Asynchronous common subset (ACS) is a powerful paradigm enabling applications such as Byzantine fault-tolerance (BFT) and multi-party computation (MPC). The most efficient ACS framework in the information-theoretic setting is due to Ben-Or, Kelmer, and Rabin (BKR, 1994). The BKR ACS protocol has been both theoretically and practically impactful. BKR ACS has an$O(\log n)$running time (where$n$is the number of replicas) due to the usage of$n$parallel asynchronous binary agreement (ABA) instances, impacting both performance and scalability. Indeed, for a network of 16$\sim$64 replicas, the parallel ABA phase occupies about 95%$\sim$97% of the total runtime. A long-standing open problem is whether we can build an ACS framework with$O(1)$time while not increasing the message or communication complexity of the BKR protocol. We resolve the open problem, presenting the first constant-time ACS protocol with$O(n^{3})$messages in the information-theoretic and signature-free settings. Our key ingredient is the first information-theoretic and constant-time multivalued validated Byzantine agreement (MVBA) protocol. Our results can improve—asymptotically and concretely—various applications using ACS and MVBA. As an example, we implement FIN, a BFT protocol instantiated using our framework. Via a 121-server deployment on Amazon EC2, we show FIN reduces the overhead of the ABA phase to as low as 1.23% of the total runtime.

Asynchronous common subset (ACS) is a powerful paradigm enabling applications such as Byzantine fault-tolerance (BFT) and multi-party computation (MPC). The most efficient ACS framework in the information-theoretic setting is due to Ben-Or, Kelmer, and Rabin (BKR, 1994). The BKR ACS protocol has been both theoretically and practically impactful. However, the BKR protocol has an O(log n) running time (where n is the number of replicas) due to the usage of n parallel asynchronous binary agreement (ABA) instances, impacting both performance and scalability. Indeed, for a network of 16 ~ 64 replicas, the parallel ABA phase occupies about 95% ~ 97% of the total runtime in BKR. A long-standing open problem is whether we can build an ACS framework with O(1) time while not increasing the message or communication complexity of the BKR protocol.

Cloud computing has become a major approach to help reproduce computational experiments. Yet there are still two main difficulties in reproducing batch based Big Data analytics (including descriptive and predictive analytics) in the cloud. The first is how to automate end-to-end scalable execution of analytics including distributed environment provisioning, analytics pipeline description, parallel execution, and resource termination. The second is that an application developed for one cloud is difficult to be reproduced in another cloud, a.k.a. vendor lock-in problem. To tackle these problems, we leverage serverless computing and containerization techniques for automated scalable execution and reproducibility, and utilize the adapter design pattern to enable application portability and reproducibility across different clouds. We propose and develop an open-source toolkit that supports 1) fully automated end-to-end execution and reproduction via a single command, 2) automated data and configuration storage for each execution, 3) flexible client modes based on user preferences, 4) execution history query, and 5) simple reproduction of existing executions in the same environment or a different environment. We did extensive experiments on both AWS and Azure using four Big Data analytics applications that run on virtual CPU/GPU clusters. The experiments show our toolkit can achieve good execution performance, scalability, and efficient reproducibility for cloud-based Big Data analytics.

A boiler heating surface is composed of hundreds of tubes, whose temperatures may be different because of their positions, the influences of attempering water, and flue gas. Using a criteria based on the Davies–Bouldin index, in this article, we propose to partition a heating surface into local ones, whose interactions in temperature are represented as a weighted heating surface graph (HSG) at each point of time, and whose current features are embedded in the HSG's nodes. Then, a local heating surface temperature prediction model based on weighted graph convolutional networks and gated recurrent units (WGCN-GRU), is proposed. Graph convolutional networks (GCNs) receive a series of HSGs, and extract the features of local heating surfaces and their spatial dependences in a time window. Features output by GCNs are finally directed to GRUs for temperature predictions. Experiments show that WGCN-GRU can averagely maintain the prediction error below 0.5 °C. Compared with other models, it can reduce the errors by a rate from 5.6% to 46.8%, and shows advantages in root-mean-squared error and$R^{2}$. It also shows that the node-to-node weights for the GCN can reduce the prediction error by 11.4%.

