A framework for estimating the relative execution time of a data-parallel algorithm in an environment capable of the SIMD and SPMD (Single Program - Multiple Data) modes of computation is presented. Given a data-paral...
For real-time radar processing, it is very desirable to have an algorithm that does not assume restricted statistics of the input data and can be implemented for high-speed processing (without a high cost) to meet rea...
详细信息
For real-time radar processing, it is very desirable to have an algorithm that does not assume restricted statistics of the input data and can be implemented for high-speed processing (without a high cost) to meet real-time requirements. We therefore apply the QR decomposition-based least-squares method for linear prediction to the problem of computing the reflection coefficients of a lattice predictor, instead of using the conventional Burg algorithm. We also propose a modified one-dimensional ring architecture for implementing the QR method of least-squares. The particular application considered in this case is that of surveillance radar systems for air traffic control.< >
The paper presents a parallel natural language processing system implemented on a marker-passing parallel AI computer, Semantic Network Array Processor (SNAP). The system uses a memory-based parsing approach in which ...
详细信息
The paper presents a parallel natural language processing system implemented on a marker-passing parallel AI computer, Semantic Network Array Processor (SNAP). The system uses a memory-based parsing approach in which parsing is viewed as a memory search process. Linguistic information is stored as phrasal patterns in a semantic network knowledge base distributed over the memory of the parallelcomputer. Parsing is performed by recognizing and linking linguistic patterns that reflect a sentence interpretation. This is achieved via propagating markers over the distributed network. The authors have developed a system capable of processing newswire articles from a particular domain. The paper presents the structure of the system, the memory-based parsing method used, and performance results obtained.< >
A technique that greatly simplifies the computational complexity of digital cyclic spectral analysis algorithms is presented. The technique, which is based on Bussgang's theorem, replaces complex multiplications i...
详细信息
A technique that greatly simplifies the computational complexity of digital cyclic spectral analysis algorithms is presented. The technique, which is based on Bussgang's theorem, replaces complex multiplications in spectral correlation operations with simple sign-change and data-multiplexing operations. Moreover, the technique is applicable to both time- and frequency-averaging algorithms. A simulation study that compares the computed results obtained using the new technique with results from standard time- and frequency-averaging algorithms shows that the new technique is very promising, particularly for frequency-averaging algorithms.
An analytical study of potential pathological performance areas of the Seamless architecture is presented. Seamless is a latency-tolerant, distributed memory, multiprocessor architecture. A key component of the philos...
详细信息
An analytical study of potential pathological performance areas of the Seamless architecture is presented. Seamless is a latency-tolerant, distributed memory, multiprocessor architecture. A key component of the philosophy of Seamless, however, is the use of standard, commodity components for a large part of the system. A discussion of the unavoidable implementation compromises imposed by this decision is presented, followed by a summary of some optimistic performance studies. Then an analytical study that parameterizes the predicts the worst-case impact of using standard components is provided. Finally, it is shown that these bottlenecks are manageable via careful generation of target machine code so that the optimistic performance studies become realistic expectations for a range of program behaviors and granularities.< >
An algorithm for converting relay ladder logic (RLL) programs for programmable logic controllers (PLCs) into sequential function chart (SFC) programs is introduced. Since an SFC (a standardization of Grafcet) better r...
详细信息
The mapping problem arises when the dependency structure of a parallel algorithm differs from the interconnection of processors in the intended parallelcomputer (topological variation) or, when the number of processe...
详细信息
The mapping problem arises when the dependency structure of a parallel algorithm differs from the interconnection of processors in the intended parallelcomputer (topological variation) or, when the number of processes required by the algorithm exceeds the number of processors available in the computer (cardinality variation). The problem discussed in this paper is to identify a distributed computing environment that best optimizes the objective function for the given problem. Distinct network permutations may result in equivalent process permutation depending upon the mapping of processes onto the processors. In this paper we study the permuting properties of dynamic interconnection networks taking process mapping into consideration. A uniform group theoretic representation for interconnection networks is developed. Finally, an algorithm to evaluate the number of passes required by an interconnection network to realize a given mapping is presented.< >
Linear filters can be designed to any frequency Specification but they exhibit smoothing of localized signal features. In this paper a new filtering scheme is introduced, which is frequency selective but at the same t...
详细信息
Linear filters can be designed to any frequency Specification but they exhibit smoothing of localized signal features. In this paper a new filtering scheme is introduced, which is frequency selective but at the same time does not blurr the signal. This scheme is based on using the ordering of samples together with frequencey specifications. Frequency specifications are used to detect the desired frequencies. Information about the samples ordering is then used to decide which filter or if any filter is to be applied on the signal.
According to the Floquet theory, an nth-order linear periodic (LP) system of the form yn+αn(t) y/sup n-1/+...+α2(t)dy(t)/dt+α1(t)y=0 can be transformed into an equivalent linear time-invariant (LTI) system whose ch...
详细信息
暂无评论