Power model is fundamental in energy efficiency optimization scheduling for cloud computing. However, systems and tools for power measurement in the cloud computing environment are relatively scarce, and power models of cloud servers cannot keep up with the times. Therefore, we propose a new CPU power model named power-exponent function model (PEFM) is proposed, which provides higher accuracy in estimating the CPU power of the latest cloud servers than the current linear, polynomial and power function models. A novel hardware-aware CPU power measurement (HCPM) is also proposed, that can select an appropriate CPU power model through the launch year of CPU without power model training. For validating the efficacy of PEFM and HCPM, a set of experiments including OpenStack cluster experiment, based on a distributed energy meter (DEM) implemented by our team were conducted. The experimental results indicate that the proposed PEFM and HCPM not only improve the accuracy of CPU power estimation in cloud servers in cloud environment, but also reduce the difficulty of model training and simplify system deployment. [ABSTRACT FROM AUTHOR]

Data centers for cloud computing must accommodate numerous parallel task executions simultaneously. Therefore, data centers have many virtual machines (VMs). Minimizing the scheduling length of parallel task sets becomes a critical requirement in cloud computing systems. In this study, we propose an efficient priority and relative distance (EPRD) algorithm to minimize the task scheduling length for precedence constrained workflow applications without violating the end-to-end deadline constraint. This algorithm consists of two processes. First, a task priority queue is established. Then, a VM is mapped for a task in accordance with its relative distance. The proposed method can effectively improve VM utilization and scheduling performance. Extensive rigorous experiments based on randomly generated and real-world workflow applications demonstrate that the resource reduction rate and scheduling length of the EPRD algorithm significantly surpass those of existing algorithms. [ABSTRACT FROM AUTHOR]

Sensor–cloud system, as one of the vital components of the Internet of Everything, has increasingly gained popularity since it bridges the physical and cyber worlds. In the smart city, sensing devices are deployed to monitor the environment and the data aggregate at the roadside Aggregation Devices (ADs). Then, Mobile Vehicles (MVs) submit the data to the data center in an opportunistic routing way when passing ADs. We propose a Content based Intelligent Trust Evaluation (CITE) scheme for trustworthy data collection over large-scale sensor-cloud systems. In CITE scheme, unmanned Mobile Electric Vehicles (MEVs) are employed to visit some sensing devices and collect their data along the designed routing path. Then, a delicate comparison algorithm is proposed to determine the credibility of devices' data operations, and trust evaluation is based on it. Similarly, the trust of MVs is obtained by comparing the data reported from ADs with the data submitted to the data center. Moreover, to reduce the cost of MEVs, we also design a path planning algorithm. The simulation results show that the CITE scheme can identify reliable sensing devices and MVs with a rate of 87.3% and 100% and increase the collection rate by 25.2% on average compared with the previous strategies. [ABSTRACT FROM AUTHOR]

The rapid development of artificial intelligence (AI) and edge computing technology has promoted the rapid growth of the number of data centers, but also caused large energy consumption. How to accurately predict the energy consumption of servers is crucial to optimize data centers. Thus, a real-time server energy consumption prediction model that integrates distance correlation and extreme gradient boosting (XGBoost) named ECDX is proposed. First, the distance correlation coefficient method is used to filter essential feature parameters and remove redundant features. Second, the cross-validation method is employed to optimize the hyperparameters. Thereafter, a data center server energy consumption prediction model is constructed using the XGBoost algorithm. Numerous experiments have been conducted to compare the performance of the ECDX model with that of benchmark models. The results show that ECDX can adapt to changes in workload, the prediction accuracy is significantly improved, and the average relative error is reduced by 4.698%. • Distance correlation aims to effectively extract a representative low-dimensional feature dataset from the original high-dimensional feature dataset. • We introduce an energy consumption prediction model named ECDX. For four workloads, the XGBoost algorithm is used for modeling to improve the model adaptability. • We conduct several simulation experiments using real server performance data to verify the effectiveness and adaptability of the ECDX model with different load types. [ABSTRACT FROM AUTHOR]

