This paper treats the data routing problem for fault-tolerant systolic arrays based on Triple Modular Redundancy (TMR) in mixed spatial-temporal domain. The number of logical links required in TMR systolic array is basically 9 times larger than the one for corresponding non-fault-tolerant systolic array. The link sharing is a promising method for reducing the number of physical links, which may, however, degrade the fault tolerance of TMR system. This paper proposes several robust data-routing and resource-sharing (plural data transfers share a physical link, or a data transfer and a computational task share a PE as a relay node for the former and as a processor for the latter), by which certain classes of fault tolerant property will be guaranteed. A stage and a dominated set are introduced to characterize the features of routing/resource-sharing in TMR systems, and conditions on the dominated set and their resultant fault-tolerant properties are derived.

Due to a large capacitance and enormous access rate, caches dissipate about a third of the total energy consumed by today's processors. In this paper we present a new architectural technique to reduce energy consumption in caches. Unlike previous approaches, which have focused on lowering cache capacitance and the number of accesses, our method exploits a new freedom in cache design, namely the voltage per access. Since in modern block-buffered caches, the loading capacitance operated on block-hit is much less than the capacitance operated on miss, the given clock cycle time is inefficiently utilized during the hit. We propose to trade-off this unused time with the supply voltage, lowering the voltage level on the hit and increasing it on the miss. Experiments show that the approach can half the cache energy dissipation without large performance and area overhead.

In this paper, a design of Programmable Logic Device (PLD) and a synthesis approach are proposed. Our PLD is derived from traditional Programmable Logic Array (PLA). The key extension is that programmable AND devices in PLA is replaced by Look-Up Tables (LUTs). A series of cascaded LUTs in the array can generate more complex terms, which we call generalized complex terms (GCTs), than product terms. In order to utilize the capability, a synthesis approach to map a given function into the array is also proposed. Our approach generates a expression of the sum of GCTs aiming to minimize the number of terms. A number of experimental results demonstrate that the number of terms for our PLD generated by our approach is 14.9% fewer than that by an existing approach. We design our PLD based on a fundamental unit named nGCT cell which can be used as LUTs in multiple sizes or random access memories. Implementation of the PLD based on a fundamental unit named nGCT cell which can be used as LUTs or random access memories is also described.

Today, people want highly-customized products to satisfy their individual requirements. However traditional manufacturing technology is not geared towards high-mix, low-volume manufacturing. Holonic Manufacturing Systems (HMS) is a new paradigm to bridge this divide. HMS offers enterprises a new breed of technology to continuously reconfigure themselves to manufacture a larger variety of products in smaller batch sizes, and do this profitably. A suitable metaphor for implementing the holonic manufacturing system is the emerging IEC function block architecture. The paper describes how function blocks can be used to build such holonic manufacturing systems. We also illustrate the merits of our approach through a real-world engine assembly line being developed by DaimlerChrysler.

A method is described that can allocate bandwidth to each user flow fairly in a scalable network architecture such as differentiated services architecture. As promising queueing techniques for providing differentiated services, class-based packet scheduling and selective packet discarding have been attracting attention. However, if we consider that bandwidth should be allocated to each flow in a weighted manner, the parameters used in these methods such as the weight assigned to each class queue should be pre-determined appropriately based on an assumption about the number of flows in each class. Thus, when the actual traffic pattern differs from the assumed one, they may not work well. Instead of assuming the traffic conditions, our method estimates the number of active flows in each class by simple traffic measurement and dynamically changes the weight assigned to each class queue based on the estimated number. Our method does not need to maintain the per-flow state, which gives it scalability. Simulation showed that this method is effective under various patterns of the number of active flows.

Generally, the wavelet transform is applied for watermarking, where a whole image is transformed to the frequency domain. However, many schemes which adopt the wavelet transform are weak for StirMark attack designed to affect locally such as rotation, extension, reduction, etc. In this paper, we propose a new scheme which applies the wavelet transform locally and has a tolerance for StirMark attack. How to maintain the quality of the watermarked image is also shown.

In this paper, two efficient codebook searching algorithms for vector quantization (VQ) are presented. The first fast search algorithm utilizes the compactness property of signal energy of orthogonal transformation. On the transformed domain, the algorithm uses geometrical relations between the input vector and codeword to discard many unlikely codewords. The second algorithm, which transforms principal components only, is proposed to alleviate some calculation overhead and the amount of storage. The relation between the principal components and the input vector is utilized in the second algorithm. Since both of the proposed algorithms reject those codewords that are impossible to be the nearest codeword, they produce the same output as conventional full search algorithm. Simulation results confirm the effectiveness of the proposed algorithms.

