Developing and Deploying Sensor Network Applications with AnduIN - - PowerPoint PPT Presentation

developing and deploying sensor network applications with
SMART_READER_LITE
LIVE PREVIEW

Developing and Deploying Sensor Network Applications with AnduIN - - PowerPoint PPT Presentation

Developing and Deploying Sensor Network Applications with AnduIN Daniel Klan Katja Hose Kai-Uwe Sattler Databases & Information Systems Group Ilmenau University of Technology, Germany Outline Introduction AnduIN Query


slide-1
SLIDE 1

Developing and Deploying Sensor Network Applications with AnduIN

Daniel Klan Katja Hose Kai-Uwe Sattler

Databases & Information Systems Group Ilmenau University of Technology, Germany

slide-2
SLIDE 2

2 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Outline

 Introduction  AnduIN  Query Specification  Query Optimization  Case Study  Conclusion

slide-3
SLIDE 3

3 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Introduction

 Anomalies in sensor networks

 Sensors measuring the temperature

s9 s6 s5 s25 s2 s15 s17 s24 s18 s10 s4 s12 s28 s13 s16 s7 s3 s1 s21 s11 s8 s19 s23 s41 s34 s38 s37 s20 s22 s27 s26 s31 s29 s35 s36 s39 s14 s40 s32 s33 s30 y

slide-4
SLIDE 4

4 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Introduction

 Anomalies in sensor networks

 Detecting anomalies in sensor data streams

y s9 s6 s5 s25 s2 s15 s17 s24 s18 s10 s4 s12 s28 s13 s16 s7 s3 s1 s21 s11 s8 s19 s23 s41 s34 s38 s37 s20 s22 s27 s26 s31 s29 s35 s36 s39 s14 s40 s32 s33 s30

slide-5
SLIDE 5

5 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Introduction

 Anomalies in sensor networks

 Determine anomaly regions

x y s9 s6 s5 s25 s2 s15 s17 s24 s18 s10 s4 s12 s28 s13 s16 s7 s3 s1 s21 s11 s8 s19 s23 s41 s34 s38 s37 s20 s22 s27 s26 s31 s29 s30 s35 s36 s39 s14 s40 s32 s33

slide-6
SLIDE 6

6 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Introduction

 Wireless sensor networks

 Continuous monitoring (environment, traffic, ...)  Limited capacities (computing power, battery lifetime, ...)

 Query processing

 Centralized processing  In-network query processing  Hybrid processing

Centralized processing INQP

slide-7
SLIDE 7

7 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Introduction

 Problem:

 Which parts of a query to evaluate within the network?

 AnduIN:

 Combining in-network query processing with data stream

processing

 Query decomposition and optimization,

  • bjective: minimization of energy consumption

 Comfortable way of specifying queries

slide-8
SLIDE 8

8 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

AnduIN

slide-9
SLIDE 9

9 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Logical Sensor Networks

 Registering a sensor  Registering a logical sensor network

CREATE STREAM net_stream (id int, temperatur double) NETWORK [ 15 (6, 9)] SAMPLING 30 SECONDS

15 6 9

ADD SENSOR 15 (temperature double, humidity double) LOCALIZATION [47° 25’, 010° 59’]

slide-10
SLIDE 10

10 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Web-based Box-and-Arrow Frontend

 Example query:

 Detecting anomaly regions based on bursts

CREATE STREAM s_burst AS SELECT timestamp, temperature FROM net_stream [ burst-detection(w => 1000, threshold =>‘forecast‘) ]; SELECT timestamp, temperature FROM s_burst [anomaly-region (t => 0.5)]

slide-11
SLIDE 11

11 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Web-based Box-and-Arrow Frontend

 Example query:

 Detecting anomaly regions based on bursts

slide-12
SLIDE 12

12 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Query Decomposition and Optimization

logical plan

anomaly region detection burst detection sampling projection

source a b p s sink

possible physical query plans

a b p s a b p s a b p s a b p s

centralized in-network

transformation

slide-13
SLIDE 13

13 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Cost Model

 Costs for in-network query processing  Factors influencing costs:

slide-14
SLIDE 14

14 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Cost Model

 Costs for in-network query processing  No leader nodes

slide-15
SLIDE 15

15 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Cost Model

 Costs for in-network query processing  Considering leader nodes

slide-16
SLIDE 16

16 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Case Study

 Detect anomaly regions based on bursts  100 nodes

slide-17
SLIDE 17

17 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Summary

 AnduIN

 Declaritive query formulation  Comprehensive optimization of complex queries  Cost model

 Future Work

 Improving the implementation  Multi-query optimization  Online query optimization

slide-18
SLIDE 18

18 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Thank you...

slide-19
SLIDE 19

19 DMSN - September 09 Developing and Deploying Sensor Network Applications with AnduIN

Case Study

 Tmote Sky sensor nodes with

 16 bit MCU MSP430F1611,  4 MHz clock rate,  IEEE 802.15.4 compatible CC2420 transceiver with 250kBit/s