ZigBee for Wireless Sensor Networks in Space and Field Science Mark Foster, CSC / NASA Ames Rick Alena, NASA Ames Intelligent Systems Division Discovery and Systems Health NASA Ames Research Center

CENIC: Expanding our Horizons UCI, March 8, 2011

Agenda • Why wireless sensors – Selected application domains

• • • •

Project Objectives Solution: TI ZigBee devkit plus SBIR leverage Recent prototyping and testing Follow on efforts

Shroud design validation

Stiffened panels with overdesigned thickness versus optimized design (Collier Research)

• Optimized: less weight (thinner construction) • Instrumentation to validate optimized design during testing and flight • Wired instrumentation scope limits (weight and location) • Wireless: more sample points, alternate datapath to provide distinct data fault assurance

Potential shroud sizes compared to Shuttle

Human and Robotic Exploration Testing

• Spacesuit design, monitoring • Human-robot enhancement • Robotic field exploration

Earth Science Technology – Sensor Webs

NASA Facilities - Smart Buildings

• Smart energy profile • Building systems management • Smart meters

NASA Ames – Sustainability Base

Wireless Sensor Network Project Objectives •

Develop “Intelligent” Wireless Sensor Network (WSN) architecture, software and applications to demonstrate fundamental concept of operations •

•

Evaluate WSN technology function and performance • • •

•

Consider constraints: power, space, weight, cost, time

Reliability Throughput Maturity (technology readiness level)

Evaluate WSN suitability for spaceflight certification •

Operational environment • Temperature, radiation, pressure, vibration, etc.

•

RF interference and compatibility • Effect on spacecraft systems • Spacecraft systems effect on WSN • Multipath distortion immunity

Crew vehicle artist concept Credit: NASA/Lockheed Martin

Intelligent Wireless Sensor Networks (WSN) Definition • • • • • • • •

Conform to IEEE 1451 Smart Transducer Interface Standards Form ad-hoc wireless networks with highreliability Provide fault tolerance through mesh routing Self-manage routing and fault tolerance Provide Transducer meta-information Provide unambiguous sensor data with temporal determinism Provide standard interface to TCP/IP networks Support open software architecture and applications

Intelligent WSN Standards / Open development •

IEEE 802.15.4 provides protocol for ad-hoc Personal Area Network (PAN) formation and management at MAC Layer

•

IEEE 1451 Standard provides architecture for WSN •

1451.0 Network Capable Application Processor (NCAP)

•

1451.4 Transducer Electronic Datasheets (TEDS)

•

1451.5 Wireless Transport Protocols (ZigBee)

•

ZigBee provides framework for network and application support

•

C language for ZigBee and NCAP firmware and bridge software

•

Texas Instruments CC2530 System on Chip (SoC) hardware

•

ARM Co-processor for NCAP

•

Simple Network Monitoring Protocol (SNMP) for external access

ZigBee Testbed Components

• • •

•

Coordinator establishes PAN Routers - forward data Sensor Nodes originate sensor data stream Gateway - connects PAN to IP network (embedded linux)

ZigBee Protocol Stack • •

Keep approach simple: APS layer and below Adapt devkit sample code •

•

•

modify parameters and specific functions

Leverage key functions within supplied object code (Z-stack) Significant learning curve, but can implement complex systems with modest coding effort.

Wireless Sensor Network Testbed Demonstration Structural Monitoring Prototype WSN Applications on IP networks

Sensor Strain Sensor End Device Signal Conditioner Strain Gauge

ARCBee A4 Zigbee Coordinator NCAP

Network Capable Application Processor (NCAP) is gateway to IP networks

SDP-1 Strain (10 - 1000 µe) 2 - 4 channels

3 axis accel (0-3g) Force (0-10 lb)

ACL-1

ARCBee A1 ARCBee A2

Router Module

4 thermistors (0-40C) FLO-1 4 thermistors (0-40C)

Humidity (10-90% RHD) ARCBee A3

Temperature (0 - 100 ˚C) Pressure (0-15 PSI) ENV-1

Transducer Electronic Data Sheet Definition for WSN Bascic TEDS Table Bit Length

Allowable Range

Manufacturer ID

14

17 - 16381

Model Number

15

0-32767

Version Letter

5

A-Z (data type Chr5)