Over recent years, "Internet-able" applications and architectures have been used to support domains where there are multiple interconnected systems that are both decentralized and autonomous. In enterprise-level data management domains, both the schema of the data repository and the individual query needs of the users evolve over time. To handle this evolution, the resulting architecture must enforce the autonomy in systems that support the client needs and constraints, in addition to maintaining the autonomy in systems that support the actual data schema and extraction mechanisms. At the MITRE Corporation, this domain has been identified in the development of a composite data repository for the Center for Advanced Aviation System Development (CAASD). In the development of such a repository, the supporting architecture includes specialized mechanisms to disseminate the data to a diverse evolving set of researchers. This paper presents the motivation and design of such an architecture to support these autonomous data extraction environments. This run-time configurable architecture is implemented using web-based technologies such as the Extensible Markup Language (XML), Java Servlets, Extensible Stylesheets (XSL), and a relational database management system (RDBMS).

When we search from a huge amount of documents, we often specify several keywords and use conjunctive queries to narrow the result of the search. Though the searched documents contain all keywords, positions of the keywords are usually not considered. As a result, the search result contains some meaningless documents. It is therefore effective to rank documents according to proximity of keywords in the documents. This ranking is regarded as a kind of text data mining. In this paper, we propose two algorithms for finding documents in which all given keywords appear in neighboring places. One is based on plane-sweep algorithm and the other is based on divide-and-conquer approach. Both algorithms run in O(n log n) time where n is the number of occurrences of given keywords. We run the algorithms on a large collection of html files and verify its effectiveness.

Multi-domain spoken dialogue systems with high degree of intelligence and domain extensibility have long been desired but difficult to achieve. When the user freely surfs among different topics during the dialogue, it will be very difficult for the system to control the switching of the topics and domains while keeping the dialogue consistent, and decide when and how to take the initiative. This paper presents a distributed agent architecture for multi-domain spoken dialogue systems with high domain extensibility and intelligence. Under this architecture, different spoken dialogue agents (SDA's) handling different domains can be developed independently, and then smoothly cooperate with one another to achieve the user's multiple goals, while a user interface agent (UIA) can access the correct spoken dialogue agent through a domain switching protocol, and carry over the dialogue state and history so as to keep the knowledge processed coherently across different domains.

A new method for object search is proposed. Conventional template matching schemes tend to fail in presence of irregularities and ill-conditions like background variations, illumination fluctuations resulting from shadowing or highlighting etc. The proposed scheme is robust against such irregularities in the real world scenes since it is based on matching gradient information around each pixel, computed in the form of orientation codes, rather than the gray levels directly. A probabilistic model for robust matching is given and verified by real image data. Experimental results for real world scenes demonstrate the effectiveness of the proposed method for object search in the presence of different potential causes of mismatches.

The Active HYpermedia Delivery System (AHYDS) facilitates the access to multimedia information over a large-scale network and wide spectrum of media. We developed intelligent access facilities that build on the access paradigms supported by current web applications. This facility generalizes not only different kinds of logical data models (relational, object, hyperlink), but also access mechanisms of multimedia applications to make them customizable and scalable. This paper proposed the distributed management mechanism of the AHYDS platform. The major contribution of this paper is the mechanism for distributed multimedia delivery management over large-scale network and heterogeneous environment. We also propose the mechanism to manage huge multimedia data.

This paper presents a new network architecture called LINA that provides node mobility. The basic concept of LINA is separation of the node identifier and the interface locator. Although there are several protocols based on such a concept, they do not address issues that arise when dealing with an entire network architecture. LINA is a holistic architecture covering the network layer to the application layer in order to support node mobility. Overhead incurred by separation of the node identifier and the interface locator is minimized in LINA by introducing the embedded addressing model. This paper also presents a new protocol called LIN6 that supports IPv6 mobility. LIN6 is an application of LINA to IPv6 and is fully compatible with IPv6. It has several advantages in comparison to Mobile IPv6, e.g. less protocol overhead. Our prototype implementation of LIN6 shows minimal overhead compared to a conventional IPv6 implementation.

