system architecture for data communication and ...system architecture for data communication and...

System Architecture for Data Communication and Localization

under Harsh Environmental Conditions in Maritime Automation

IEEE 10th International Conference on Industrial InformaticsBeijing, China

Thorsten Wehs, Manuel Janssen, Carsten Koch, Gerd von Cölln

Hochschule Emden/Leer, GermanyUniversity of Applied Sciences, Department of Informatics and Electronics

Email: {koch, coelln}@technik-emden.de

27. July 2012

IntroductionSystem overviewRadio technology

Experiment and performanceSummary and outlook

OutlineMotivation

Outline

Outline

I motivation

I system overview

I radio technology

I experiment and performance

I summary and outlook

T. Wehs System Architecture for a WSN 27. July 2012 2/14



OutlineMotivation

Motivation

O�shore operations

I part of research project SOOP - Safe O�shore OPerations

I contribution to the industrialization of o�shore wind energy

I today: manual process monitoring (TETRA radio, visual, . . . )

I system architecture as base to generate an overview of the operation

I wireless sensor network (WSN) which combines communication andlocalization for harsh environmental conditions




OutlineMotivation

Motivation

Wireless sensor network

I most available approaches focus on localization or communication

I Active-RFID networks ⇒ good localizationI ZigBee networks ⇒ good communication

I usage of a rugged radio link for industrial applicationsI use case: rough maritime environment

I approach can be easily mapped to many problems in the industries




Overall architectureWSN architectureSensor node

Overall architecture

Mission assistance

I distribution of wireless sensor nodes tomobile and stationary parts on vessels

I a gateway concentrates whole sensor data

I assistance system determines possiblehazards and warns the crew [3]

Clustering

I interconnection ofindependent WSNs on

I very huge vesselsI supply and jack up

vesselsI . . .

WSN 1

WSN n

WSN 2

Assistance system

SOOP-Gateway

Sensormote

Cable (Ethernet) / wireless (GSM)

Legend





WSN architecture

WSN

node #S0

gatewaynode #M0

node #Mn

node #Sn

node #S1

node #S2

node #M1

eth1

eth0uwb0

Legendinternal Ethernet:

external Ethernet:

UWB radio:

UART:

node #Sx: stationary node

node #Mx: mobile node

NMEA-

device





Sensor node

Features

I wireless transceiver module

I decentralized localization technique

I NMEA device wrapper

I interface for sensors

I communication via SCAI protocol [2]

UWBTransceiver

Base board

ARM Cortex M3

Sensors

AccelerationTemperature

NMEA...

Powersupply

3,7 V

Figure: block diagram of the sensor node Figure: prototypical sensor node (stacked,power supply not shown)




State of the artUltra Wideband

State of the art

Some popular technologies

I WLAN (IEEE 802.11)

I IEEE 802.15.4 (Zigbee, WirelessHart, ISA 100.11a) [1], [7]

I Chirp Spread Spectrum (CSS) [4]

WLAN IEEE 802.15.4 CSS UWB

Frequency band [GHz] 2.4/5 (ISM) 2.4 (ISM) 2.4 (ISM) 3.1 - 5.3Range [m] 30-100 220-250 > 800 88Low Energy - + o +Data Rate 300 mbps 250 kbps 250 kbps 159 kbpsRobustness - o o +Distance determination few meters 1-2 meter sub meter < 7 cm

Due to the strong requirements

I robustness in harsh environmental conditions and

I precise distance estimation for precise localization,

there is only one solution - Ultra Wideband (UWB)





State of the art










there is only one solution

- Ultra Wideband (UWB)





State of the art










there is only one solution - Ultra Wideband (UWB)





Ultra Wideband

Features

I no carrier waves ⇒ pulse based

I very short duration pulses (≈ 500 ps)

I frequency bandwidth: 2.2GHz

I only about 50µW transmit power [5]

I causes no interferences to existing radio systems

I high accuracy distance determination [5]I ≈ 3 cm in line-of-sight (LOS)I ≈ 7 cm in moderate non-line-of-sight (NLOS)