Version Number

6

0-63

Serial Number

24

0-16777215

TEDS generation code snippet void InitStructBasicData(struct BasicData *Basic) { strcpy(Basic->Portal_Number, "192.168.2.12"); Basic->Sensor_Number =1; Basic->AD_Channel =0; Basic->TEDS_ID =25; Basic->Manufacturer_ID = 55; Basic->Model_Number = 0; Basic->Version_Letter = 'A'; Basic->Version_Number = 1; Basic->Serial_Number = 123456; strcpy(Basic->User_ASCII_Data,"data"); } void InitStructThermocoupleData(struct ThermocoupleData * thermocouple) { strcpy(thermocouple->Portal_Number,"192.168.2.12"); thermocouple->Sensor_Number =1; thermocouple->AD_Channel =0; thermocouple->Maximum_Physical_Value_Volts = 3.5; thermocouple->Minimum_Physical_Value_Volts = 0; thermocouple->Maximum_Electrical_Value_Volts = 3.3; thermocouple->Minimum_Electrical_Value_Volts = 0.3; thermocouple->Thermocouple_Type ='B'; strcpy(thermocouple->Cold_Junction_Source,"CJC required"); thermocouple->Sensor_Impedance_Ohms = 100; thermocouple->Transducer_Response_Time_Sec = 1.035; strcpy(thermocouple->Calibration_Date,"2007-09-13"); strcpy(thermocouple->Calibration_Initials,"TED"); thermocouple->Calibration_Period_days = 7; thermocouple->Measurement_Location_ID = 89; }

Wireless Sensor Network Development Task Integrate new sensors for specific structural and environmental monitoring – Multi-channel temperature, atmospheric environmental sensors, load cell and accelerometer – New strain gauge sensors, acoustic emission sensors and other sensors relevant to structural health monitoring – Circuits for sensor to SoC connection compatible with battery power 3.0 VDC. – Modify ZigBee firmware and produce new IEEE 1451 Transducer Data Sheets (TEDS) representing new sensor classes and specific prototype sensors – Test new sensors and determine accuracy of measurement

WSN Testbed Hardware/Software Integration Sensor data streams plus TEDS meta-information and WSN status transferred using ZigBee Protocol

ARCBee Firmware ARCBee Sensor Module

802.15.4 ARCBee Firmware

Gateway between ZigBee and TCP/IP networks using SNMP for defining sensor objects

Sensor Data Display TEDS info Display WSN Status Display

Data Logging and Error Detection Data Error Checking

Mobitrum NCAP Module MOBEE-NET CC2430 Firmware

Serial

Data Logs SNMP Queries

PXA-270 SNMPAgent Mn_Driver TinyOS

Sensor Info Display Application SNMP Queries

Ethernet

Computer Module

ARCBee Sensor Module

SNMP Queries access sensor data streams plus TEDS meta-information and WSN status information

WSN Prototype Demonstration GUI Mockup Strain Sensor Chart STR-1

TEDS FLO-1 RSSI

active STR-2

FLO-1 Battery

TEDS:
ENV‐1:ENV‐T—Resistance
temperature detectors
(RTDs)

Function

Select Property/Cmd Description

ID

—

TEMPLATE

STR-3

Acce ss Bits Data
type
(and
range)

Template
ID —

8

Integer
(value
=
37)

Units —

Measurement —

Minimum ConRes
(–200
to
1,846, %MinPhysVal temperature CAL 11 step
1) ºC

—

Maximum ConRes
(–200
to
1,846, %MaxPhysVal temperature CAL 11 step
1) ºC

STR-4

Reliability and RF Compatibility Test Methods •

failover behavior • • • •

• • •

loss rates under nominal conditions and monitor RF spectrum in ISM band packet loss rate vs external interference throughput vs external interference • • • •

•

PAN association time PAN re-association time • Single hop and double-hop through router 1, 5, 10 node clusters 2 sec, 1 sec, 0.5 sec data rates

802.11 b 802.11 g 802.11 n Bluetooth

multipath environment • • •

reflections from conductive surfaces can prevent data transfer by creating standing wave pattern metallic enclosures of varying size packet and throughput loss rates

WSN Reliability and RF Interference Test Protocols •

ZigBee and 802.15.4 packet analyzer • association time, orphan detection time and reassociation time • Fail sensor node • Fail router node

• • •

WLAN sources to create high duty-cycle interference WiSpy for ISM RF Spectrum packet loss with SmartRF Studio • Directly runs CC2530 chip

•

throughput with Transmit App • Send data as quickly as possible

Sensor Failover: Single and Dual Routers

•

Failover from router to coordinator

•

Failover from router to alternate router

One-Hop PAN Association and Orphan Transition Times • •

Scales well Reasonably fast and consistent

•

Dependent upon data rate Reasonably fast and consistent

•

Two-Hop PAN Association and Reconstruction Times !"#$%#&'()*')++#,-./-#0'.01'23,#0+/45,/-#0'!-63 *$!!!"!!

? > 4=3?< /1 '$!!!"!!

