Tomu DB
Overview
Sensor network is a promising tool for collecting information about the real world without human's intervention. Instead of asking for raw data, query processing is a convenient method to obtain desired information from a sensor network. In general, people are likely to perform a real-world search; they search things related to their daily lives in the real world ranging from straight-forward queries to complex ones. However, a single sensor network cannot respond to a variety of queries by using its own sensing data. Currently, each organization deploys sensor networks for their own uses. Each sensor network has many different characteristics such as sensor types, network size, sensing resolution, etc. In order to realize the real-world searches, a collaboration of such heterogeneous sensor networks is a promising solution.
This technology proposes a TomuDB architecture for performing multi-resolution queries in heterogenous sensor networks. In order to cope with the real-world searches, heterogeneous sensor networks work collaboratively by forwarding queries to other networks that are likely to answer the queries. Each query
also includes the requirements of resolutions determined by the users. If sensing data collected from all collaborated networks
do not meet the required resolutions, an interpolation is used to calculate approximate results from coarse answers. A prototype
implementation is presented and the experiments are performed to study the performance of TomuDB architecture in the realworld
environment.

News
- 2008 8/5 ver1.01 Release
- 2008 3/31 ver1.0 Release
Implementation Details
- Programming Language:C
- Proper Operation Environment
- Fedora7
- gcc 4.1.2
- glibc 2.6
- Hardware Platform:(CPU: Pentium3 1G, Memory:256MB)
- DBMS:PostgreSQL
Publication
-
Yoh Shiraishi, Niwat Thepvilojanapong, Yosuke Tamura, Tatsuro Endo, Koichi Yamada, Nayuta Ishii, Hiroki Ishizuka, Keisuke Kanai, Yoshito Tobe
"TomuDB: Multi-Resolution Queries in Heterogeneous Sensor Networks through Overlay Network",
Poster session, The Fifth ACM Conference on Embedded Networked Sensor Systems (SenSys 2007),
Sydney, Australia, Nov. 2007.
Download
- 2008 8/4 ver1.01 Release
- A readme document is included in the archive file above.