The proceedings contain 8 papers. The topics discussed include: TardisTM: incremental repair for transactional memory;bounded incoherence: a programming model for non-cache-coherent shared memory architectures;ELSE: a...
ISBN:
(纸本)9781450375221
The proceedings contain 8 papers. The topics discussed include: TardisTM: incremental repair for transactional memory;bounded incoherence: a programming model for non-cache-coherent shared memory architectures;ELSE: an efficient link-time static instrumentation tool for embedded system;self-adjusting task granularity for global load balancer library on clusters of many-core processors;towards a portable hierarchical view of distributed shared memory systems: challenges and solutions;lock-free transactional vector;exploring accelerator and parallel graph algorithmic choices for temporal graphs;and generating energy-efficient code for parallel applications specified by streaming task graphs with dynamic elements.
This paper proposes a distributed rolling horizon optimization framework for energy trading problems among multiple microgrids (MMGs). The rolling horizon optimization based on model predictive control (MPC) method al...
详细信息
ISBN:
(纸本)9781665425582
This paper proposes a distributed rolling horizon optimization framework for energy trading problems among multiple microgrids (MMGs). The rolling horizon optimization based on model predictive control (MPC) method alleviates impacts of uncertainties in renewable energy resources and loads while the independent decision making for MMGs can be guaranteed by distributed algorithms in realtime electricity markets. A mixed linear programming model with rolling timing windows is developed to describe energy transactions among MMGs. In order to enable different entities to meet their energy requirement closer to real-time operation, online alternative direction method of multipliers (ADMM) is implemented to solve the given model, in which coupled variables from energy transactions are separated by consistency constraints. In rolling timing windows, operating statuses could be updated based on real-time information to improve the precision of energy transactions. Case simulations based on different methods are presented to demonstrate the effectiveness of the proposed method in coordinating energy transactions among MMGs under uncertainties.
With the rapid development of edge computing technology, the application of edge computing in smart grids has become more and more extensive. But edge computing has not yet been applied to the operation control of dis...
详细信息
With the rapid development of edge computing technology, the application of edge computing in smart grids has become more and more extensive. But edge computing has not yet been applied to the operation control of distributed power generation microgrid systems. This article proposes a microgrid-oriented edge computing architecture. First, we introduce the main functions of edge-cloud collaboration. Then we explain the construction plan of the architecture, including the realization of data processing, network communication and security mechanisms. Finally, we introduce the architecture application practice in a rural community in Central China.
This article proposes a washout filter-based distributed resilient current control scheme that can achieve current sharing of distributed generations (DG) and voltage restoration of dc bus almost surely under communic...
详细信息
This article proposes a washout filter-based distributed resilient current control scheme that can achieve current sharing of distributed generations (DG) and voltage restoration of dc bus almost surely under communication channel noise interferences. Since the cyber channels are exposed to inherent noise interferences, the stability and reliability of cyber-physical dc microgrids may terribly be reduced. To dispel the adverse influences of noise interferences, a novel discrete time washout filter is developed. Accordingly, a discrete-time-distributed noise-resiliency control algorithm is developed for dc microgrids based on washout filters, in which only the current state variable is allowed to exchange among DGs with the local decentralized voltage controller. This makes the proposed control mode different from the conventional parallel mode usually requiring a leader–follower consensus-based voltage control by exchanging two state variables (i.e., terminal voltage and current output). Thus, the proposed control mode decreases the complexity to the formulation and enhances the robustness to the leader communication failure. By introducing stochastic theory and Lyapunov stability function, the sufficient conditions considering noise interferences is derived to ensure the stability operation of the whole closed-loop dc network. The obtained results show the effectiveness of the proposed strategy by a tested dc microgrid in OPAL-RT real-time simulator. IEEE
We proposed a new automatic speech recognition (ASR) service architecture that is extendable to medium-scale ASR service and more flexible than the previous architecture. Improvement aims to substitute the distributed...
详细信息
We proposed a new automatic speech recognition (ASR) service architecture that is extendable to medium-scale ASR service and more flexible than the previous architecture. Improvement aims to substitute the distributed processing approach with an asynchronous parallel thread for decoding multiple voice streams. We replace our TCP-based communication protocol with a remote procedure call developed by Google (gRPC) that makes our ASR service become a developer-friendly, less overhead connection. Besides, the API gateway is employed to reinforce the ASR services by multiple servers so that we can increase our new ASR service to a larger scale. The experimental result shows that our new architecture performs faster than the previous architecture in terms of real-time factor.
This paper proposes a communication strategy for decentralized learning on wireless systems. Our discussion is based on the decentralized parallel stochastic gradient descent (D-PSGD), which is one of the state-of-the...
