TinyOS [1] is the current state of the art in operating systems for sensor network research. Event-based programming model of TinyOS presents concept of Task to allow postponing processing. For little processing and m...
详细信息
ISBN:
(纸本)9780769534718
TinyOS [1] is the current state of the art in operating systems for sensor network research. Event-based programming model of TinyOS presents concept of Task to allow postponing processing. For little processing and memory overhead and to avoid race conditions, tasks are non-preemptive. This causes executing long running task reduce system responsiveness. In general two approaches suggested for solving this problem: cooperative and multithreaded multitasking. In this paper we propose a new TinyOS task scheduler to integrate these approaches with new type of tasks. We argue that this approach improves the overall system responsiveness without concerning about data races or complicate programming for developers.
We report on our experience of integrating Fault-Tolerant parallel and distributed Computing (FTPDC) concepts in three core and elective courses in spring 2014 at Michigan Technological University, namely Model-Driven...
详细信息
ISBN:
(纸本)9781467376846
We report on our experience of integrating Fault-Tolerant parallel and distributed Computing (FTPDC) concepts in three core and elective courses in spring 2014 at Michigan Technological University, namely Model-Driven Software Development, Software Quality Assurance and Operating Systems. Our objective is to educate senior undergraduate Software Engineering students on the design, verification and implementation of fault tolerance in parallel and distributed Computing (PDC) systems. Towards this end, we design three course modules that include several lectures, demonstrations and homework assignments. We implement these course modules based on a discovery-based method of teaching where students are required (individually and in groups) to solve problems by modeling, simulating and verifying models of FTPDC systems. The students' performances and their feedback show that the designed course modules for the modeling and automated analysis of FTPDC have significantly helped them to understand the basic concepts of FTPDC and explore more advanced concepts. We will improve the design of the third course module to enhance students' skills in the implementation of fault-tolerance mechanisms.
A reconfigurable architecture using distributed logic block processing elements (PEs) is presented. This distributed processor uses a low-cost interconnection network and local indirect VLIW memories to provide effici...
详细信息
ISBN:
(纸本)354067442X
A reconfigurable architecture using distributed logic block processing elements (PEs) is presented. This distributed processor uses a low-cost interconnection network and local indirect VLIW memories to provide efficient algorithm implementations for portable battery operated products. In order to provide optimal algorithm performance, the VLIWs loaded to each PE configure that PE for processing. By reloading the local VLIW memories, each PE is reconfigured for a new algorithm. Different levels of flexibility are feasible by varying the complexity of the distributed PEs in this architecture.
We are developing a compiler that translates ordinary MATLAB scripts into code suitable for compilation and execution on parallel computers supporting C and the MPI message-passing library. In this paper we report the...
详细信息
ISBN:
(纸本)0818684038
We are developing a compiler that translates ordinary MATLAB scripts into code suitable for compilation and execution on parallel computers supporting C and the MPI message-passing library. In this paper we report the speedup achieved for several MATLAB scripts on three diverse parallel architectures: a distributed-memory multicomputer (Meiko CS-2), a symmetric multiprocessor (Sun Enterprise Server 4000), and a cluster of symmetric multiprocessors (Sun SPARCserver 20s). By generating code suitable far execution on parallel computers, our system multiplies the gains achievable by compiling, rather than interpreting, MATLAB scripts. Generating parallel code has an additional advantage: the amount of primary memory available on most parallel computers makes it possible to solve problems too large to solve on a single workstation.
As comparison tools, BinDiff, PatchDiff2, Darun-Grim, eEye Binary Diffing Suite provide diffing functionality for binary codes files. In this paper, we propose one framework DCDFNE for building dataset. The built data...
详细信息
ISBN:
(纸本)9781665414852
As comparison tools, BinDiff, PatchDiff2, Darun-Grim, eEye Binary Diffing Suite provide diffing functionality for binary codes files. In this paper, we propose one framework DCDFNE for building dataset. The built dataset could be used in two aspects: evaluate the performance of the binary codes comparison tools;train deep learning models which predict the binary codes are similar or not. DCDFNE includes four sections: the first section collects information about multiple projects from GitHub, the collected information is stored in the database which includes source codes of different versions and commit information between parent and child versions;the second section implements the function of distributed compilation of vim source codes;the third section extracts the differential function name which represents the modification between the two versions of the commit files;the fourth section combines the results of the second section and the third section, namely the binary codes and the differential function names, then extracts the binary codes of the differential functions. The experiment results show that the distributed compilation algorithm improves 99.80% compared to serial compilation algorithm;the extraction method extracts the differential function names effectively.
In this paper, we introduce a memory storage scheme allowing conflict-free parallel access to rows, columns, square blocks, distributed blocks, and positive and negative diagonals of two dimensional arrays. Unlike the...
详细信息
ISBN:
(纸本)081864222X
In this paper, we introduce a memory storage scheme allowing conflict-free parallel access to rows, columns, square blocks, distributed blocks, and positive and negative diagonals of two dimensional arrays. Unlike the existing schemes, the proposed scheme can be used for an arbitrary number of memory modules and an arbitrary size of the arrays. We develop a systematic procedure for the memory allocation based on a placement matrix constructed using circulant matrices.
In this paper we propose a mechanism for improving the performance of name resolution operations in distributed file systems. The mechanism is based on the idea of reducing the number of message exchanges required for...
详细信息
ISBN:
(纸本)081864222X
In this paper we propose a mechanism for improving the performance of name resolution operations in distributed file systems. The mechanism is based on the idea of reducing the number of message exchanges required for the resolution of a particular file name. In this mechanism, users are given the flexibility to dynamically define and change their performance requirements for the various file names being used by them.
The workstation model of parallelprocessing presents specific challenges caused by the latency of the communications network and the workload imbalance that arises from the heterogeneity of the nodes. Data partitioni...
详细信息
ISBN:
(纸本)081864222X
The workstation model of parallelprocessing presents specific challenges caused by the latency of the communications network and the workload imbalance that arises from the heterogeneity of the nodes. Data partitioning is critically important for parallel processsing in this environment. We mathematically characterize the communication costs for four data decomposition schemes: scatter, contiguous point, contiguous row, and block. These methods are analyzed in terms of problem size, number of processors, network speed, and communication pattern. Bounds are established for the performance of these decomposition schemes that can be used to make better-informed data partitioning decisions.
The focus of the research work presented in this paper is on theoretical and practical aspects of communication optimization in distributed systems. Communication is a key issue in every distributed system and, since ...
详细信息
ISBN:
(纸本)9781424437504
The focus of the research work presented in this paper is on theoretical and practical aspects of communication optimization in distributed systems. Communication is a key issue in every distributed system and, since good communication performance is often difficult to achieve, optimization techniques are welcome, if not even strictly required. We present both the achieved results and future research plans.
暂无评论