A virtualization layer makes it possible to compose multiple functionalities on a multi-core processor with minimum modifications of OS kernels and applications. A multi-core processor is a good candidate to compose various software independently developed for dedicated processors into one multi-core processor to reduce both the hardware and development cost. In this paper, we present SPUMONE, which is a virtualization layer suitable for developing multi-core processor based-information appliances.

Embedded systems are becoming increasingly sophisticated and there exists a wide variety of requirements such as traditional realtime requirements, multimedia support, etc. It is hard to satisfy all of the requirements by a single OS. Would they be satisfied, the system would become complex and this would cause new problems. A multi OS environment is an efficient approach to deal with these problems and to satisfy complex requirements while keeping the system simple. We propose a multi OS environment named the SIGMA system targeted especially at multiprocessor architectures. On the SIGMA system, guest OSes are corresponded one-to-one with cores. As a result, opposing to existing multi OS environment using virtualization techniques, the system does not degrade the performance of the guest OSes. In addition, the guest OS running on the SIGMA system requires almost no modification to its source code.

Future ambient intelligence environments will embed powerful multi-core processors to compose various functionalities into a smaller number of hardware components. This makes the maintainability of intelligent environments better because it is not easy to manage massively distributed processors.
A composition kernel makes it possible to compose multiple functionalities on a multi-core processor with the minimum modification of OS kernels and applications. A multi-core processor is a good candidate to compose various software developed independently for dedicated processors into one multi-core processor to reduce both the hardware and development cost. In this paper, we present SPUMONE which is a composition kernel for developing future smart products.

This paper proposes SPLiT (Scalable Performance Library Tool) as the methodology to improve performance of applications on multicore processors, through CPU, and cache optimizations. SPLiT analyzes applications based on cycle counts, and cache misses, and predicts behavior of the applications according to the analysis. With these acquired knowledge on the target application, SPLiT improves CPU resource allocation determined by operating systems.
SPLiT, and SPLiT library are designed to achieve resource optimizations based on the combination of hardware information collected by Operating System kernel, and software information collected by applications. Additionally, all the features of SPLiT are immediately available with the small modifications to the application source codes, and the modifications are basically to implant SPLiT library calls into the source code. This simple requirement diminishes programmers' difficulties for performance tuning of multicore applications. Through empirical experiments, we validated the efficiency of SPLiT, and observed that performance of the web application was improved by 26%.

The recent increase in complexity and functionality in embedded systems makes them more vulnerable to rootkit-type attacks, raising the need for integrity management systems. However, as of today there is no such system that can guarantee the system's safety while matching the low-resource, real-time and multi-core requirements of embedded systems. In this paper, we present a Virtual Machine Monitor (VMM) based monitoring service for embedded systems that checks the actual kernel data against a safe data specification. However, due to the VMM and multi-core nature of the system, the guest OS can be preempted at any time, leading to the checking of potentially inconsistent states. We evaluated two approaches to solve this problem: detecting such invalid states by checking specific kernel data, and detecting system calls using the VMM. © 2010 IEEE.

In this paper, Anomaly Detection by Resource Monitoring (Ayaka), a novel lightweight anomaly and fault detection infrastructure, is presented for Information Appliances.
Ayaka provides a general monitoring method for detecting anomalies using only resource usage information on systems independent of its domain, target application and programming languages. Ayaka modifies the kernel to detect faults and uses a completely application black-box approach based on machine learning methods. It uses the clustering method to quantize the resource usage vector data and learn the normal patterns with Hidden Markov Model. In the running phase, Ayaka finds anomalies by comparing the application resource usage with learned model. The evaluation experiment indicates that our prototype system is able to detect anomalies, such as SQL injection and buffer overrun, without significant overheads.

