Transcript Industrial Wireless Sensor Networks_ Challenges, Design Principles

INDUSTRIAL WIRELESS SENSOR
NETWORKS
CHALLENGES, DESIGN PRINCIPLES, AND TECHNICAL
APPROACHES
Presented By:
1
Jesmin Jahan Tithi
Std no: 0409052065
S.M.Arifuzzaman
OUTLINE
 WSN
(Wireless Sensor Network)
 Industrial Monitoring
 Applications of WSN in Industry
 Challenges & Design Goals
 Standardized Activities
 Open Issues
2
WIRELESS SENSOR NETWORK
•consists of spatially distributed autonomous sensors
•cooperatively monitor physical or environmental conditions
such as temperature, sound, vibration, pressure, motion or
pollutants
3
WIRELESS SENSOR NETWORK
Sensor
4
APPLICATIONS OF WSN
military applications e.g. battlefield surveillance
 environment and habitat monitoring
 health monitoring & healthcare applications
 home automation
 traffic control
 industrial process monitoring and control
machine

5
WIRELESS SENSOR NETWORK
AT INDUSTRIES
6
INDUSTRIAL MONITORING AND
CONTROLLING

Three types of monitoring
 Process monitoring
 Staff monitoring
 Machineries monitoring and controlling

companies often use manual labor-intensive
techniques.
• increases the cost
• human errors
7
INDUSTRIAL MONITORING AND
CONTROLLING AND SENSORS

some monitoring process can not be done by
human beings


they are out of reach
it is dangerous to monitor them directly ( for
example because of RF interference/Highly caustic
or corrosive environments/High humidity levels
/Vibrations /Dirt and dust)
Without sensors these
types of monitoring are
very difficult or
impossible!!
8
APPLICATIONS OF WSN IN
INDUSTRY

Building automation

Building access controls , HVAC controls , Lighting
controllers, Thermostat , Lifts / Elevators /
Escalators , Remote alarm triggering , Water
Management, Electrical blinds
9
APPLICATIONS OF WSN IN
INDUSTRY

Industrial process automation
Water/Wastewater Monitoring
• Landfill Ground Well Level Monitoring and
Pump Counter
• Flare Stack Monitoring
• Water Tower Level Monitoring
 Vehicle Detection
 Agriculture
• Windrow Composting
• Greenhouse Monitoring

10
APPLICATIONS OF WSN IN
INDUSTRY

Electric utility automation

Monitoring device parameters

Automatic meter reading

Inventory management

Monitoring the inventory product conditions and
environment
11
MACHINE HEALTH MONITORING OR
CONDITION BASED MAINTENANCE

Condition-based maintenance (CBM)
-significant cost savings and enable new functionalities.

US Navy shipboard systems
-reduced manning levels
-automated maintenance monitoring systems.
Inaccessible locations, rotating machinery, hazardous or restricted
areas, and mobile assets can now be reached with wireless
sensors.
12
WHAT HAPPENS AT INDUSTRY

Wireless tiny sensor nodes are installed on industrial
equipment

Sensors monitor the parameters critical to each equipment

based on a combination of measurements such as vibration,
temperature, pressure, and power quality
13
WHAT HAPPENS AT INDUSTRY (CONTD.)

Data are then wirelessly transmitted to a sink node that
analyzes the data from each sensor

Any potential problems are notified to the plant personnel as
an advanced warning system.

This enables plant personnel to repair or replace equipment
before their efficiency drops or they fail entirely.

In this way, catastrophic equipment failures and the associated
repairing can be prevented in advance.
14
CHALLENGES, DESIGN GOALS AND
STATE OF ART CONDITIONS OF
WSN & IWSN
15
CHALLENGE:

RESOURCE CONSTRAINTS
Constraints
 Battery energy
 Limited memory
 Limited Processing Capabilities
 Bandwidth constraint
16
DESIGN GOAL: RESOURCE-EFFICIENT DESIGN

Energy saving with energy-efficient protocols


Energy-aware routing on network layer
Energy-saving mode on MAC layer


For certain FEC (forward error correction) codes, hop-length
extension decreases energy consumption
Hardware optimizations
Sleeping schedules to keep electronics inactive most of the time,
dynamic optimization of voltage, and clock rate
 System-on-chip (SOC) technology for low power consumption by
integrating a complete system on a single chip ( ZigBee SOC,
CC2430, EM250)

•
Local data processing
17
DESIGN GOAL:
RESOURCE-EFFICIENT
DESIGN

Energy Recovery/Acquisition: Energy harvesting
technique
 Extracts energy from environment
Some approaches

Photovoltaic cell with rechargeable battery

Background radio signal: small energy
 vibrations, thermoelectric conversion, human
body

RF signal transmission: safety issue

employing piezoelectric materials
18
CHALLENGE: DATA REDUNDANCY

High Density in network topology cause redundant
data in both spatial and temporal domain

Spatial correlation: redundant data possibly from nearby
sensors
 Temporal correlation: redundant data from consecutive
observation
19
DESIGN GOALS:
DATA FUSION AND
LOCALIZED PROCESSING
Data aggregation and fusion


