Local differential privacy (LDP) is a recently proposed standard for collecting and analyzing data, which has been used, e.g., in the Chrome browser, iOS macOS. In LDP, each user perturbs her information locally, only sends randomized version to an aggregator who performs analyses, protects both users against private leaks. Although LDP attracted much research attention recent years, majority of existing work focuses on applying complex data and/or analysis tasks. this paper, we point out...

Recently, there have been several promising techniques developed for schedulability analysis and response time multiprocessor systems based on over-approximation. This paper contains two contributions. First, to improve the precision, we apply Baruah's window framework poradic tasks where deadlines of are within their periods. The crucial observation is that global fixed priority scheduling, a bound each task can be efficiently estimated by fixed-point computation without enumerating all...

Differential privacy (DP) is a promising scheme for releasing the results of statistical queries on sensitive data, with strong guarantees against adversaries arbitrary background knowledge. Existing studies DP mostly focus simple aggregations such as counts. This paper investigates publication DP-compliant histograms, which an important analytical tool showing distribution random variable, e.g., hospital bill size certain patients. Compared to whose are purely numerical, histogram query...

A mobile operating system often needs to collect frequent new terms from users in order build and maintain a comprehensive dictionary. Collecting keyboard usage data, however, raises privacy concerns. Local differential (LDP) has been established as strong standard for collecting sensitive information users. Currently, the best known solution LDP-compliant term discovery transforms problem into n-grams under LDP, subsequently reconstructs collected by modelling latter graph, identifying...

Due to the complexity of blockchain technology, it usually costs too much effort build, maintain and monitor a system that supports targeted application. To this end, emerging "Blockchain as Service" (BaaS) makes distributed ledgers more accessible, particularly for businesses, by reducing overheads. BaaS combines high computing power cloud computing, pervasiveness IoT decentralization blockchain, allowing people build their own applications while ensuring transparency openness system. This...

Hypergraphs can model higher-order relationships among data objects that are found in applications such as social networks and bioinformatics. However, recent studies on hypergraph learning extend graph convolutional to hypergraphs cannot learn effectively from features of unlabeled data. To learning, we propose a contrastive neural network, CHGNN, exploits self-supervised techniques labeled First, CHGNN includes an adaptive view generator adopts auto-augmentation strategy learns perturbed...

The major obstacle to use multicores for real-time applications is that we may not predict and provide any guarantee on properties of embedded software such platforms; the way handling on-chip shared resources as L2 cache have a significant impact timing predictability. In this paper, propose space isolation techniques avoid contention hard tasks running with caches. We present scheduling strategy both constraints, which allows each task fixed number partitions, makes sure at time partition...

Liu and Layland discovered the famous utilization bound for fixed-priority scheduling on single processor systems in 1970's. Since then, it has been a long standing open problem to find algorithms with same multiprocessor systems. In this paper, we present partitioning-based algorithm Layland's bound.

It is predicted that multicores will be increasingly used in future embedded real-time systems for high performance and low energy consumption. The major obstacle we may not predict provide any guarantee on properties of software such platforms. shared memory bus among the most critical resources, which severely degrade timing predictability multicore due to access contention between cores. In this paper, study a architecture where each core has local L1 cache all cores use off-chip memory....

The production and real-time usage of streaming data bring new challenges for systems due to huge volume quick response request applications. Message queuing that offer high throughput low latency play an important role in today's big processing. There are several popular message also many in-lab academia. These with different design philosopies have characteristics. It is non-trivial a non-expert choose suitable system meet his specific requirement. With this premise, our primary...

With the continuous development of blockchain technology, many applications, such as digital currencies in form tokens, are deployed. However, due to lack data and value inter-blockchain transmission methods, these chain applications islands. Therefore, cross-chain technology is proposed applied distributed transaction platforms, finance, e-government other different fields. Although just started, its gradually improving meet needs system. At same time, they have already realized asset...

The graph structure is a very important means to model schemaless data with complicated structures, such as protein-protein interaction networks, chemical compounds, knowledge query inferring systems, and road networks. This paper focuses on the index for similarity search set of large sparse graphs proposes an efficient indexing mechanism by introducing Q-Gram idea. By decomposing small grams (organized κ-Adjacent Tree patterns) pairing-up those patterns, lower bound estimation their edit...

Density Peaks (DP) is a recently proposed clustering algorithm that has distinctive advantages over existing algorithms. It already been used in wide range of applications. However, DP requires computing the distance between every pair input points, therefore incurring quadratic computation overhead, which prohibitive for large data sets. In this paper, we study efficient distributed algorithms DP. We first show naive MapReduce solution (Basic-DDP) high communication and overhead. Then,...

Magnetic flux leakage (MFL) inspection in nondestructive testing (NDT) has been widely used damaged pipeline defect inversion. The changeable environment and the complexity of MFL signal have brought severe challenges to accurate estimation sizes inversion issue. This article proposes a novel method (WT-STACK) based on stacking learning. consists two parts. First, multi-domain feature extraction with three-axis (axial, radial, circumferential) signals is constructed. To avoid information...

GNN's training needs to resolve issues of vertex dependencies, i.e., each representation's update depends on its neighbors. Existing distributed GNN systems adopt either a dependencies-cached approach or dependencies-communicated approach. Having made intensive experiments and analysis, we find that decision choose one the other for best performance is determined by set factors, including graph inputs, model configurations, an underlying computing cluster environment. If various trainings...

Blockchain serves as a replicated transactional processing system in trustless distributed environment. Existing blockchain systems all rely on an explicit ordering step to determine the global order of transactions that are collected from multiple peers. The consensus can be bottleneck since it must Byzantine-fault tolerant and scarcely benefit parallel execution. In this paper, we propose ordering-free architecture makes implicit through deterministic Based novel architecture, develop...