In this paper we describe our approach to improve dependability of a commodity OS for embedded systems. Usually it is too difficult for end-users to resolve the problem inside a single OS, especially for embedded systems. We propose a self-healing mechanism for Linux kernel to improve the system dependability without any operations by administrators. This paper presents our white box approach for monitoring and recovering Linux kernel. Key components are a system monitor and a virtual machine monitor. The system monitor is used to detect the inconsistency of data structures inside Linux kernel. The virtual machine monitor provides a multi-OS environment and it isolates the system monitor from Linux kernel. In a multi-OS environment, the system monitor is able to resolve failures inside Linux kernel without stopping crucial services running on another OS. We have developed a prototype for an embedded system to verify our approach. The experiment results show that our system can remove hidden processes and reload buggy kernel modules. The performance evaluation results show that our self-healing mechanism can be used even when Linux kernel is heavily-loaded and the overhead of the system monitor is vanishingly small in actual use. Copyright 2009 ACM.

Constructing an embedded device with a real-time and a general-purpose operating system has attracted attention as a promising approach to let the device balance real-time responsiveness and rich functionalities. This paper introduces our methodology for constructing such multi-OS platform with minimal engineering cost by assuming asymmetric OS combinations unique to embedded systems. Our methodology consists of two parts. One is a simple hypervisor for multiplexing resources to be shared between operating systems. The other is modifying operating systems to allow them to be aware of each other. We constructed an experimental system executing TOPPERS and Linux simultaneously on a hardware equipped with an SH-4A processor. The modification to each operating system kernel limited to a few dozen lines of code and do not introduce any overhead that would compromise real-time responsiveness or system throughput.

In this paper, a novel lightweight anomaly and fault detection infrastructure called Anomaly Detection by Resource Monitoring (Ayaka) is presented for Information Appliances.
Ayaka provides a general monitoring method for detecting anomalies using only resource usage information on systems independent of its domain, target application, and programming languages. Ayaka modifies the kernel to detect faults and uses a completely application black-box approach based on machine learning methods. It uses the clustering method to quantize the resource usage vector data and learn the normal patterns with a hidden Markov Model. In the running phase, Ayaka finds anomalies by comparing the application resource usage with the learned model. The evaluation experiment indicates that our prototype system is able to detect anomalies, such as SQL injection and buffer overrun, without significant overheads.

This paper explores system issues for enabling incremental deploy- ment of pervasive application - the problem of how to deploy and gradually enhance the functionalities of applications in a pervasive environment. We present a system architecture, FedNet that pro- vides the foundation for incremental deployment and uses a docu- ment centric approach utilizing a profile based artefact framework and a task based application framework. Our artefact framework represents an instrumented physical artefact as a collection of ser- vice profiles and expresses these services in generic documents. Pervasive applications are expressed as a collection of functional tasks (independent of the implementation) in a corresponding doc- ument. A runtime component provides the foundation for mapping these tasks to the corresponding service provider artefacts. This mapping is spontaneous and thus enables gradual addition of ser- vices. Primary advantages of our approach are twofold- firstly, it allows end users to deploy pervasive applications and instrumented artefacts easily and gradually. Secondly, it allows developers to write applications in a generic way regardless of the constraints of the target environment. We describe an implemented prototype of FedNet, and show examples of its use in a real life deployment by the end users to illustrate its feasibility.

The emergence of functional embedded systems such as cell-phones and digital appliances brought up a new issue, building a system supporting both real-time and rich services. One of the solutions is leveraging a hypervisor to integrate an RTOS and a commodity OS into a single device. However, this approach induces the limitation of application deployment: all the high priority tasks should reside in the RTOS.
In this paper, we propose a task grain scheduling for a real-time hypervisor, which enables a flexible application deployment between an RTOS and a commodity OS. We constructed a prototype system with an existing hypervisor an RTOS. and a commodity OS. We measured some basic overheads, and fixed some tasks which were missing their deadlines using the task grain scheduling to meet their deadlines. The overheads were small and the task grain scheduling achieved a flexible real-time scheduling for the hypervisor based system.