Abstract Nowadays, a large number of cloud services have been published and hosted by geo-distributed cloud data centers (Geo-2DCs). In spite of numerous benefits, those Geo-2DCs face significant challenges such as dynamic resource scaling where workload forecasting plays a crucial role in addressing such a challenge. High accuracy and fast learning are key indicators for workload forecasting and the literature has witnessed a lot of efforts. This work proposes an integrated forecasting method, equipped with noise filtering and data frequency representation, named Savitzky-Golay and Wavelet-supported Stochastic Configuration Networks (SGW-SCN), to predict the amount of workload in future time slots. In this approach, the workload time series is first smoothed by a Savitzky-Golay filter and then decomposed into multiple components via wavelet decomposition. With stochastic configuration networks, an integrated model is established to characterize statistical characteristics of both trend and detail components. Extensive results have demonstrated that the proposed method achieves higher forecasting accuracy and faster learning speed than typical forecasting methods. [ABSTRACT FROM AUTHOR]

Highlights • Optimization techniques for exploiting DCs' electrical and thermal energy flexibility. • Optimization techniques for delay-tolerant workload relocation in federated DCs. • New revenue streams by acting as active energy players. • Identify synergistic opportunities for cross multi-utility networks optimization. • A "lighter" deployment where the DC collaborates with operators of the smart energy grids • An "extended deployment" in which DCs collaborates with other energy players in the grids. Abstract In this paper, we have considered Data Centres (DCs) as computing facilities functioning at the crossroad of electrical, thermal and data networks and have defined optimisation techniques to exploit their energy flexibility. Our methods are leveraging on non-electrical cooling devices such as thermal storage and heat pumps for waste heat reuse and IT workload execution time shifting and spatial relocation in federated DCs. To trade energy flexibility we have defined an Energy Marketplace which allows DCs to act as active energy players integrated into the smart grid, contributing to smart city-level efficiency goals. Reinforcing this vision, we have proposed four innovative business scenarios that enable next generation smart Net-zero Energy DCs acting as energy prosumers at the interface with smart energy grids within smart city environments. Simulation experiments are conducted to determine the DCs potential electrical and thermal energy flexibility in meeting various network level goals and to assess the financial viability of the defined business scenarios. The results show that DCs have a significant amount of energy flexibility which may be shifted and traded to interested stakeholders thus allowing them to gain new revenue streams not foreseen before. [ABSTRACT FROM AUTHOR]

Deploying person re-identification (Re-ID) models in edge devices is of great importance in the field of intelligent surveillance. It can not only relieve the computational pressure on cloud data centers, but also effectively protects data privacy (e.g., pedestrian identity). However, current Re-ID methods mostly rely on the large-scale backbone with huge parameters, making them impossible to deploy in edge devices. To solve this problem, light-weight Re-ID backbones came into being. The current proposed light-weight Re-ID backbones are all multi-branch parallel structures based on multi-scale ideas. Nonetheless, an excessive number of parallel branches tend to extract redundant information, which takes up too much storage resources and easily leads to overfitting of the model structure. These problems limit the performance of the current models. In this paper, we innovatively propose a blocks-by-blocks combine-and-expand (BB-CE) training method. First we construct a base architecture using the proposed single-branch multi-scale (SMS) module. Then, by combining and expanding the blocks in the base architecture stage by stage during training, BB-CE training method can reduce the redundancy of the structure, utilize the multi-scale information more efficiently, and thus improve the matching accuracy of re-identification. Extensive experimental results on four generic person Re-ID datasets demonstrate that the proposed method outperforms the state-of-the-art algorithms. • We innovatively propose a BB-CE training method to design a novel lightweight person re-identification backbone. • We are the first to adopt the structural Rep to solve the model overfitting problem in the lightweight person Re-ID. • Our final architecture outperforms state-of-the-art methods on four generic person re-identification datasets. [ABSTRACT FROM AUTHOR]

Abstract This paper presents a novel emotion modeling methodology for incorporating human emotion into intelligent computer systems. The proposed approach includes a method to elicit emotion information from users, a new representation of emotion (AV-AT model) that is modelled using a genetically optimized adaptive fuzzy logic technique, and a framework for predicting and tracking user's affective trajectory over time. The fuzzy technique is evaluated in terms of its ability to model affective states in comparison to other existing machine learning approaches. The performance of the proposed affect modeling methodology is tested through the deployment of a personalised learning system, and series of offline and online experiments. A hybrid cloud intelligence infrastructure is used to conduct large-scale experiments to analyze user sentiments and associated emotions, using data from a million Facebook users. A performance analysis of the infrastructure on processing, analyzing, and data storage has been carried out, illustrating its viability for large-scale data processing tasks. A comparison of the proposed emotion categorizing approach with Facebook's sentiment analysis API demonstrates that our approach can achieve comparable performance. Finally, discussions on research contributions to cloud intelligence using sentiment analysis, emotion modeling, big data, and comparisons with other approaches are presented in detail. [ABSTRACT FROM AUTHOR]