In this letter, a new scheme of designing multilevel BTC coding is proposed. The optimal quantization can obtain by selecting the quantization threshold with an exhaustive search. However, it requires an enormous amount of computation and is thus impractical while we take an exhaustive search for the multilevel BTC. A two-steps searching method is applied to reduce the computational complexity. Comparisons of results with various methods have verified that the proposed method approach to the optimal quantization with little computation complexity.

Magnetic random access memory (MRAM) possesses the attractive properties of non-volatility, radiation hardness, nondestructive readout, low voltage, high access speed, unlimited read and write endurance and high density. MRAM technology is described for the devices using giant magnetoresistance (GMR) and tunneling magnetoresistance (TMR) materials in this paper. The TMR type MRAM architectures using ferromagnetic tunneling junctions (MTJ) are more attractive for mainstream RAM applications than the GMR type, because the signal of the TMR type is larger than that of the GMR type. A MRAM device with an MTJ plus MOS transistor switch architecture, which can provide large signal-to noise ratio, is detailed. A design of the MTJ element is discussed and the requirements for the junction resistance and the TMR needed for the memory device are demonstrated based on the simple signal voltage calculations. The TMR significantly decreases with increasing bias voltage, which leads to the reduction of the signal voltage for the actual MRAM. A ferromagnetic double tunneling junction is proposed for the high density MRAM application demanding large signal voltage, because of the smaller degradation of the TMR for the bias voltage, compared with that of the conventional single junctions. Recent trials of MRAM fabrication are introduced, which demonstrates high-speed access time. Finally, challenges for the higher bit density MRAM above Gb are discussed, and it is noticed that higher signal voltage, lower power consumption for writing and novel cell designs are needed for the achievement.

Access control plays an important role in the area of information security, which guarantees that any access to data is authorized. Hierarchical access control is a special access control model in distributed environment, in which each user protects his local data using a secret key; moreover, for any two related users ui and uj, ui can access uj's data if, and only if, ui's priority is higher than uj. Therefore, there should be a way for ui to obtain the secret key of uj if ui's priority is higher than uj. This paper presents an efficient solution to the problem. A special kind of function called secure filter is used as the building block of the proposed solution. In the solution, an authorized user can acquire correct keys efficiently and securely via secure filters. The proposed solution is also well-performed while inserting/deleting users, injecting/removing relations, and changing secret keys. Especially, only deleting users and removing relations will change some keys in the system, other operations can be performed freely without affecting other keys in the system; only secure filters need to be modified in these cases.

We propose the problem of how to transmit an information-theoretically secure bit using random deals of cards among players in hierarchical groups and a computationally unlimited eavesdropper. A player in the highest group wants to send players in lower groups a secret bit which is secure from the eavesdropper and some other players. We formalize this problem and design protocols for constructing secret key exchange spanning trees on hierarchical groups. For each protocol we give sufficient conditions to successfully construct a secret key exchange spanning tree for the hand sizes of the players and the eavesdropper.

In this paper, we investigate the performance characteristics of parallel switching architectures constructed by a stack of multistage switching networks. We first find that the performances of the previously proposed parallel switching architectures are much worse than the expected ones from analytic models which are based on the assumption that traffic is uniformly distributed at each stage of a switching network. We show that this phenomenon is closely related to a traffic-distribution capability of a parallel switching system and has a large influence on the performance. From these results, we then propose an architectural solution based on the Generalized Shuffle Network (GSN) and analyze its performance by proposing a new iterative analysis method. The proposed architecture uses self-routing and deflection routing, and inherently has a traffic-distribution capability to improve switch performances such as cell loss and delay in a cost-effective manner. From the comparison of simulation and analysis results, it is shown that the developed models are quite accurate in predicting the performance of a new parallel switching system.

This paper describes simple and efficient (linear-preserving) reductions between the Decision Diffie-Hellman problem and related problems.

This paper describes the design, implementation and testing of wideband code division multiple access (CDMA) base station demodulator for the international mobile telecommunication-2000 (IMT-2000) system test plant based on cdma2000 radio transmission technology (RTT). The performance of the implemented base station demodulator is measured and compared with the theoretical performance bound. The system test plant equipped with this demodulator provides wireless services, such as high quality speech (9.6 kbps), real-time video (384 kbps) and internet protocol (IP) based data services (144 kbps) in a mobile radio environment.