First-in First-out Queue Research Articles

Designing systolic, distributed buffers with bounded response time

Several buffer designs are derived by applying a design methodology that is based on so-called abstract states. Abstract states are euivalence classes of communication histories. These abstract states are very useful in the verification of program transformations, since they facilitate the definition of a function mapping the states of the transformed automaton onto the states of the original one. Three kinds of bufferes are discussed: the stack, the first-in first-out queue, and the priority queue. The designs are systolic and offer bounded response time, which means that all permissible communications are accepted within a time bounded by a constant. The design of the stack offers maximum storage utilization as well. We show that the properties of bounded response time and maximum storage utilization cannot be combined in distributed systolic queues.

VME imaging boards target machine vision: Birenbaum, R I and King, D RDigital Des. Vol 16 No 8 (May 1986) pp 28–32

In order to conduct an accurate hydrological analysis of a watershed, certain conditions need to be understood. Flow direction and flow accumulation are important watershed characteristics that need to be determined before an analysis can be made. Other important characteristics which can be gleaned from analysing the digital elevation model (DEM) of a watershed include channel networks, stream lengths and watershed boundaries. Determining flow direction and flow accumulation is usually carried out in separate steps. Flat regions are types of terrain in raster DEMs without local elevation gradients. Evaluating flow direction and flow accumulation in flat regions using DEMs is a well-known problem in watershed analysis because of the occurrence of problematic parallel flow lines. Calculations also tend to be time-consuming. We have developed an efficient and comprehensive integrated approach to assign flow directions and flow accumulation in flat regions. This approach uses values for non-flat flow accumulation and a maze algorithm with a weight value (MW method) to determine several things: a main flow line through the flat area to the local outlet, an octree tree, and first-in first-out queue structures to calculate flow accumulation. The MW method can be applied to hydrologically corrected DEMs and a single flow path can be provided to resolve all flat areas. To investigate the influence on the topological properties of the channel networks, we used this integrated algorithm to extract three sets of flow accumulation areas from existing DEMs. Using this new integrated method was faster than using the two existing methods and produced continuous channel networks without the occurrence of problematic parallel flow lines.

Algorithm and Hardware for a Merge Sort Using Multiple Processors

An algorithm is described that allows log (n) processors to sort n records in just over 2n write cycles, together with suitable hardware to support the algorithm. The algorithm is a parallel version of the straight merge sort. The passes of the merge sort are run overlapped, with each pass supported by a separate processor. The intermediate files of a serial merge sort are replaced by first-in first-out queues. The processors and queues may be implemented in conventional solid logic technology or in bubble technology. A hybrid technology is also appropriate.

Detection of auditory signals presented at random times, II

The paper includes experimental results obtained using two procedures in which signals are presented at random times. A simple three-stage theory consisting of a sensory process, followed by a response bias, and ending in a response process, involving a random delay, and that mayor may not have a memory, is compared with the data. The free-response procedure yielded data that could not be accounted for by a first version of the theory in which the crucial assumption about the response process was that once a response process is under way it locks out all further inputs until the response is made. Alternative models, including a race between response processes with the first response suppressing all others and a simple first-in first-out queue, were equally inadequate. On the assumption that the difficulties lie in the complexities of the response process when there are two or more inputs close in time, we decided to avoid these theoretical difficulties by using a modified reaction-time procedure. The initial results are encouraging. The tails of the latency distributions appear to be exponentially distributed and a theoretical prediction that two of the time constants should be the same appears to be supported. The major indications of difficulties are inconclusive evidence that the response bias parameter may not be constant and, possibly related, that the initial portion of the tails may overshoot the predicted value. In spite of the possible variation in the bias parameter, estimates of the signal-plus-noise to noise-alone parameters increase in a systematic manner with signal strength