Naoya Namatame

A distributed resource management architecture for interconnecting Web-of-Things using uBox

Although there are many smart devices and networked embedded object applications using World Wide Web technologies, it is still a big step to go towards a true Web of Things. It is e.g. difficult to build ubiquitous WoT applications that work in and accross multiple environments. Approaches which aggregate WoT ressources by centralizing all the resource information, have problems: total dependency on external infrasture, lack of private WoT management, inflexible communication patterns and limited dynamic ressource discovery and mapping. To solve these problems, we propose uBox, a local WoT platform which can be a stand-alone server to make your WoT environment, with interfaces to connect the other local WoT platforms. This way, which we call uBoXing, we can create World Wide WoT platform with a distributed architecture. This paper describes the concept of a distributed resource management architecture, and how we implement the concept into software. Also, we will discuss the platform with the example application in SmartTecO environment.

SmartTecO: context-based ambient sensing and monitoring for optimizing energy consumption

This paper presents an energy-saving concept for home/office environments, which proposes to design a multi-layered architecture for an automatic monitoring and control. Based on wireless sensor networks and a context awareness system, the acquired data will be interpreted into different energy-related contextual information. A correlation module using Hidden Markov Models could then give the actuation module a certain context, which allows managing and saving the energy consumption of home/office appliances.

Logical Sensor Network: An Abstraction of Sensor Data Processing over Multidomain Sensor Network

This paper focuses on a sensor network virtualization over multidomain sensor network and proposes an abstraction called “logical sensor network (LSN)” for sensor data processing. In the proposed abstraction, processing is a directed acyclic graph that consists of nodes and streams, which represents a small data processor and communication rules between them, respectively. We have added a notion of a trigger to this graph. A trigger represents a timing of the process execution. We have implemented the middleware named LSN-Middle to run a virtualized sensor network and proved its feasibility.

LSN: An architecture for dynamic sensor mapping for Logical Sensor Networks

We propose Logical Sensor Network(LSN) and LSN Operator to create a dedicated sensor network over a multi-institutional world-wide sensor network. LSN can logically define sensor network environment with XML and LSNO is a middleware that dynamically map and remap the logical sensor to real sensor. With this mechanism, applications can operate and be developed on the premise of detailed knowledge about the sensor network. In this paper, we propose an architecture which operates LSN mechanism with low latency and high scalability, as well as LSN schema.

Pattern-based matrix-size optimization algorithm for compressive sensing in real-world wireless sensor networks

Compressive Sensing (CS) is a novel approach for data representation, which can represent signals at a rate below the Nyquist rate with low computation costs on encoder. For these characteristics, CS is very suitable for low power sensor nodes to save power consumption that is a primary problem in Wireless Sensor Networks (WSN). But there are many problems when using CS in a real environment. One of these is that pattern of sensor values change dynamically. It decreases the efficiency of power consumption and accuracy of recovery. To solve the problem, we propose Pattern-based Matrix-size Optimization Algorithm (PMOA), which aims to improve the accuracy of exact recovery and power consumption.

Uninterruptible Data Supply for sustainable context aware system

Various context mining algorithms are proposed aiming at improving accuracy. The major problem of the existing algorithms is that they assume no data loss in sensor data input, thus are unable to function sufficiently in a practical environment, where sensor data frequently drop. This paper proposes a novel middleware system called Uninterruptible Data Supply (UDS) system, which compensates the missing data with a probabilistic manner. Applications can benefit from UDS on the occurrence of constant and temporal deficit of sequential or discrete data. The evaluation shows that UDS can sustain the accuracy of context over 80% even with 40% data missing.

Bubbles: Overlay sensor network framework for smart spaces

We present a framework for overlay sensor network, which creates virtual sensors and their networks over a physical sensor network. This framework targets a sensor network under a smart space, that is the space with many different types of sensor-related applications running. This work enables a such physical sensor network to cope with various requirements that the applications request.

UDS: sustaining quality of context using uninterruptible data supply system

Context mining algorithms from sensor data have been researched and successful results have been shown. However, since these existing works are focused on improving the accuracy of context mining, they are established on the assumption that they can acquire a complete set of necessary data. Therefore, the context mining algorithms do not work sufficiently since the data drops easily in the reality. In this paper, to cope with this problem, we propose a middleware named UDS (Uninterruptible Data Supply System). The system compensates the missing data, creates virtually complete dataset and provides upper layer applications. Applications operating over UDS can work sufficiently with some data actually missing. We have defined two types of characteristic data deficit patterns and created a robust model for both patterns utilizing Bayesian Network. In the evaluation, we show UDS can sustain the quality of context over 80% with 40% data missing.

Life2Guard: A Physical Disorder Detection in Private Rooms

We present a research result of our project named Life2Guard. The project aims at detecting a person facing a sudden physical disorder in private rooms such as restrooms. In this paper, firstly we specify the required functions for the system to detect physical disorders in private rooms, which are privacy-aware sensing, high availability, quick responsiveness and false-call avoidance. Secondly, we propose how we have implemented the system to fulfill the requirements using following key approaches: IRSensor, Bayesian network, approximate inference, Naive Bayes and user interaction. In this way we have succeeded in detecting physical disorders such as faint or convulsion in a certain situation for 100% within about 30 seconds.