In this paper a lightweight system level monitoring infrastructure known as Kernel Objects Monitoring Infrastructure (KOMI) is presented for commercial-off-the-shelf (COTS) embedded systems. The kernel objects consist of certain critical kernel data structures and entry points of system calls, which are protected as first-class objects inside the system. KOMI provides specific runtime protections to different kernel objects: kernel data structures are protected by the periodic detection and recovery, the interception of arguments is used to protect vulnerable system calls. Both protection methods can provide not only consistency regulations but also recovery actions for the system. During its runtime deployment, once any system inconsistency has been detected, predefined recovery actions will be invoked. Since KOMI requires few modifications to kernel source code, it is easy to integrate into existing embedded systems. The evaluation experiment results indicate our prototype system can correctly detect the inconsistent kernel data structures caused by security attacks and also prevent kernel from exploits due to vulnerable system calls with acceptable penalty to the system performance. Moreover KOMI is fully software-based without introducing any specific hardware and requires no modifications to system call APIs, therefore legacy applications can be also easily reused.

An upcoming trend for automobile manufacturers is to provide firmware updates over the air (FOTA) as a service. Since the firmware controls the functionality of a vehicle, security is important. To this end, several secure FOTA protocols have been d veloped. However, the secure FOTA protocols only solve the security for the transmission of the firmware binary. Once the firm ware is downloaded, an attacker could potentially modify its contents before it is flashed to the corresponding ECU's ROM. Thus, there is a need to extend the flashing procedure to also verify that the correct firmware has been flashed to the ECU. We present a framework for self-verification of firmware updates over the air. We include a verification code in the transmission to the vehicle, and after the firmware has been flashed, the integrity of the memory contents can be verified using the verification code. The verification procedure entails only simple hash functionnsand is thus suitable for the limited resources in the vehicle. Virtualization techniques are employed to establish a truste computing base in the ECU, which is then u ed to perform the verification. The proposed framework allows the ECU it selfto perform self-verification and can thus ensure the successful flashing of the firmwa e © 2008 IEEE.

Device drivers are the major cause of operating system failure. Prior research proposed frameworks to improve the reliability of device drivers by means of driver restart. While avoiding any instrumentation of the driver, this approach does not always allow graceful recovery. In this paper, we propose a framework for self-healing device drivers that lets the driver developer consider and implement the failure recovery of device drivers. For this purpose, our framework provides easy to use and light-weight persistent memory that preserves the state of the driver needed to successfully recover We developped a prototype on top of the L4 microkernel, and were able to achieve full recovery of crashed drivers as fast as 0.2 ms for different device drivers. In all cases, recovery was totally transparent for the user.

A software platform for developing future information appliances requires to satisfy various diverse requirements. The operating system architecture presented in this paper enhances the flexibility and dependability through virtualization techniques. The architecture allows a system to use multiple operating systems simultaneously, and to use multi-core processors in a flexible way. Also, dependability mechanisms in our architecture will avoid crashing or hanging a system as much as possible in order to improve the user experience when defects in the software are exposed. We present a brief overview of each component in the operating system architecture and some sample scenarios that illustrate the effectiveness of the architecture.

The kernel objects consist of critical kernel data structures and system call functions, which are, the most important data for a system, should be protected as first-class candidates. In this paper, a lightweight system-level detection and recovery infrastructure is presented for embedded systems. Inside the infrastructure, specific runtime protections have been implemented for different kernel objects, kernel data structures are protected by the periodic detection and recovery, the interception of arguments is used to protect vulnerable system calls. At runtime once any system inconsistency has been detected, predefined recovery actions will be invoked. The consistency detection regulations and corresponding recovery actions can also be flexibly customized by system developers. The infrastructure requires few modifications to kernel source code, thus it is easy to integrate into existing embedded systems. The evaluation experiment results indicate our prototype system can correctly detect the inconsistent kernel data structures caused by security attacks and also prevent kernel from exploits due to vulnerable system calls with acceptable penalty, to system performance. Moreover it is fully software-based without introducing any specific hardware and requires no modifications to system call APIs, therefore legacy commercial-off-the-shelf (COTS) applications can be also easily reused.

