A Sensor Data Gathering Framework for Agricultural-Fields: Implementation and Experiment Report Hideya Ochiai The University of Tokyo / NICT, JAPAN 1
Background • Weather stations at remote sites • How can we collect sensor data ? Mechanical observation Weather Station 2 引用 : http://www.twin.ne.jp/~saineria/flowerknowledge/flowerknowledge01.html 引用 : http://www.edu.city.yokohama.jp/sch/es/hie/6-1ho-mu/page1.htm
<< Background >> By making use of PHS or 3G Network Real-time data collection from remote sites PHS PHS PHS Access Point PHS PHS PHS ○ Easy to deploy × Communication fee for carrier Agricultural Field □ Real-time is not always important 3
<< Proposal >> DTN-Based Data Gathering Use the movement of truck as a network carrier WiFi WiFi WiFi WiFi WiFi WiFi WiFi WiFi Agricultural Field ○ Easy to deploy ○ No communication fee for carrier 4 △ Delivery delay is physically bound to the movement of the truck
What is DTN ??? • Delay or Disruption Tolerant Networking Message B A Circle: Transmission Range Features Isolated Networks 2 1 7 4 3 8 Intermittent 6 Connectivity 5 5 Mobile Node
Routing in DTN: Potential-Based Routing (PBR) Potential Data Flow Data Server Truck Sensor Data Flow To deliver sensor readings to the data server 6 : Wireless device
Experiment Scenario 1. The sensors and the server are isolated. 2. The sensors continuously send data to the server. 3. The movement of "mobile nodes" carries data to the server. Data Server (#99) Mobile Node (#4) Mobile Node (#3) Mobile Node (#5) Sensor Sensor Deploy (#1) 7 Deploy (#2)
Armadillo220 Storage(2GByte) Wifi 802.11g Power Circuit Battery(6.0V 2100mAh) 8
Implementation PEAR for sensor data gathering PEAR: potential-based entropy adaptive routing 1. Autonomously develops routing information 2. Adaptively changes message delivery form according to the physical mobility pattern. • Software: http://sourceforge.net/projects/pear/files/ • IP packet buffer: 4096 entries • Source code: 3225 lines in C . • Foot print: 34kbyte (object code) • OS: linux-2.6.12.3-a9-15 • Hardware: Armadillo-220 (Atmark Techno) "Mobility entropy and message routing in community-structured 9 delay tolerant networks", ACM AINTEC 2008.
10
11
12
13
14
Weather Station with DTN Weather Sensor (WXT510/520) - Temperature - Humidity Observe and send data to - Pressure the data server (#99) - Rain Fall - Wind Direction (every ten second) - Wind Speed DTN node (#1, #2) 15 RS232C(Serial)
16 Mobile Node -- Carry by hand!! DTN node
17 Data Server 40 [min] 50m
Contact Graph The number of received neighbor advertisement packets. 18 Mobility Entropy � 2.1 e.g., node 8 received 45 advertisement from sensor 1.
Potential for the Data Server Potentials developed by PEAR potential construction algorithm 19
Message Flow (1/2) Red arrow: messages from sensor 1 (#1) 20 Blue arrow: messages from sensor 2 (#2)
21 Message Flow (2/2)
Distribution of Delivery Latency 22 E.g., 50 = [0,100]
Collected Temperature Data from Sensor 1 Delivery Rate = 100 % 23
Collected Temperature Data from Sensor 2 Delivery Rate = 100 % 24
Conclusion • DTN-Based Sensor Data Gathering – Data server � Truck � Sensors • Implementation – DTN-Communication Boxes – Potential-Based Entropy Adaptive Routing (PEAR) • Experiment – in the University of Tokyo – 100% delivery (with 20min – 30min delay) 25
Thank you ... DTN-Based Sensor Data Gathering 26 DTN Communication Boxes Sensor data delivery pattern
Recommend
More recommend
