state-of-the-art method for several artificial intelligence domains in recent years. The increasingly complex CNN models are both computation-bound and I/O-bound. Fieldprogrammable gate array-based accelerators driven by custom instruction set architecture (ISA) achieve a balance between generality and efficiency, but there is much on them left to be optimized. We propose the full-stack compiler deep neural network virtual machine (DNNVM), which is an integration of optimizers for graphs, loops and data layouts, an assembler, a runtime supporter, and a validation environment. The DNNVM works in the context of deep learning frameworks and transforms CNN models into the directed acyclic graph: XGraph. Based on XGraph, we transform the optimization challenges for both data layout and pipeline into graph-level problems. DNNVM enumerates all potentially profitable fusion opportunities by a heuristic subgraph isomorphism algorithm to leverage pipeline and data layout optimizations, and searches for the best choice of execution strategies of the whole computing graph. On the Xilinx ZU2@330 MHz and ZU9@330 MHz, we achieve equivalently state-of-the-art performance on our benchmarks by naïve implementations without optimizations, and the throughput is further improved up to 1.26× by leveraging heterogeneous optimizations in DNNVM. Finally, with ZU9@330 MHz, we achieve state-of-the-art performance for VGG and ResNet50. We achieve a throughput of 2.82 TOPs/s and an energy efficiency of 123.7 GOPs/s/W for VGG. Additionally, we achieve 1.38 TOPs/s for ResNet50 and 1.41 TOPs/s for GoogleNet.

In recent years, Convolutional Neural Network(CNN) is becoming the state-of-the-art method in a wide range of Artificial Intelligence(AI) domains. The increasingly large and complex CNN models are both computation bound and I/O bound. FPGA-based accelerators driven by custom Instruction Set Architecture(ISA) achieve a balance between generality and efficiency, and leave much room for optimization. Operation fusion which fuses adjacent operations without saving intermediate results back to off-chip DDR can greatly alleviate bandwidth pressure, operations can be executed by different computation engines concurrently for latency hiding. To leverage optimizations, especially operation fusion on custom instruction-based accelerators, we propose a full-stack compiler DNNVM(Deep Neural Network Virtual Machine). DNNVM is an integration of optimizers for framework-independent computing graph, loops and data layouts, an assembler, a runtime supporter and a validation environment. DNNVM works in the context of deep learning frameworks and transforms CNN models into a directed acyclic graph, XGraph. After analyzing the interaction among fusion depth, tiling across multiple stages and on-chip memory capacity, DNNVM enumerates all potentially profitable fusion opportunities according to custom fusion templates upon XGraph, by a subgraph isomorphism algorithm. In addition, DNNVM searches for the optimal execution strategies by a heuristic shortest-path algorithm. On Xilinx [email protected], we achieve up to 1.26x speedup than naïve implementations without fusion on GoogLeNet. On Xilinx [email protected], we achieve the throughput of 2.82 TOPs/s for VGG, 1.38 TOPs/s for ResNet50 - he fastest ever reported on comparable FPGAs.

Multi-tenancy in the cloud is a double-edged sword. While it enables cost-effective resource sharing, it increases security risks for the hosted applications. Indeed, multiplexing virtual resources belonging to different tenants on the same physical substrate may lead to critical security concerns such as cross-tenants data leakage and denial of service. Particularly, virtual networks isolation failures are among the foremost security concerns in the cloud. To remedy these, automated tools are needed to verify security mechanisms compliance with relevant security policies and standards. However, auditing virtual networks isolation is challenging due to the dynamic and layered nature of the cloud. Particularly, inconsistencies in network isolation mechanisms across cloud-stack layers, namely, the infrastructure management and the implementation layers, may lead to virtual networks isolation breaches that are undetectable at a single layer. In this article, we propose an offline automated framework for auditing consistent isolation between virtual networks in OpenStack-managed cloud spanning over overlay and layer 2 by considering both cloud layers’ views. To capture the semantics of the audited data and its relation to consistent isolation requirement, we devise a multi-layered model for data related to each cloud-stack layer’s view. Furthermore, we integrate our auditing system into OpenStack, and present our experimental results on assessing several properties related to virtual network isolation and consistency. Our results show that our approach can be successfully used to detect virtual network isolation breaches for large OpenStack-based data centers in reasonable time.