This paper explores system issues for involving end users in constructing and enhancing a smart home. In support of this involvement we present an infrastructure and a tangible deployment tool. Active participation of users is essential in a domestic environment as it offers simplicity, greater usercentric control, lower deployment costs and better support for personalization. Our proposed infrastructure provides the foundation for end user deployment utilizing a loosely coupled framework to represent an artefact and its augmented functionalities. Pervasive applications are built independently and are expressed as a collection of functional tasks. A runtime component, FedNet maps these tasks to corresponding service provider artefacts. The tangible deployment tool uses FedNet and allows end users to deploy and control artefacts and applications only by manipulating RFID cards. Primary advantages of our approach are two-fold. Firstly, it allows end users to deploy ubicomp systems easily in a Do-it-Yourself fashion. Secondly, it allows developers to write applications and to build augmented artefacts in a generic way regardless of the constraints of the target environment. We describe an implemented prototype and illustrate its feasibility in a real life deployment session by the end users. Our study shows that the end users might be involved in deploying future ubicomp systems if appropriate tools and supporting infrastructure are provided. Copyright 2008 ACM.

Sentient artefacts are everyday objects augmented with digital services. Being the building blocks of our surroundings, these artefacts can incrementally integrate computing into environment and can convert it into an intelligent one economically. In this paper, we report our experiences with sentient artefacts to rationalize intelligent environment. We discuss the design principles of sentient artefacts and present a sensor selection framework to convert a regular artefact into a sentient one. In addition, we discuss the application development guideline and lessons that we have learnt through prototyping sentient artefacts and proactive applications.

Sentient artefacts are everyday objects augmented with value added digital services. Being the building blocks of our surroundings, these artefacts can incrementally integrate computing into environment and can convert it into an intelligent one economically. In this paper, we report our experiences with sentient artefacts to rationalise intelligent environment. We discuss the design principles of sentient artefacts and present a sensor selection framework to convert a regular artefact into a sentient one. In addition, we discuss the application development guideline and lessons that we have learnt through prototyping sentient artefacts and proactive applications.

This paper proposes a technique to achieve asymmetric multiple OSes environment for symmetric multiprocessors. The system has a host OS and guest OSes: L4 microkernel and their servers run as the host OS, and modified Linux runs as the guest OS. OS manager which is one of the servers on the host OS manages the guest OSes. Our approach avoids a lot of execution overheads and modification costs of the guest OSes because the each OS can control hardware directly without any virtualization. The results of the evaluation show that our system is much better than the existing virtual machine systems in point of the performance. In addition, a guest OS in our system requires a few amount of modification costs. Consequently, the experiments prove that our system is a practical approach for both performance and engineering cost sensitive systems.

Microkernels and virtual machine monitors are both utilized as platforms for running operating systems. Although there are many similarities in their designs and features, they have opposite advantages and drawbacks. A microkernel based system is highly portable. However, the interface it exposes is inflexible and incompatible with other real hardware interfaces. In contrast, a virtual machine monitor interface is identical to a specific real hardware interface. However, the implementation of virtual machine monitors highly depends on processor architectures and specific hardwares.
In this paper, we present a new model of virtual machine monitor, a flexible dynamic translator constructed on a portable microkernel. Our model offers both high portability and compatibility. Moreover, its flexible interface could be reconfigured to support various types of hardware interfaces. The results of the evaluation show that the performance of our prototype system is unsatisfactory, so we propose some techniques to improve its performance as future work.

Developing and deploying augmented reality (AR) services in pervasive computing environments is quite difficult because almost of all current systems require heavy and bulky head-mounted displays (HMDs) and are based on inflexible centralized architectures for detecting service locations and superimposing AR images. We propose a light-weight mobile AR service framework that combines personal mobile devices most of people own nowadays, visual tags as inexpensive AR techniques, and mobile code that enables easy-to-deploy environments. Our framework enables developers to easily deploy mobile AR services in pervasive computing environments and users to interact them in a both of practical and intuitive way.

In this paper, we present a new model of virtual execution environment, a flexible machine emulator with dynamic translator constructed on a portable microkernel. Our model offers both high portability and compatibility. Moreover, its flexible interface could be reconfigured to support various types of hardware interfaces.

