COM-0992: PIXSAR: Disk reclustering for faster XML databases PIXSAR is an algorithm that is capable of efficiently rearranging an XML-based database. It does so by treating database clustering as an augmented tree partitioning problem and adjusting the data physical layout according to the dynamic workload. iPIXSAR, an extension, is able to do so in the presence of multiple-indexes over the database. Read more...
COM-1030: Efficient and general Voronoi diagrams algorithm Voronoi diagrams are a powerful computational tool. They appear in many areas in science and technology, and have diverse applications. One can find the diagrams and their applications in chemistry, computer science, geography, economics, communication, art, materials engineering and more. Our algorithm is simple and efficient, and significantly improves current calculation capabilities of existing algorithms. For example, it can compute Voronoi diagrams of general sites in general normed spaces (of any dimension), in manifolds and in other settings. In addition, it supports parallel computing. Read more...
COM-1035: Decision support tool for project management Despite the common use of project management software, many projects still fail to meet their goals on time and within budget. This is partly because these tools do not allow the project team to actively experiment with "what-if" scenarios during the course of the project. Our technology allows for the interactive simulation of stochastic events on real-world projects, enabling the management team or students to discuss, analyze and find more optiomal solutions while developing better judgment. Read more...
COM-0712: GeomCore - outstanding 3D graphics GeomCore offers a whole framework that allows three-dimensional geometry to be presented over the World Wide Web. The geometry is presented in an efficient way and allows for a high degree of interaction and collaboration, all in real time. GeomCore’s technology can serve as the basis for numerous applications such as MCAD 3D visualization over the Web, 3D icons and collaborative 3D games. Read more...
COM-1104: Ternary content-addressable memory with fast error detection Ternary Content-Addressable Memory (TCAM) devices are increasingly being used for performing high-speed classification of data packets required by networking applications such as routing, monitoring and security. The parallel-access properties of TCAM devices make them prone to errors which standard error detection schemes are not designed to handle. Our technology is a parallel error detection scheme which is optimized for TCAM. It allows an order of magnitude reduction in the number of lookups required for the error location process, thereby increasing performance and reliability. Read more...
COM-1103: Complex barycentric coordinates Barycentric coordinates are heavily used in computer graphics applications to generalize a set of given data values. Traditionally, the coordinates are required to be real and positive. We relax this requirement by allowing the barycentric coordinates to be complex numbers which lead to several powerful advantages. These new coordinates are highly effective in generating conformal mappings from a given planar source region to a planar target region. Complex barycentric coordinates can be used for image manipulation, shape deformation, animation, data interpolation, color interpolation and parameterization. Read more...
COM-1163: Optimal inspection of multidimensional regions Securing a complex terrain or mold designs are examples of seemingly unrelated geometric problems that deal with region /object coverage
Our novel algorithm can compute an (almost) optimal solution to the complete coverage of multidimensional region. The algorithm has two major stages. In the first, visibility is determined from a set of positions. In the second, an approximation to the set-cover problem is computed. Read more...
COM-1030: Efficient and general Voronoi diagrams algorithm Voronoi diagrams are a powerful computational tool. They appear in many areas in science and technology, and have diverse applications. One can find the diagrams and their applications in chemistry, computer science, geography, economics, communication, art, materials engineering and more. Our algorithm is simple and efficient, and significantly improves current calculation capabilities of existing algorithms. For example, it can compute Voronoi diagrams of general sites in general normed spaces (of any dimension), in manifolds and in other settings. In addition, it supports parallel computing. Read more...
COM-1104: Ternary content-addressable memory with fast error detection Ternary Content-Addressable Memory (TCAM) devices are increasingly being used for performing high-speed classification of data packets required by networking applications such as routing, monitoring and security. The parallel-access properties of TCAM devices make them prone to errors which standard error detection schemes are not designed to handle. Our technology is a parallel error detection scheme which is optimized for TCAM. It allows an order of magnitude reduction in the number of lookups required for the error location process, thereby increasing performance and reliability. Read more...
COM-0712: GeomCore - outstanding 3D graphics GeomCore offers a whole framework that allows three-dimensional geometry to be presented over the World Wide Web. The geometry is presented in an efficient way and allows for a high degree of interaction and collaboration, all in real time. GeomCore’s technology can serve as the basis for numerous applications such as MCAD 3D visualization over the Web, 3D icons and collaborative 3D games. Read more...
COM-0992: PIXSAR: Disk reclustering for faster XML databases PIXSAR is an algorithm that is capable of efficiently rearranging an XML-based database. It does so by treating database clustering as an augmented tree partitioning problem and adjusting the data physical layout according to the dynamic workload. iPIXSAR, an extension, is able to do so in the presence of multiple-indexes over the database. Read more...
