The Linda tuplespace paradigm is increasingly being used in the design of large-scale Internet-based systems. However, centralized tublespace implementations sfufer from both availability and scalability limitations. GLOBE is a distributed and replicated tuplespace that combines high availability and scalability in such heterogeneous and loosely couple systems. The level of availability and scalability of a GLOBE tuplespace can dynamically and independently be adjusted by modifying the number of replicas and partitions in the system. A tuplespace semantics is defined specifically tailored for large-scale, loosely coupled networks. The GLOBE replica update protocols are based on a combination of active operation propagation and anti-entropy propagation in the case of failures. A prototype of GLOBE has been implemented by extending a centralized JavaSpaces implementation. Evaluations of the prototype implementation indicate that distributed and replicated GLOBE tuplespace can outperform a highly loaded centralized tuplespace.

While distributed middleware technologies, such as CORBA, DCOM and Jini have enabled the development of complex software applications in varied domains, they have also presented challenges. One major challenge is middleware transparency - abstractions that capture those elements of the application specific to the middleware and allow seamless integration of the abstracted elements into an application. Aspect oriented programming (AOP) promises to make the possibility of achieving middleware transparency a reality. This paper describes our attempt to realize middleware transparency by applying AOP with AspectJ to a distributed Jini "stock broker service" application. We successfully encapsulated Jini code into a number of aspects that greatly simplified the application development process. In addition, we also implemented a number of visualization utility aspects having applicability in a variety of different applications and scenarios. We describe both the benefits and challenges of this approach.

Current mainstream distributed Java architectures offer great capabilities embracing conventional enterprise architecture patterns and designs. These traditional systems provide robust transaction oriented environments that are in large part focused on data and host processors. Typically, these implementations require that an entire application be deployed on every machine that will be used as a compute resource. In order for this to happen, the application is usually taken down, installed and started with all systems in-sync and knowing about each other. Static environments such as these present an extremely difficult environment to setup, deploy and administer.

A number of imaging systems rely on the reconstruction of a 3- dimensional model from its projections through the process of computed tomography (CT). In medical imaging, for example magnetic resonance imaging (MRI), positron emission tomography (PET), and Single Computer Tomography (SPECT) acquire two-dimensional projections of a three dimensional projections of a three dimensional object. In order to calculate the 3-dimensional representation of the object, i.e. its voxel distribution, several reconstruction algorithms have been developed. Currently, mainly two reconstruct use: the filtered back projection(FBP) and iterative methods. Although the quality of iterative reconstructed SPECT slices is better than that of FBP slices, such iterative algorithms are rarely used for clinical routine studies because of their low availability and increased reconstruction time. We used Jini and a self-developed iterative reconstructions algorithm to design and implement a Jini reconstruction service. With this service, the physician selects the patient study from a database and a Jini client automatically discovers the registered Jini reconstruction services in the department's Intranet. After downloading the proxy object the this Jini service, the SPECT acquisition data are reconstructed. The resulting transaxial slices are visualized using a Jini slice viewer, which can be used for various imaging modalities.

It is important to be able to provide a variety of ways of visualising any data set if the differences between users in terms of ability and preferences are to be successfully overcome. Human variability is a wonderful thing; except perhaps when there is a visualisation being evaluated! There is not likely to be a representation that fulfils the needs of any given group of users and still remains likeable to them all. This provided the impetus to create a visualisation framework that is able to a variety of visualisations and data through a service mechanism. This allows for matching services to be located, and utilised, as necessary, using a contain clie3nt application that acts to hold the component parts of the framework together. This provides representations each user is happiest with to be employed (of those made available), and also for a comprehensive analysis and understanding tool because of the multiple sources of data that can be integrated. In order to best exploit the possible distributed nature of visualisations and data sources, and to provide extensibility for collaboration and Grid integration the implementation make use of Jini.

This paper describes a Publish and Subscribe capability developed under the Air Force Research Laboratory s (AFRL) Joint Battlespace Infosphere (JBI) project. The paper will give a brief description of the JBI and it s core service components of publish, subscribe and query. A detailed description fo the Pub/Sub system design and implementation will then be given describing how and where Java, Jini, and XML technologies were used to describe information objects, match subscribers to appropriate dissemination nodes, and disseminate information objects to subscribing clients. Fianlly we describe a number of applications that are currently using the Pub/Sub capability.

We describe two distinct distributed systems, MeasureMe and GeoPresence. As with all distributed software systems, these require support for communication, coordination, task distribution and persistent shared data. In this paper, we show how the tuple spaces paradigm provides the required communication and coordination, and how its use facilitates our achieving scalability, extensibility, data persistency and ease of software component upgrades.

The target of ubiquitous computing environment is to support users to get necessary information and services in a situation-dependent form. Therefore, we propose a location-based information support system by using Compact Battery-less Information Terminal (CoBIT). A CoBIT can communicate with the environmental system and with the user by only the energy supply from the environment. It has a solar cell and get a modulated light from an environmental optical beam transmitter. The current from the solar cell is directly (or through passive circuit) introduced into an earphone, which generates sound for the user. The current is also used to make vibration, LED signal or electrical stimulus on the skin. The sizes of CoBITs are about 2cm in diameter, 3cm in length, which can be hanged on ears conveniently. The cost of it would be only about 1 dollar if produced massively. The CoBIT also has sheet type corner reflector, which reflect optical beam back in the direction of the light source. Therefore the environmental system can easily detect the terminal position and direction as well as some simple signs from the user by multiple cameras with infra-red LEDs. The system identifies the sign by the modulated patterns of the reflected light, which the user makes by occluding the reflector by hand. The environmental system also recognizes other objects using other sensors and displays video information on a nearby monitor in order to realize situated support.