We propose a novel computer gaming style in which users can enjoy games through their daily lives without paying too much attention and time for the games, we call it lifestyle ubiquitous gaming. Lifestyle ubiquitous games track human daily activities implicitly and incorporate the tracked activities in their game logics. Then they represent decorated and virtualized those activities on ambient displays. We believe that this gaming style enables not only making their boring and messy daily tasks fun but improving their lazy lifestyle and customs. In this paper, we describe our lifestyle ubiquitous gaming concept and framework, and then we show our two case studies implementing distinguishing scenarios.

Microkernels and virtual machine monitors are both utilized as platforms for running operating systems. Although there are many similarities in their designs and features, they have opposite advantages and drawbacks. A microkernel based system is highly portable. However, the interface it exposes is inflexible and incompatible with other real hardware interfaces. In contrast, a virtual machine monitor interface is identical to a specific real hardware interface. However, the implementation of virtual machine monitors highly depends on processor architectures and specific hardwares.
In this paper, we present a new model of virtual machine monitor, a flexible dynamic translator constructed on a portable microkernel. Our model offers both high portability and compatibility. Moreover, its flexible interface could be reconfigured to support various types of hardware interfaces. The results of the evaluation show that the performance of our prototype system is unsatisfactory, so we propose some techniques to improve its performance as future work.

In this paper, we propose a novel computer gaming style Lifestyle Ubiquitous Gaming, where our experience is enriched in our daily lives by using gaming concepts. In lifestyle ubiquitous gaming, human daily activities are implicitly tracked and interpreted. Then, the feedback of the activities are returned to a user as a game system' presentation. The user can enjoy the game without too much conscious. This characteristic has a very big difference from traditional gaming.
We describe the motivation and conceptual framework of lifestyle ubiquitous gaming, and show two case studies that we have conducted to demonstrate our proposal's feasibility and effectiveness.

In this paper we present a wearable teddy, RoonRoon (a physical embodiment artefact) that acts as a user interface for information services. It can monitor user's physical activity state (walking, standing, running and sitting) and can provide personalized information in a contextual manner minimizing interruptions. RoonRoon has four notification mechanisms: audible beeps, human voice, ambient sound and vibration. Since RoonRoon is intended primarily for mobile use, interaction is managed completely by voice using natural language. This paper describes the design and implementation of RoonRoon. We also present an informal study that reveals end users' impressions towards wearable toys as social interfaces.

We propose an effective design process for tangible media using a novel framework "CookieFlavors". CookieFlavors provides developers with some abstractions which hide cumbersome implementation phase to develop tangible media. Moreover the framework supports developers to do prototyping them easily. From three case studies to develop tangible media using our framework, we have confirmed our design process and derived some future works of our framework.

Emergence of social web services like YouTube[1], Flickr[2] etc. is constantly transforming the way we share our lifestyles with family, friends and colleagues. ne significance of specific contents can be enhanced if shared with the right person, at right location and time. This paper explores this contextual social content sharing mechanism and presents DroPicks that utilizes augmented everyday artefacts as containers for digital information. Many everyday artefacts are immobile, which implicitly restricts contents stored to a location. Such indirect, contextual sharing has advantages over direct communication mechanisms (E-mail, SMS, IM...) as it provides the content in correct context, seeking minimal attention. Considering the lightness of social contents, we argue that our approach is appropriate, as it does not seek immediate attention, but rather offers the content in the background.

This paper proposes a technique to achieve asymmetric multiple OSes environment for symmetric multiprocessors. The system has a host OS and guest OSes: L4 microkernel and their servers run as the host OS, and modified Linux runs as the guest OS. OS manager which is one of the servers on the host OS manages the guest OSes. Our approach avoids a lot of execution overheads and modification costs of the guest OSes because the each OS can control hardware directly without any virtualization. The results of the evaluation show that our system is much better than the existing virtual machine systems in point of the performance. In addition, a guest OS in our system requires a few amount of modification costs. Consequently, the experiments prove that our system is a practical approach for both performance and engineering cost sensitive systems.