COM-1035: Decision support tool for project management Despite the common use of project management software, many projects still fail to meet their goals on time and within budget. This is partly because these tools do not allow the project team to actively experiment with "what-if" scenarios during the course of the project. Our technology allows for the interactive simulation of stochastic events on real-world projects, enabling the management team or students to discuss, analyze and find more optiomal solutions while developing better judgment. Read more...
COM-1103: Complex barycentric coordinates Barycentric coordinates are heavily used in computer graphics applications to generalize a set of given data values. Traditionally, the coordinates are required to be real and positive. We relax this requirement by allowing the barycentric coordinates to be complex numbers which lead to several powerful advantages. These new coordinates are highly effective in generating conformal mappings from a given planar source region to a planar target region. Complex barycentric coordinates can be used for image manipulation, shape deformation, animation, data interpolation, color interpolation and parameterization. Read more...
COM-1163: Optimal inspection of multidimensional regions Securing a complex terrain or mold designs are examples of seemingly unrelated geometric problems that deal with region /object coverage
Our novel algorithm can compute an (almost) optimal solution to the complete coverage of multidimensional region. The algorithm has two major stages. In the first, visibility is determined from a set of positions. In the second, an approximation to the set-cover problem is computed. Read more...
COM-1150: Monitoring threshold functions over multiple nodes in distributed data sets Various systems and applications perform monitoring of data streams in a distributed environment. Such applications include, for example, sensor networked, distributed web-sites, etc. Existing methods for data stream monitoring require high computational costs when monitoring non-linear functions. They are also limited in the number of distributed nodes and monitored objects. This method employs a novel geometric approach to monitor the data stream at much greater efficiency and with many fewer restrictions. Read more...
COM-0812: Monitoring multi-variate threshold functions over distributed data streams A common requirement in many emerging applications is the ability to process, in real time, a continuous high-volume stream of data. Consequently, data stream monitoring in a distributed system is the focus of much research in recent years. Examples of such applications are sensor networks, real-time analysis of financial data and intrusion detection. We present a novel geometric approach utilizing naive, centralized algorithms for efficient arbitrary threshold functions monitoring over distributed data streams without requiring the very high communication overhead common in other methods. Read more...
COM-1282: Mitigation of inter-cell coupling effects in NAND flash memory NAND flash memory is a popular data storage facility at various electronic devices. However, it suffers from an inter-cell coupling effect which reduces the memory functionality. An innovative method improves the NAND flash memory operations by diminish the inter-cell coupling effect. Read more...
COM-1394: Aquatic event detection algorithm Identifying if and when a contamination event in a water distribution system occurred is of great importance. Available software are either inflexible, lack probability updates according to existing conditions or "black boxed", and therefore hard to adapt and customize for a specific water distribution system. We propose a novel algorithm that can be utilized by water utilities worldwide as an add-on to existing on-line monitoring systems. The novelty of the system is threefold: By using a generic formulation of the problem the system may be used with various inputs; Very promising results are obtained by updating event probabilities as additional information is obtained; The univariate and multivariate functions in the algorithm provide a unique, user friendly and precise evaluation of real time status of the water distribution system. Read more...
COM-1351: Learning adaptive personalized terms semantic relatedness measurement from examples Semantic relatedness is the quantification of the intensity with which two objects are related to each other. In recent years, the problem of automatically determining semantic relatedness has been steadily gaining attention among statistical natural language processing and artificial intelligence researchers. This method is a novel structure –free supervised approach to learning semantic relatedness from examples. With this model, semantic relatedness learning is treated as a binary classification problem where each instance encodes the relative relatedness of two term pairs. Read more...
COM-1342: Supply Chain Simulator Management of supply chains is a major issue in operations, and teaching supply chain management and supporting decisions in this area is a real need for many businesses as well as academic institutions. However, no analytical model can express the true richness of any real world supply chain problem. Available models either have a very limited range of scenarios or are so all-inclusive that they lose the focus on supply chains. We have designed software that enables simulating complicated supply chain models and can interface with existing software as well as generate scenarios based on real supply chains stored in the system. The simulation allows students on the one hand, and business users (like inventory managers) on the other hand to gain a deeper understanding of how supply chains work. Read more...
COM-1451: Parallelized decision forest builder for streaming multiprocessors Very large data sets with large numbers of attributes are currently only able to be processed on expensive and complex computing clusters. The present technology is a new method which takes advantage of the modern large (1GB+) onboard GPU device memory as well as large onboard CPU RAM in order to map the building of a single randomized decision tree to a GPU, multiple GPUs or multiple CPUs. The technology can be applied to very large datasets of training data on commodity hardware for a fraction of the startup and runtime costs required to perform the same learning method on a cluster. Read more...
COM-1343: Accelerating flash memory performance by speculative early sensing Flash memory is a non-volatile memory component, widely used in modern electronic systems. The growing demand for high performance computing along with high capacity data storage stimulates flash technology development. The increase in capacity, often accompanied by performance reduction (due to the use of multiple charge levels), makes increasing performance a critical task. Current methods used to accelerate flash memory performance induce errors in the data. The technique presented uses a novel algorithm to exploit the actual capabilities of each part of the memory chip at any given time, thereby performing significantly accelerated read, write and erase operations without inducing undetected errors. Yet, implementation of the technique requires no changes in the core technology or fabrication process. Read more...