详细信息
ISBN:
(纸本)9781728150895
This paper proposes a communication strategy for decentralized learning on wireless systems. Our discussion is based on the decentralized parallel stochastic gradient descent (D-PSGD), which is one of the state-of-the-art algorithms for decentralized learning. The main contribution of this paper is to raise a novel open question for decentralized learning on wireless systems: there is a possibility that the density of a network topology significantly influences the runtime performance of DPSGD. In general, it is difficult to guarantee delay-free communications without any communication deterioration in real wireless network systems because of path loss and multi-path fading. These factors significantly degrade the runtime performance of D-PSGD. To alleviate such problems, we first analyze the runtime performance of D-PSGD by considering real wireless systems. This analysis yields the key insights that dense network topology (1) does not significantly gain the training accuracy of D-PSGD compared to sparse one, and (2) strongly degrades the runtime performance because this setting generally requires to utilize a low-rate transmission. Based on these findings, we propose a novel communication strategy, in which each node estimates optimal transmission rates such that communication time during the D-PSGD optimization is minimized under the constraint of network density, which is characterized by radio propagation property. The proposed strategy enables to improve the runtime performance of D-PSGD in wireless systems. Numerical simulations reveal that the proposed strategy is capable of enhancing the runtime performance of D-PSGD.
parallel filesystems (PFSs) are one of the most critical high-availability components of High Performance Computing (HPC) systems. Most HPC workloads are dependent on the availability of a POSIX compliant parallel fil...
详细信息
ISBN:
(纸本)9781728166773
parallel filesystems (PFSs) are one of the most critical high-availability components of High Performance Computing (HPC) systems. Most HPC workloads are dependent on the availability of a POSIX compliant parallel filesystem that provides a globally consistent view of data to all compute nodes of a HPC system. Because of this central role, failure or performance degradation events in the PFS can impact every user of a HPC resource. There is typically insufficient information available to users and even many HPC staff to identify the causes of these PFS events, impeding the implementation of timely and targeted remedies to PFS issues. The relevant information is distributed across PFS servers;however, access to these servers is highly restricted due to the sensitive role they play in the operations of a HPC system. Additionally, the information is challenging to aggregate and interpret, relegating diagnosis and treatment of PFS issues to a select few experts with privileged system access. To democratize this information, we are developing an opensource and user-facing parallel FileSystem TRacing and Analysis SErvice (PFSTRASE) that analyzes the requisite data to establish causal relationships between PFS activity and events detrimental to stability and performance. We are implementing the service for the open-source Lustre filesystem, which is the most commonly used PFS at large-scale HPC sites. Server loads for specific PFS I/O operations (IOPs) will be measured and aggregated by the service to automatically estimate an effective load generated by every client, job, and user. The infrastructure provides a real-time, user accessible text-based interface and a publicly accessible web interface displaying both real-time and historical data.
Graph computing recently receives intensive interests due to a wide range of needs to analyse big graphs in realtime. Many single-machine out-of-core graph computing systems have been developed to process large-scale...
详细信息
Graph computing recently receives intensive interests due to a wide range of needs to analyse big graphs in realtime. Many single-machine out-of-core graph computing systems have been developed to process large-scale graphs by utilizing effective memory access techniques. The current systems either focus on graph compression or pay attention to disk scheduling strategies. In this paper, we propose an efficient and unified out-of-core graph computing model on a single machine named D 2 Graph which combines the differential storage strategy and the dynamic caching mechanism. D 2 Graph improves the spatial locality of graph data and the temporal locality of graph computing at the same time. The main contributions of our work lie in three aspects: firstly, we design a fine-grained differential storage(FGDS) strategy to reorganize graph data on disk; secondly, we propose a dynamic caching mechanism (DCM) to selectively load the necessary edge chunks; thirdly, we combine FGDS and DCM to construct D 2 Graph and evaluate it on five public graph data sets. What’s more, to optimize the performance of D 2 Graph, we also make use of the multi-threaded parallel technology. A series of results show that D 2 Graph outperforms state-of-the-art out-of-core graph computing systems by 1.3x-17.9x.
Industry 4.0 and the vision of smart factories drive the need for real-time communication. time-Sensitive Networking (TSN) augments the IEEE Std 802.1Q with a family of mechanisms enabling real-time communication. One...
详细信息
ISBN:
(数字)9781728152974
ISBN:
(纸本)9781728152974
Industry 4.0 and the vision of smart factories drive the need for real-time communication. time-Sensitive Networking (TSN) augments the IEEE Std 802.1Q with a family of mechanisms enabling real-time communication. One of the key mechanisms is the time-Aware Shaper (TAS) implementing a TDMA scheme on a traffic class basis. With proper synchronization it can even be used to schedule individual frames or streams. With this capability, the network can guarantee communication deadlines, bounded latency, and bounded jitter. However, for these guarantees a system-wide schedule needs to be calculated, which is an NP-hard problem. Current approaches are mainly based on constraint programming and optimization problems, and, therefore do not scale well for larger topologies and number of streams. In this paper, our contribution is twofold: first, we propose a scheduling model for converged networks supporting different traffic types and, secondly, we introduce a novel procedure for schedule planning of isochronous traffic which exploits the hierarchical structure of factory networks. To this end, we split the network into sub-networks and use a two-stage approach based on a heuristic and tracing. Our evaluation shows that the new scheduling approach outperforms the reference scheduler by more than two orders of magnitude with regard to execution time.
暂无评论