he design, architecture, and implementation for the high-throughput data transmission and high-performance computing,which are applicable for various real-time radar signal transmission applications over the data network, are presented. With a client-server model, the multiple processes and threads on the end systems operate simultaneously and collaborately to meet the real-time requirement. The design covers the Digitized Radar Signal (DRS) data acquisition and data transmission on the DRS server end as well as DRS data receiving, radar signal parameter computation and parameter transmission on the DRS receiver end. Generic packet and data structures for transmission and inter-process data sharing are constructed. The architecture was successfully implemented on Sun/Solaris workstations with dual 750 MHz UltraSPARC-III processors containing Gigabit Ethernet card. The comparison in transmission throughput over gigabit link between with computation and without computation clearly shows the importance of the signal processing capability on the end-to-end performance. Profiling analysis on the DRS receiver process shows the work-loaded functions and provides guides for improving computing capabilities.

Today s large diversity of network resources, computing resources and heterogeneous devices can lead to a mismatch between network-capacity and the volume of data transferred, or between devices and data (consider transferring a large video clip to a handheld over a congested wireless network). Traditional distributed file systems including NFS [1], AFS [2], and Coda [3], fail to address this issue, as they cannot adapt their services according to environmental constraints. We address the problem of adaption in a DFS by modeling the file system as an object-oriented framework using persistent distributed objects. Files (persistent objects) in the DFS participate when the system reacts to environmental constraints (e.g. network congestion) and/or user-imposed constraints (e.g. security). For instance, when the user wants to play a video clip stored on a remote server with an untrusted network in between, the system chooses a version of the clip appropriate to the current network traffic, encrypts it over the wire, and plays it on the client. Further, the behavior of our general-purpose framework can also be specialized according to specific application requirements, to implement domain-specific optimizations (e.g. modifying concurrent- write semantics).

Optimizing area and speed in parallel prefix circuit have been considered important for a long time. The issue of power consumption in these circuits, however, has not been addressed. The paper presents a comparative study of different parallel prefix circuits form the point of view of power-speed trade-off. The power consumption and the power-delay product of seven parallel prefix circuits were compared. A linear output capacitance assumption, combined with PSpice simulations, is used to investigate the power consumption in the parallel prefix circuits. The degrees of freedom studied include different parallel prefix algorithms and voltage scaling. The results show that the use of the linear output capacitance assumption provides results that are consistent with those obtained using PSpice simulations. The study can help identify parallel prefix algorithms with the desirable power consumption with a given throughput.

Recently the demand of high speed and high-density optical recording media using direct overwrite scheme is very high. Among some of the potential candidates, AgInSbTe alloy appears to be one of the latest promising materials that has drawn world wide attention. The optical disks of this material with overwrite cyclability of more than 10⁵ times and data rate 22Mbps have been reported for DVD 4.5GB. Results of X-ray diffraction analysis of amorphous and crystalline (AgSbTe)_x(In_1-ySb_y)_1-x films (x = 0.2, 0.4 and y = 0.7) deposited by thermal evaporation technique are presented here. The difference in crystallization behavior of the crystalline phases formed after 1hr. of thermal annealing at temperature between 200-400°c are studied through X-ray diffraction analysis. The optical band gap of above mentioned amorphous and crystalline films were also calculated from transmittance spectra. It is observed that transmittivity increase by about 20% to give significant contrast between amorphous and crystalline marks. This relative change in transmittivity varies with chemical composition also. The results show that as the annealing temperature is incresed, film becomes more crystalline and with lower value of x, i.e. with x = 0.2 better results are obtained. These results were also confirmed through microstructural analysis of the films, involving surface detail using SEM. It has been observed that grain size depends of the annealing temperature as well as on the composition.

An optical interconnection network with parallel optical links for multiprocessor cluster system of 256 nodes is described. There are 16 subsystems in the system, in which each subsystem includes 16 computers and an optical fiver-ribbon interconnection plate of 16X16 crossbar interconnection network with VCSEL-based opto-electronic interface. An optical data rate between computers is 5Gb/s which is transmitted by the optical fiber-ribbon based parallel optical data links with 4 channels at data rates of 1.25Gb/s per channel. Every interface between optical interconnection network and each computer includes 16X4 VCSEL pixels, 16X4 PIN pixels and (1X16)X4 electrical switch. The whole optical interconnection network is composed of two level optical networks. There are sixteen optical fiber-ribbon interconnection plates of 16X16 crossbar interconnection network in the bottom level. The top level optical interconnection network would be an optical fiber-ribbon interconnection plate with a total of 2084 data channels propagating for communication of 16 subsystems.

In this paper a software method for electrical capacitance sensor design of electrical capacitance tomography system is presented by using finite element analysis techniques. Electrical capacitance sensor theory model is established, and capacitance sensitivity field distribution and structure parameters of the sensor to measurement effect are analyzed by the method. The optimized design and simulation of electrical capacitance sensor is based on the method, and performances of electrical capacitance sensor are improved clearly.

Expert system development tool ESDT-HKD is a general-purpose language for knowledge engineering. Rule plus frame plus black board is used as knowledge structure of the system. It is a perfect system implemented on personal computer. There is particular knowledge structure, and generating and running environment for users. As a tool of expert system assistant design, ESDT-HKD many original aspects such as citation of explanation framework, opened rule describing language PBRL, dynamic query of black board, maintenance of knowledge base, two-double scheduling of task and so on with which system has the complete function. This article introduces scheduling and reasoning of ESDT-HKD system tasks during consultation in details.