Broadcast is becoming an increasingly attractive data dissemination method for large client populations. In order to effectively utilize a broadcast medium for sucha service, it is necessary to have efficient, on-line scheduling algorithms that can balance individual and overall performance, and can scale in terms of data set sizes, client populations, and broadcast bandwidth. We propose an algorithm, called RxW, that provides good performance across all of these criteria and that can be tuned to trade o# average and worst case waiting time. Unlike previous work on lowoverhead scheduling, the algorithm does not use estimates of the access probabilities of items, but rather, it makes scheduling decisions based on the current queue state, allowing it to easily adapt to changes in the intensity and distribution of the workload. We demonstrate the performance advantages of the algorithm under a range of scenarios using a simulation model and present analytical results that describe the intrinsic behavior of the algorithm.

Many solid modeling applications require information not only about the geometry of an object but also about its topology. Most interchange formats do not provide this information, which the application must then derive as it builds its own topological data structure from unordered, "polygon soup" input. For very large data sets, the topological data structure itself can be bigger than core memory, so that a naive algorithm for building it that doesn't take virtual memory access patterns into account can become prohibitively slow due to thrashing. In this paper, we describe a new out-of-core algorithm that can build a topological data structure efficiently from very large data sets, improving performance by two orders of magnitude over a naive approach.

The paper proposes designs of interconnection networks (graphs) which can tolerate link failures. The networks under study belong to a subclass of Cayley graphs whose generators are subsets of all possible transpositions. We specifically focus on star and bubble-sort networks. Our approach is to augment existing dimensions (or generators) with one or more dimensions. If the added dimension is capable of replacing any arbitrary failed dimension, it is called a wildcard dimension. It is shown that, up to isomorphism among digits used in labeling the vertices, the generators of the star graph are unique. The minimum number of extra dimensions needed to acquire i wildcard dimensions is derived for the star and bubble-sort networks. Interestingly, the optimally augmented star network coincides with the Transposition network, Tn. Transposition networks are studied rigorously. These networks are shown to be optimally fault-tolerant. Tn is also shown to possess wide containers with short length. Fault-diameter of Tn is shown to be n. While the Tn can efficiently embed star and bubble-sort graphs, it can also lend itself to an efficient embedding of meshes and hypercubes.

Solid freeform fabrication (SFF) refers to a class of technologies used for making rapid prototypes of 3-D parts. With these processes, a triangulated boundary representation of the CAD model of the part is "sliced" into horizontal 2.5-D layers of uniform thickness that are successively deposited, hardened, fused, or cut, depending on the particular process, and attached to the layer beneath. The stacked layers form the final part. The current de facto standard interface to these machines, STL, has many shortcomings. We have developed a new "Solid Interchange Format" (SIF) for use as a digital interface to SFF machines. SIF includes constructs for specifying surface and volume properties, precision information, and transmitting unevaluated Boolean trees. We have also develope...

This paper describes algorithms for key deletion in B+-trees. There are published algorithms and pseudocode for searching and inserting keys, but deletion, due to its greater complexity and perceived lesser importance, is glossed over completely or left as an exercise to the reader. To remedy this situation, we provide a well documented flowchart, algorithm, and pseudo-code for deletion, their relation to search and insertion algorithms, and a reference to a freely available, complete B+-tree library written in the C programming language.

Hashing is a fundamental technique in Computer Science, and perfect hashing, which guarantees constant-time lookup, has particular theoretical and practical importance. The algorithm of Fredman, Koml'os, and Szemer'edi [8], augmented by Dietzfelbinger, et al. [7], is a simple perfect hashing scheme that has not seen extensive use due to poor practical performance. In this paper, we develop modifications to the previous techniques that result in a fast, space-efficient algorithm. We show the resulting algorithm is comparable to hash algorithms used in several popular dictionary libraries but is more flexible and has better theoretical performance. 1

Broadcast is becoming an increasingly attractive data dissemination method for large client populations. In order to effectively utilize a broadcast medium for such a service, it is necessary to have efficient, on-line scheduling algorithms that can balance individual and overall performance, and can scale in terms of data set sizes, client populations, and broadcast bandwidth. We propose an algorithm, called RxW, that provides good performance across all of these criteria and that can be tuned to trade off average and worst case waiting time. Unlike previous work on low overhead scheduling, the algorithm does not use estimates of the access probabilities of items, but rather, it makes scheduling decisions based on the current queue state, allowing it to easily adapt to changes in the intensity and distribution of the workload. We demonstrate the performance advantages of the algorithm under a range of scenarios using a simulation model and present analytical results that describe the ...

Abstract. In the past, several attempts have been made to deal with the state space explosion problem by equipping a depth-first search (DFS) algorithm with a state cache, or by avoiding collision detection, thereby keeping the state hash table at a fixed size. Most of these attempts are tailored specifically for DFS, and are often not guaranteed to terminate and/or to exhaustively visit all the states. In this paper, we propose a general framework of hierarchical caches which can also be used by breadth-first searches (BFS). Our method, based on an adequate sampling of BFS levels during the traversal, guarantees that the BFS terminates and traverses all transitions of the state space. We define several (static or adaptive) configurations of hierarchical caches and we study experimentally their effectiveness on benchmark examples of state spaces and on several communication protocols, using a generic implementation of the cache framework that we developed within the CADP toolbox. 1

In this paper, we present a parallel SIMD algorithm for sorting of N numbers of log N bits each on a mesh connected array without comparators. While most previous AT 2 -optimal sorting algorithms on the mesh explicitly or implicitly assume O#1# time comparison of two operands of O#log N# bits, our algorithm does not require O#1# time comparison. Rather we assume O#log N# time comparison, which makes our algorithm realizable with current VLSI technology. To retain the AT 2 optimality with increased ( by a factor of O#log N# ) comparison time, we develop a new mapping technique which combines radix sort, shear sort, block merge and column sort in a creative way. The time complexity of sorting of N numbers on a (two-dimensional) mesh of size N 1=2 #N 1=2 without comparotors is O#N 1=2 #. 1 Introduction The problem of sorting numbers on a two-dimensional array has been extensively studied and the lower bound is achieved for the mesh connected parallel computer[3, 4, 6, 8] an...

We consider radix sorting problems on two-dimensional mesh and three-dimensional mesh. By employing the radix sorting, we removed the necessity of comparator...