Experiment and performance

Experiment and performance

I measurements in comparable environments

LOS NLOS (heavy NLOS (25 cm rein-

metal obstructions) forced concrete wall)

Ground truth [cm] 100 600 100 600 100 600

Mean [cm] 101 599 106 594 137 629Median [cm] 103 600 110 602 140 629Std. deviation [cm] 4.1 3.2 7.8 8.4 4.5 1.9Min [cm] 87 588 77 552 127 620Max [cm] 104 604 113 606 143 632Failures [%] 2 0 0 0 1 1

Table: 100 single measurements for each scenario and distance were performed




Summary and outlookLiterature

Summary and outlook

The aim of the paper was to depict a new approach for wireless sensor networkswhich combines localization and communication for industrial applications

Summary

I basic architecture for o�shorelocalization and communicationsystem conceived

I prototypical hardware system foroutdoor and indoor localization

I sensor node: up to 80mdistances

I gateway: SCAI protocolimplemented

I veri�cation of UWB as robustradio technology

I promising measurement results(≈ 8 cm even in heavy NLOS)

Outlook

I LOS/NLOS detection

I NLOS compensation

I scaling the number of nodes

I algorithm optimized referencenode positioning





Literature

[1] M.R. Akhondi, A. Talevski, S. Carlsen, and S. Petersen.Applications of wireless sensor networks in the oil, gas and resources industries.In Advanced Information Networking and Applications (AINA), 2010 24th IEEE InternationalConference on, pages 941 �948, april 2010.

[2] C. Busemann and S. Behrensen.SCAMPI project report, requirements and �nal speci�cation of the SCAI-protocol.Technical report, OFFIS - Institute for Information Technology, 2011.

[3] R. Droste, E. Böde, A. Hahn, T. Peikenkamp, and J. Lenk.Security analysis for o�shore operations, 2011.

[4] Dr. Frank Schlichting.Präzise Abstandsbestimmung und Lokalisierung mittels Laufzeitmessungen (RTOF) durch Einsatzder 2,4 GHz Chirp Spreiztechnologie (CSS).In Wireless Technologies Kongress, 2007.

[5] Time Domain.Data Sheet PulsON P400 RCM, 2011.

[6] W. S. Murphy, Jr. and W. Hereman.Determination of position in three dimensions using trilateration and approximate distances.Technical report, Department of Mathematical and Computer Sciences, Colorado School of Mines,Golden, Colorado, October 1995.

[7] R. S. Wagner and Richard J Barton.Standards-based wireless sensor networking protocols for space�ight applications, 2010.





Localization

Method: lateration (no angles!)

I known positions: {P1, ..., P4}

I unkown position: P0

Three phases algorithm

1. determine distances {r01, ..., r04}

2. calculate distances {d12, ..., d14}

3. solve system of linear equations [6]

P3 = {x3, y3, z3}

P2 = {x2, y2, z2}

P0 = {x0, y0, z0}

P4 = {x4, y4, z4}

r03

r04

r01r02

d13

d12

d14

P1 = {x1, y1, z1}

x2 − x1 y2 − y1 z2 − z1x3 − x1 y3 − y1 z3 − z1x4 − x1 y4 − y1 z4 − z1

.

.

.

.

.

.

.

.

.xn − x1 yn − y1 zn − z1

·[x0 − x1y0 − y1z0 − z1

]=

1

2·

r201 − r202 + d212r201 − r203 + d213r201 − r204 + d214

.

.

.

r201 − r20n + d21n





Gateway

Features

I suited for installation in anavigation bridge

I retail x86 architectureI running linux OS

I coordination of the networkI prioritizationI fault detectionI sensor node con�guration

I sensor data preprocessingI plausibility checkI aggregation to higher level

informations

I provide data to assistance systemFigure: IPC connected with a sensor node

(sensor node unstacked)


system architecture for data communication and ...system architecture for data communication and...

Documents