Developing and deploying augmented reality (AR) services in pervasive computing environments is quite difficult because almost of all current systems require heavy and bulky head mounted displays (HMD) and are based on unflexible centralized architectures for detecting service locations and superimposing AR images. We propose a light-weight mobile AR service framework that combines personal mobile devices, visual tags and mobile codes, which enable developers to easily deploy mobile AR services in pervasive computing environments and users to interact them in a both of practical and intuitive way.

In ubiquitous computing environments, context-awareness is one of the most important research topics. Computers embedded in our surrounding can extract information about a user and the information makes it possible to offer personalized services according to the user's preference. To extract a large amount of context information, wearable sensor devices will become more important in the near future. However, it is not easy to develop context-aware services that use context information from wearable sensor devices because of the gap between low-level sensor information and high-level context that the services require.
In this paper, we propose a software infrastructure for wearable sensor networks. We first discuss the requirements to retrieve context information from wearable sensor networks. Then, we introduce our software infrastructure named Cinnamon that extracts high-level context from low-level context information retrieved from wearable sensor networks. The software infrastructure makes it dramatically easy to develop context-aware services for wearable sensor networks. We present the design and implementation of Cinnamon, and discuss our current prototype implementation.

In ubiquitous computing environments, our daily lives will be made convenient by embedded intelligent devices. Those devices, such as car navigation systems, personal digital assistances, and cellular phones, provide various kinds of the complex services. Those devices are networked with each other and provide complicated services, through the Internet. While they provide useful services, there is an increasing possibility of security attacks, which include the unexpected execution of un-secure codes. Current information appliances have not yet fully embodied a resource protection mechanism that prevents misbehaved applications from consuming the whole CPU capacity of system resources.
In this paper we propose Accounting System, and describe its design and implementation. The system is a resource monitoring and restriction system that has the purpose of improving the system's reliability and security. We developed the system on Linux. Our system is a very generic to offer various services, such as security improvement, overload control, and class-based accounting, that require CPU resource control.

In this paper, we describe an augmentation of everyday artefact called sentient artefact. A sentient artefact is expected to capture the user's specific context implicitly and naturally from its original usage since such an everyday artefact has inherent roles and functionalities. Therefore, a context-aware space is built incrementally using the specific contextual information. We show three types of everyday artefact augmentation, and propose a sensor selection framework that allows an artefact developer to systematically identify desirable sensors. Also, we discuss expectations and issues on the augmentation through prototyping. © IFIP International Federation for Information Processing 2005.

In ubiquitous computing era, the notion of context-awareness will play an important role. An application should be aware of its operating context for supporting and enriching human activities. Such contextual information is required to be captured as seamlessly as possible through interaction between users and surrounding environments. This leads to the need for dealing with a wide variety of contextual information from a physical world. In this paper, we propose a conceptual framework, Bazaar, for modeling the physical world and for manipulating the model. It constructs the model with self-descriptive objects represented as a set of triples. Also, a programming model allows a developer to intuitively manipulate the model and develop an application. Moreover, we report experiences with building sample applications. Copyright 2005 ACM.

In ubiquitous computing era, the notion of context-awaxeness will play an important role. An application should be aware of its operating context for supporting and enriching human activities. Such contextual information is required to be extracted as seamlessly as possible through interaction between users and surrounding environments. This leads to the need for dealing with a wide variety of contextual information from a physical world.
In this paper, we propose a conceptual framework, Bazaar, for modeling the physical world and for manipulating the model. It constructs the model with self-descriptive objects represented as a set of triples. Additionally, it provides a programming model for a developer so that he/she can intuitively manipulate the model and develop an application. We also report experiences with building a sample application.

In this paper, we propose universal interaction for networked home appliances, which provides a simple mechanism to fill the gap between traditional user interface systems and advanced user interaction devices. Our middleware enables us to control appliances in a uniform way at any places, and the system allows us to select suitable input and output interaction devices according to our preferences and situations. Our middleware has based on the stateless thin-client system, and translates input and output interaction events according to user interaction devices. Therefore, our system allows us to use a variety of interaction devices without modifying applications using existing GUI toolkits.