CDE3BA-3F1=473?< /1 A;12DG13BA-3F1=473?< /1

&$!!!"!!

A;13BA-3A> > 4=3?< /13 93F4.H123ID< @ .21 A;13BA-3F1=473?< /13 93 F4.H123ID< @ .21

%$!!!"!!

#$!!!"!!

!"!! !

#

%

&

'

(

)

*

+

,

#!

-./01234536789:1;

• •

Scales well, but more traffic during PAN reconstruction as number of nodes increases Reasonably fast and consistent

RF Interference Test Configuration 802.11g, 802.11n 802.15.4 WLAN WLAN Client Client Adapter Adapter

1 ft

1 ft

Zigbee Collector

1 ft

1 ft WiSPY RF Spectrum Measurement

Zigbee Collector

iperf

iperf Smart RF Studio Computer

• • •

WiFi Access Point (WAP)

Smart RF Studio Computer

WLAN interference source - 802.11bgn Access Point and Client running iperf Two node ZigBee test set using Smart RF Studio WiSpy RF Spectrum capture

2.4 GHz ISM Spectrum Diagram and Baseline

• • •

WiSpy Spectrum Monitor trace for RF baseline 802.11b/g (ARC-WLAN) WLAN on Channel 4 Control experimental variables for each test run

WLAN G and N Mode Interference Spectrum

• • •

802.11g on Chan 1 ZigBee on Chan 11 802.11n on Chan 4

Simple RF Multipath Test Configuration

• •

Run within metallic drawers (12” X 20” X 6”) and (12” X 20” X 12”) 1 and -19 dBm ZigBee power output

Packet Loss Rate with Multipath and WLAN Interference

• • • • •

Case 1: Baseline - no packet loss Case 2 and 3: Multipath yields ZERO Loss rate Case 4: WLAN-G yields significant packet loss Case 5: WLAN-N yields some packet loss RSSI is Received Signal Strength Indication –

Keep near the same level for comparison

*75/43784-&98:,&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

Throughput with WLAN-G and WLAN-N Interference Throughput w/ WLAN-G Interference •

Zigbee Throughput - 104 Kbps to 15 Kbps with interference WLAN Throughput - 18.8 Mbps to 15 Mbps with interference

!"""""

•

!""""

#

%$!"""

#$%&;2-7&#()%

!""

#()%&;2-7&#$% #()%

!" ! #$%&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&#$%&'&#()%&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&#()%

*+,-&./0123456-2/0

Throughput w/ WLAN-N Interference •

*75/43784-&9:8,

!""""" !"""" #

%$!"""

#$%&;2-7&#()%

!""

#()%&;2-7&#$% #()%

!" ! #$%&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&#$%&'&#()%&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&#()%

*+,-&./0123456-2/0

•

Zigbee Throughput - 106 Kbps to 45 Kbps with interference WLAN Throughput - 30.5 Mbps to 21 Mbps with interference

Future WSN Development Activities •

Define single fault-tolerant Developmental and Flight Instrumentation (DFI) architecture as baseline

•

Examine scaling limitations/tradeoffs

•

Extend types of sensors and TEDS supported

•

Examine software/application interface alternatives – Perform tradeoffs of SNMP, DDS, JMS and SQL – Define WNS Data Interface Protocol

•

Higher-performance interface mode for instruments and increase measurement sampling rates

•

Assess acceptable environmental operating conditions

•

Better characterize reliability, fault tolerance and compatibility

WSN Two-Tier Fault-Tolerant Mesh Network Architecture Sensor End Device Coordinator A Gateway 1

TCP/IP Network

Sensor End Device Router Module

Coord Fault

Sensor End Device Router Fault

Router Module

Coordinator B Gateway 2 Sensor End Device

TBD

• • • •

Sensor End Device

Redundant sensors in each module cover sensor failures Redundant Sensor Modules cover Module failures Redundant Routers cover router failures Redundant Coordinators/Gateways cover PAN formation faults and Gateway faults

Sensor End Device Sensor End Device Sensor End Device

X

Sensor Fault Module Fault

WSN Development Team •

NASA Ames Code TI and TN development team – Jeff Becker, Mark Foster, Thom Stone, Ray Gilstrap – John Ossenfort, Pete Wilson, Rick Alena

•

Education Associates Program - Interns – Jarren Baldwin (now at Stanford) – Adrienne Haynes (while at Norfolk State U.)

•

NASA Stennis Space Center – Fernando Figueroa

•

Mobitrum Corp – Ray Wang and Suman Gumandevelli – NASA SBIR Phase I and II

Questions?

go, Glory!

 [email protected] TaurusXL @ Vandenberg

Aerosol Polarimetry Sensor for Earth climate study