Yutaka Arakawa

Experiments on Binary Sensor Networks for Estimation of Target Perimeter and Size

This paper presents experiments on a method of estimating the size and perimeter length of a target object by using networked binary sensors whose locations are not known. Although these sensors only report whether the target object is detected or not, our proposed estimation method yields the size and the perimeter length of the target object by using sensors that have different sensing area sizes. Based on theoretical results in our previous work, we implement and experiment with our proposed estimation method by using infrared distance measurement sensors with communication capabilities. As a result, we can confirm that our proposed estimation method works if we appropriately choose parameter values of experiments.

uTupleSpace: A Bi-Directional Shared Data Space for Wide-Area Sensor Network

A sensor network covering a large area enables connection to various types of sensors and actuators, but application development becomes complicated because of the uncontrollable behavior of such a large number of devices. We propose the u Tuple Space model for uniform and indirect communication with two extensions to the original tuple space model. The extensions enable efficient range search for multi-dimensional keys and bi-directional communications. Our implementation of our proposed model also integrates load balancing by using a distributed hash table, and the experimental results indicate good scalability in multiple servers. The trial application of gathering and plotting GPS sensor data works well in the field.

Proposal of Web Framework for Ubiquitous Sensor Network and Its Trial Application Using NO2 Sensor Mounted on Bicycle

This paper proposes a web framework for a ubiquitous sensor network (u-framework), which supports application development on the uTupleSpace, a special database management system for sensor data. Moreover, an evaluation of this framework applied to a practical system development is also explained in this paper. We first analyzed some examples of vertical-domain services using a wireless sensor network, and found that server-side functions in each of the services share some degree of similarity. The proposed u-framework is built on Ruby on Rails and can save a great deal of effort in the development of such functions. We applied the u-framework to server development for our field trial. In the trial, we ran bicycles equipped with small and high-precision NO2 sensors to gather and share information on air pollution. We detail here the results of the field trial and the effects of the u-framework on server-side development.

Design and Implementation of New uTupleSpace Enabling Storage and Retrieval of Large Amount of Schema-less Sensor Data

This paper proposes the design and implementation of new uTupleSpace to meet increases in variety and in quantity of sensor data. We introduce the following two extensions to our storage and delivery system for sensor data in order to share much of sensor data for a lot of applications to utilize them. One is the feature to store schema-less sensor data and search into them. This enables to introduce various types of applications for a sensor network one after another and to share the data stored through such applications. The other is to create chunks of sensor data. This method fundamentally improves processing overheads caused by existence of a huge amount of small sensor data. We implemented the uTupleSpace with the proposed enhancement and experimented it in performance improvement by creating chunks of data.

UBI-Tree: Indexing Method for Schema-Less Search

We propose an indexing method called UBI-Tree for improving the efficiency of a new type of data search called schema-less search. Schema-less search is a multi-dimensional range search from a wide variety of data, such as sensor data, collected through participatory sensing. Such data have different types and number of dimensions because a participant uses various devices. Therefore, applications must search for their target data within the sensor data in a cross-schema manner. UBI-Tree is a tree-structured index based on R-Tree. The insert algorithm classifies various data into nodes according to newly introduced scores to estimate the inefficiency of classification. The score can uniformly represent the difference in the types of dimensions between data as well as the difference in dimension values. By classifying data that have a similar dimension set into the same node, UBI-Tree suppresses the curse of dimensionality and makes schema-less searches efficient. The validity of UBI-Tree was evaluated through experiments.

Load-Adaptive Indexing Method for Schema-less Searches

There is a growing need for a horizontal-integrated platform for machine to machine (M2M) applications. To facilitate high reusability of the various data stored on the platform, the platforms search function must be able to locate all data useful to a particular application. We refer to this cross-schema-manner search as a schema less search. We have proposed a tree-structured indexing method, called UBI-Tree, for efficient schema-less searches. UBI-Tree predicts usage frequencies of each query based on characteristics of stored data. Moreover, UBI-Tree classifies incoming data to improve the search performance of frequently used queries. One disadvantage of the conventional UBI-Tree is decreased search performance as a result of differences between predicted and actual frequencies. Therefore, we propose a new method to improve search performance. This method utilizes the query history and is based on the real search load. We show experimentally that our proposed method improves the average latency and throughput of search processes.