With the popularity of Internet of Things (IoT), edge computing and cloud computing, more and more stream analytics applications are being developed including real-time trend prediction and object detection on top of IoT sensing data. One popular type of stream analytics is the recurrent neural network (RNN) deep learning model based time series or sequence data prediction and forecasting. Different from traditional analytics that assumes data are available ahead of time and will not change, stream analytics deals with data that are being generated continuously and data trend/distribution could change (a.k.a. concept drift), which will cause prediction/forecasting accuracy to drop over time. One other challenge is to find the best resource provisioning for stream analytics to achieve good overall latency. In this paper, we study how to best leverage edge and cloud resources to achieve better accuracy and latency for stream analytics using a type of RNN model called long short-term memory (LSTM). We propose a novel edge–cloud integrated framework for hybrid stream analytics that supports low latency inference on the edge and high capacity training on the cloud. To achieve flexible deployment, we study different approaches of deploying our hybrid learning framework including edge-centric, cloud-centric and edge–cloud integrated. Further, our hybrid learning framework can dynamically combine inference results from an LSTM model pre-trained based on historical data and another LSTM model re-trained periodically based on the most recent data. Using real-world and simulated stream datasets, our experiments show the proposed edge–cloud deployment is the best among all three deployment types in terms of latency. For accuracy, the experiments show our dynamic learning approach performs the best among all learning approaches for all three concept drift scenarios.

Data-driven causality discovery is a common way to understand causal relationships among different components of a system. We study how to achieve scalable data-driven causality discovery on Amazon Web Services (AWS) and Microsoft Azure cloud and propose a causality discovery as a service (CDaaS) framework. With this framework, users can easily re-run previous causality discovery experiments or run causality discovery with different setups (such as new datasets or causality discovery parameters). Our CDaaS leverages Cloud Container Registry service and Virtual Machine service to achieve scalable causality discovery with different discovery algorithms. We further did extensive experiments and benchmarking of our CDaaS to understand the effects of seven factors (big data engine parameter setting, virtual machine instance number, type, subtype, size, cloud service, cloud provider) and how to best provision cloud resources for our causality discovery service based on certain goals including execution time, budgetary cost and cost-performance ratio. We report our findings from the benchmarking, which can help obtain optimal configurations based on each application’s characteristics. The findings show proper configurations could lead to both faster execution time and less budgetary cost.

Adversarial attacks attempt to disrupt the training, retraining, and utilizing of artificial intelligence (AI) and machine learning models in large-scale distributed machine learning systems. This causes security risks on its prediction outcome. For example, attackers attempt to poison the model by either presenting inaccurate misrepresentative data or altering the models' parameters. In addition, Byzantine faults including software, hardware, network issues occur in distributed systems which also lead to a negative impact on the prediction outcome. In this paper, we propose a novel distributed training algorithm, partial synchronous stochastic gradient descent (ParSGD), which defends adversarial attacks and/or tolerates Byzantine faults. We demonstrate the effectiveness of our algorithm under three common adversarial attacks again the ML models and a Byzantine fault during the training phase. Our results show that using ParSGD, ML models can still produce accurate predictions as if it is not being attacked nor having failures at all when almost half of the nodes are being compromised or failed. We will report the experimental evaluations of ParSGD in comparison with other algorithms.

We present a framework for evaluating the performance of Byzantine fault-tolerant (BFT) protocols theoretically. Our motivation is to identify protocols suitable for a particular power grid application. In this application, replicas are located in a LAN network where latency is the priority. To fully understand the performance of BFT, we provide a generic approach that quantifies the performance of BFT protocols based on the number of cryptographic operations under five different scenarios (in the presence of failures and without failures). We present the performance of three representative BFT protocols: PBFT, Prime, and SBFT. To validate our framework, we also evaluate the protocols experimentally in the CloudLab testbed. Our experimental results match the findings predicted by the framework. Although a variety of factors may affect the performance of the protocols, our framework can be used as a valuable reference to understand the performance of BFT.

We present Chios, an intrusion-tolerant publish/subscribe system which protects against Byzantine failures. Chios is the first publish/subscribe system achieving decentralized confidentiality with fine-grained access control and strong publication order guarantees. This is in contrast to existing publish/subscribe systems achieving much weaker security and reliability properties. Chios is flexible and modular, consisting of four fully-fledged publish/subscribe configurations (each designed to meet different goals). We have deployed and evaluated our system on Amazon EC2. We compare Chios with various publish/subscribe systems. Chios is as efficient as an unreplicated, single-broker publish/subscribe implementation, only marginally slower than Kafka and Kafka with passive replication, and at least an order of magnitude faster than all Hyperledger Fabric modules and publish/subscribe systems using Fabric.