Cloud computing is emerging as a promising IT solution for enabling ubiquitous, convenient, and on-demand accesses to a shared pool of configurable computing resources. However, the widespread adoption of cloud is still being hindered by the lack of transparency and accountability, which has traditionally been ensured through security auditing techniques. Auditing in cloud poses many unique challenges in data collection and processing (e.g., data format inconsistency and lack of correlation due to the heterogeneity of cloud infrastructures), and in verification (e.g., prohibitive performance overhead due to the sheer scale of cloud infrastructures and need of runtime verification for the dynamic nature of cloud). To this end, existing runtime auditing techniques do not offer a practical response time to verify a wide-range of user-level security properties for a large cloud. In this paper, we propose a runtime security auditing framework for the cloud with special focus on the user-level including common access control and authentication mechanisms e.g., RBAC, ABAC, SSO, and we implement and evaluate the framework based on OpenStack, a widely deployed cloud management system. The main idea towards reducing the response time to a practical level is to perform the costly operations only once, which is followed by significantly more efficient incremental runtime verification. Our experimental results show that runtime security auditing in a large cloud environment is realistic under our approach (e.g., our solution performs runtime auditing of 100,000 users within 500 milliseconds).

Cloud computing has seen a lot of interests and adoption lately. Nonetheless, the widespread adoption of cloud is still being hindered by the lack of transparency and accountability, which has traditionally been ensured through security compliance auditing techniques. Auditing in cloud, however, presents many new challenges in data collection and processing (e.g., data format inconsistency and lack of correlation due to the heterogeneity of cloud infrastructures) and in verification (e.g., prohibitive performance overhead due to the sheer scale of cloud infrastructures and their self-provisioning, elastic, and dynamic nature). In this paper, we propose a security compliance auditing framework for cloud, with special focus on identity and access management, and we implement and evaluate the framework based on OpenStack, one of the most popular cloud management systems. Our experimental results show that auditing with formal methods in large cloud environment is realistic (e.g., our auditing solution can handle 60 thousand users in less than one minute).

This paper investigates the problem of state feedback H∞control for singular systems through delta operator approach. Firstly, a bounded real lemma corresponding to a singular continuous system under the framework of the delta operator model is obtained. Then, the existence condition and explicit expression of a desirable H∞controller for the singular delta operator system are presented. As special cases, the results of admissible control for the singular delta operator system and H∞control as well as admissible control for the singular continuous system are also derived. All required conditions in this paper are characterized in the form of strict linear matrix inequalities whose feasible solutions can be obtained easily and directly. Finally, some numerical examples are provided to illustrate the effectiveness of the obtained theoretical results in the paper.

Face image partially occluded or damaged can be repaired automatically. We proposed a new inpainting algorithm, based on patch guidance deduced from an existing face database, to recover the damaged portions. This newly proposed concept of guided inpainting method produces seamless faces which are hardly seen drawbacks. Examples of our results can be retrieved from http://member.mine.tku.edu.tw/www/ISCAS09/.

The problem of parallel machine multi-objective scheduling subjected to special process constraint in the textile industries, as one of the most important combinational optimization problems, is different from other parallel machine scheduling problems in the following characteristics. On one hand, processing machines are non-identical; on the other hand, the sort of job processed on every machine can be restricted Considering one of the multi-objective problems, either minimizing the maximum completion time among all the machines(makespan) or minimizing the total earliness/tardiness penalty of all the jobs has been cornerstone of most studies done so far. However, under special process constraint, taking them into account as a multi-objective problem has not been well studied Therefore, in this paper, a multi-objective model based on them is presented and a new parallel genetic algorithm based on a vector group coding method is also proposed in order to effectively solve this model. The algorithm shows the following advantages: the coding method is simple and can effectively reflect the virtual scheduling policy, which can vividly reflect the numbers and sequences of these processed jobs on every machine, and then enables the individuals generated by crossover and mutation to satisfy process constraint. Numerical experiments show that it is efficient, and is better than the common genetic algorithm, and has the better parallel efficiency. A much better prospect of application can be optimistically expected.