COM-1488: Storage system with inexpensive disk add-ons for inter- and intra-disk redundancy Contemporary storage systems often maintain at least three replicas of all data, reducing the risk of permanent data loss due to simultaneous disk failures. The cost of the additional copies is a reduction in usable storage space, with increased network traffic and higher power consumption. The presented technology employs an inexpensive add-on device attached to each disk, enabling storage systems to withstand simultaneous failures while requiring only two data replicas and minimizing cost, network, and power overheads. Read more...
COM-1483: Retired page utilization for improved write capacity of solid state drives NAND Flash is presently the most prominent solid-state memory technology. While SSD offer higher performance and other advantages over mechanical magnetic hard disk drives (HDD), their per-bit cost is much higher. Current methods to reduce cost however sharply degrade the, namely the tolerable number of erasures of a cell, thereby limiting product lifetime. The presented technology is an innovative memory management scheme that significantly boosts current write-capacity enhancing schemes with negligible capacity and performance overhead. This is done by managing “where” to write in addition to “what” to write. Specifically, RPU adds spatial consideration to any write-once memory coding. Read more...
COM-1401: A parallel twig join algorithm for XML processing using a GPGPU Current algorithms to match XML query twig patterns in a parallel GPU multi-threaded computing platform are only fit for systems with a maximum of several dozen threads. The technology presented is a novel algorithm for matching twig patterns in large XML documents, using GPU. The algorithm efficiently exploits both the high-bandwidth GPU memory interface and the lower-bandwidth CPU main memory and fits for systems with many hundreds of threads. The invention can be used to upgrade database systems that support XML documents to work in multi-core GPU environments and can significantly improve their performance. Read more...
COM-1482: Performance enhancing multi-level-cell memory architecture via minimal maximum-level programming NAND Flash is currently the most prominent non-volatile semiconductor memory technology, used mostly for storage. Both Flash and Phase-Change Memory employ multi-level cells (MLC), while beneficial in terms of storage capacity and cost per bit, MLC comes at a performance penalty. The technology presented is an innovative memory architecture, composed of encoder, decoder and address-to-cells mapping. The technology improves the best known average page write time, has an accelerated read time, and is proven to achieve the upper bound of write speed. The proposed scheme does not require data redundancy, reduces energy consumption and is suitable to every product that employs Flash or Phase Change memory. Read more...
COM-1484: Continuous flow multithreading architecture (CFMT Architecture) Switch on Event Multithreading is a low power process but suffers from performance drawbacks due to the relatively high thread switch penalty. Simultaneous Multithreading has limitations in the number of threads running in the machine due to its complexity. The presented technology is a new processor architecture that can use SRAM or emerging memory technologies to enhance the performance of the processor and support more threads while consuming significantly less power and with no other drawbacks. This architecture can be used in any general purpose processors suitable for mobile products, personal computers, embedded system and servers. Read more...
COM-1468: Processing query to graph database The presented technology is a new method for matching graph database query tree patterns using General-Purpose GPU that allows massive parallelization of the task and hence a quick response. Any graph database provider can use this algorithm as a parallel query processor core. It can also serve as a core technology for designing extremely fast utilities for graph databases using SIMD technology. Read more...
COM-1352: 3D presentation software Many industries are intensive users of 3D. However, their inability to present 3D information in their everyday workplace and outside their natural 3D working environment (i.e. a CT scanner) is a significant operational and financial drawback. Current 2D presentation tools are less efficient and are prone to misunderstandings and errors of misinterpretation, when 3D content is involved; this has a direct and substantial economic implication on these industries. Currently, presentations with 3D content are typically constructed by specialized production teams that are extremely expensive. The technology presented is of a multi-modalities presentation software that will also provide the average user with the ability to create 3D presentations with 3D content at a fraction of the cost and will allow the presenter to annotate, transform and animate, and arbitrarily manipulate (i.e. slice) 3D content in a seamless manner within the presentations. Read more...
COM-0944: XPathL – logical language for integrating relational and XML databases The joint processing of XML and relational data is of major importance in information and knowledge management. Many database management systems support XML data, which is the de facto standard for data representation and exchange over the web. Our method handles conjunctive queries, comprising relational predicates and external predicates, such as XML predicates. The language can be utilized as a target language for higher level constructs or as an embedded query language. Read more...
COM-1204: Multi-core XML query processing Multi-core XML processing is a method that allows better processing performance of an XML Twig Pattern query. The known processing method use sequential algorithms that deal with a single query processing, but since modern computers have multi-core processors we managed to optimize the processing performance of a twig pattern XML query by parallelizing two existing sequential algorithms and in this way create new novel and efficient algorithms . Read more...