In this paper, we propose a personalized display, "AwareMirror: an augmented mirror". AwareMirror presents information relevant to a person in front of it by super-imposing his/her image. A toothbrush has been chosen as an identification tool while proximity sensors have been utilized to detect a person's position (in front of the mirror). Also, three types of information that can affect a user's decision have been selected. The mirror has been constructed using an acrylic magic mirror board and ordinal computer monitor. The acrylic board has been attached in front of the monitor, and only bright color from the display can penetrate the board. As a result of preliminary evaluation, we found that the mirror is useful to offer information in an unobtrusive manner while preserving its metaphor.

In this paper, we propose a personal home server that makes it possible to coordinate various home appliances. Since each person has a different personal home server, it allows us to personalize how to use the appliances according to his preference. Also, our personal home server can discover and configure various appliances at any places, such as railroad stations, cars, and streets as well as houses in a seamless way. We show the design and implementation of a personal home server, and discuss the current design and future work.

The paper proposes a framework to support spontaneous interaction among information appliances in our daily computing environments by creating a society of appliances. Our framework, that we call Pervasive Servers, embeds micro-servers called stationary pervasive servers everywhere. Also, a personal pervasive server that is carried by each person coordinates the embedded servers that are near to the person. The framework is very attractive because it is easy to personalise the coordination according to each person's policy. We also show four applications using our framework.

Ubiquitous computing will dramatically change our lives due to the enhancement of our real world. Augmented reality (AR) is a promising technique for realizing the enhancement by superimposing computer generated images on video images. However, it is not easy to build applications using augmented reality techniques since the developer needs to deal with issues like distribution and context-awareness. It is desirable to provide a software infrastructure to hide the complexities from programmers.
In this paper, we propose a middleware called TEAR (Toolkit for Easy Augmented Reality) supporting augmented reality for ubiquitous computing environments. Our middleware provides several multimedia components that process video streams using AR techniques. New components to provide more complex functionality can be developed by composing these components.

Ubiquitous computing proliferates complexity and heterogeneity of software. Component software provides better productivity and configurability by assembling software from several components. The purpose of this paper is to investigate system configurations on a component-based system and the side effects of the configurations. We have implemented a component-based Java virtual machine named Earl Gray, by modifying an existing Java virtual machine. The case study revealed several problems to use the current component framework when configuring software. We report three experiments of those problems and present a future direction to solve the problem.

In this paper, we propose a middleware system for building adaptive migratory continuous media applications. In future computing environments, a variety of objects at home and in offices will embed computers, and various applications will be moved among these computers according to the user's locations. For example, a computer that executes an application may be changed to another computer according to the location of the ap plication's user for monitoring his behavior. However, since the computers may have dramatically different hardware and software configurations, the application cannot be moved without taking into account the configurations of the computers that the application is migrated. Therefore, migratory applications should be aware of environments where they are executed.
The paper focuses on middleware for building adaptive migratory continuous media applications that are one of the most important classes of migratory applications. Our middleware consists of three components. The first component is an environment server that monitors computing environments that applications run, and it notifies the changes of the environments to the applications. The second component is a continuous media toolkit that enables us to build adaptive migratory continuous media applications easily. The toolkit enables us to build continuous media applications in a highly configurable way. This makes applications to be adapted to every computing environment by changing their configurations. The third component is a migration manager that makes applications migratory. We also show an example of a migratory application, and describe some experiences with building the application.

In this paper, we describe our currently ongoing work to build distributed middleware for networked audio and visual home appliances, which is executed on commodity software components, Linux and Java.

This paper proposes an extension To the standard CORBA interfaces to support explicit binding. The extended interface enables a program to select a suitable transport protocol for. each method invocation. and it provides a flexible and versatile framework to control various transport protocols. The interface is compatible with the object model of CORBA, and it is carefully designed to take into account the portability of an application to other CORBA implementations. The extended interface has implemented in omniORB2, which is a CORBA 2.0 compliant system that has been developing at. AT&T Laboratories, Cambridge.