Abstract As the usability of Cloud-based solutions has increased for various types of users with different needs, from scientists that want to process big data sets collected from sensors or business analysts that want to take decisions based on the huge amount of gathered data to simple users that store or share documents via a Cloud platform, the generated data is increasing more and more. For example, the ATLAS and other detectors at CERN generate petabytes of data and Facebook stores data with a rate of around 600 TB daily. In the current context, efficient scheduling for Big Data applications is a challenge and an appropriate scheduling technique is required for different types of incoming requests. In this paper we propose a scheduling algorithm for different types of computation requests: independent tasks, like bag of tasks (BoT) model or tasks with dependencies modeled as directed acyclic graphs (DAG), and they will be scheduled for execution in a Cloud datacenter. The tasks in the requests are scheduled on the available resources using the suitable scheduling algorithm for each request. We rely on a machine learning toolbox, named as MLBox, to find what algorithm should be used for a certain request. We implemented four heuristics for scheduling BoTs and four heuristics for DAGs scheduling and generated the training data for the machine learning algorithm by running multiple traditional scheduling algorithms and selecting the 'best' one for a given request. We evaluate the performance by comparing the scheduling of different tasks requests using some of the traditional algorithms and our machine learning based scheduling algorithm. [ABSTRACT FROM AUTHOR]

Abstract Due to their decoupling and scalability properties, multicast technologies are widely adopted in all the software projects aiming at integrating off-the-shelf systems, in order to compose large-scale complex infrastructures. This also holds within the context of mission critical systems, where strong requirements for reliable and timely data sharing are imposed. In the current literature, reliable multicast is always achieved at the expenses of violations of the temporal constraints, since retransmissions are used to recover lost messages. When timeliness requirements assume the same importance as the reliability ones, the techniques based on spatial redundancy, such as forward error correction, are preferable. However, such coding techniques are scarcely adopted within multicast services due to the difficulties related to the optimal tuning of the introduced redundancy. In this paper, we present a solution applying a proper coding scheme that jointly achieves reliability and timeliness in all those scenarios involving tree-based multicasting over the Internet. Such a solution employs game theory in order to select the best locations within the multicast tree where coding operations have to be performed. We prove the quality of this solution by using a series of simulations running on OMNET++, that compare the achievable quality with respect to the optimal solution of the underlying optimization problem. [ABSTRACT FROM AUTHOR]

In this paper, we proposed an unconditionally secure quantum Private Set Intersection Cardinality (PSI-CA) protocol. Compared with classical PSI-CA protocols, the proposed protocol can dramatically reduce the communication complexity, because it only requires O (1) communication cost, which is fully independent of the size of the sets. Furthermore, based on the proposed quantum PSI-CA protocol, we constructed a novel anonymous authentication scheme. This scheme can not only achieve two basic secure goals: secure authentication and anonymity, but can also dynamically update the authorized clients. When revoking any authorized client or adding a new client, it only needs to simply compute several set operations without any complex cryptographic operation, and thus it is very suitable for applications in some dynamic environments, e.g., large-scale client-server networks. [ABSTRACT FROM AUTHOR]

Recently, large-scale relational data analysis has gained much attention. Several Hadoop-based distributed systems have been proposed for scalable relational data analysis. Because the column-store approach is very suitable for analytic queries, many studies on column-oriented storage and query processing for Hadoop-based distributed systems have been conducted. However, two problems have arisen in existing studies, the first of which is that only a small amount of data is processed per task during distributed processing. Each task reads only the necessary data using the columnar structure. Because the task initialization in Hadoop requires a large overhead, it is inefficient that each task processes a small amount of data. The second problem is the lack of support for techniques that optimize columnar execution. Although many such techniques have been proposed for efficient columnar query execution, existing column-store methods for Hadoop-based distributed systems cannot support them efficiently. In this paper, we propose a novel column-store method called SSFile for Hadoop-based distributed systems. SSFile increases the actual amount of data processed per task and supports representative columnar execution techniques for efficient query processing. Through extensive experiments, we show that SSFile significantly improves the performance of distributed processing. [ABSTRACT FROM AUTHOR]