Locally filter the sensed data and transmit only the
processed one
Only necessary information is transported to the end-user
Intermediate node checks the contents of incoming
data and then combines them by eliminating
redundant information under some accuracy
constraints
20
PACKET ERRORS AND
VARIABLE-LINK CAPACITY
CHALLENGE:

Attainable capacity and delay at each link depends
on




Location
Interference level perceived at the receiver
Varying characteristics of the link over space and
time due to obstructions and noisy environment
High bit error rates
21
INTERFERENCE


Broadband interference

Generated by motors, inverters, computers, electric-switch
contacts, voltage regulators, pulse generators, thermostats,
and welding equipment

Have constant energy spectrum over all frequencies and
high energy

Emitted unintentionally from radiating sources
Narrowband interference

Intentional and have less Energy

Caused by UPS system, electronic ballasts, test
equipment, cellular networks, radio–TV transmitters, signal
generators, and micro wave equipment
22
DESIGN GOALS:
FAULT TOLERANCE AND
RELIABILITY

Sensed data should be reliably transferred to the sink node (specially
mission-critical information)

Programming/command and queries should be reliably delivered to
the target sensor node to assure the proper functioning

To combat the unreliability, verification and correction on each
communication layer are required
automatic repeat request (ARQ): not suitable for real time system
 forward error correction (FEC)
 hybrid schemes.

23
DESIGN GOAL: FAULT TOLERANCE AND
RELIABILITY

Forward error correction (FEC)

Improve the error resiliency more than ARQ
Radio-modulation
techniques to reduce interferences
and improve reliability
Direct- sequence spread spectrum
 Frequency-hopping spread spectrum
Benefits of SSM:




Multiple access
Anti-multipath fading
Anti jamming
24
CHALLENGE:
SECURITY
Security for external attacks and intrusion

Passive attacks: eavesdropping on transmissions , traffic
analysis, disclosure of message contents

Active attacks: modification, fabrication, and interruption
(in case of IWSN, node capturing, routing attacks, or
flooding)

External denial-of-service attacks and intrusion
25
DESIGN GOAL: SECURE DESIGN


Low level and high level security should be addressed

key establishment and trust control, secrecy and
authentication, privacy, robustness to communication DoS,
secure routing, resilience to node capture

secure group management, intrusion detection, secure data
aggregation
Security overhead should be balanced against QoS
26
CHALLENGE: DYNAMIC TOPOLOGIES AND
HARSH ENVIRONMENTAL CONDITIONS
In harsh industrial environments, the topology and
connectivity of the network may vary due to
 link and sensor-node failures
 a portion of sensor nodes to malfunction
27
DESIGN GOAL:
ADAPTIVE NETWORK
OPERATION

Adaptability enables to cope with dynamic wirelesschannel conditions and new connectivity requirements for
new industrial processes

Adaptive signal-processing algorithms and
communication protocols are required to balance the
trade offs among
Resources
 Accuracy
 Latency
 time synchronization requirements

28
CHALLENGE: QUALITY-OF-SERVICE
REQUIREMENTS

Accuracy between the data reported and what is
actually occurring in the industrial environment

Time sensitive data should be reached in a timely
manner

Different IWSNs have different QoS requirements
and specifications
29
DESIGN GOAL: APPLICATION-SPECIFIC
DESIGN AND
TIME SYNCHRONIZATION

Designs and techniques should be based on the
application-specific QoS requirements

Existing time synchronization strategies designed
for other traditional wired and wireless networks
may not be appropriate for IWSNs due to:
resource and size limitations
 lack of a fixed infrastructure
 dynamic topologies


Adaptive and scalable time-synchronization
protocols are required for IWSNs
30
CHALLENGE: LARGE-SCALE DEPLOYMENT AND
AD HOC ARCHITECTURE

Large number of sensor nodes

Randomly spread over the deployment field

Need for autonomous establishment of connections and
maintenance of network connectivity
31
DESIGN GOAL: LOW-COST AND SMALL SENSOR
NODES AND
SELF-CONFIGURATION AND SELFORGANIZATION

To accomplish large scale deployments feasible hardware cost
should be minimized

Commercial release:
Smart Dust motes
 uAMPS
 CC2430 and EM250
 ZigBee SOC


self-organizing architectures and protocols are required for
supporting the dynamic topologies caused by node failure/mobility/
temporary power-down/addition of new nodes
 large-scale node deployments

32
CHALLENGE: INTEGRATION WITH
INTERNET AND OTHER NETWORKS
IWSN needs to provide service for querying the network
to retrieve useful information from anywhere and
anytime

Should be remotely accessible from the Internet

Need to be integrated with the Internet Protocol(IP)
architecture
33
DESIGN GOAL: SCALABLE ARCHITECTURES AND
EFFICIENT PROTOCOLS
•
Needs to support heterogeneous industrial applications

necessary to develop flexible and scalable architectures to
accommodate the requirements of various applications in the
same infrastructure
•
Modular and hierarchical systems
•
Interoperability with existing legacy solutions such as
fieldbus and Ethernet-based systems
34
SOFTWARE DEVELOPMENT: API

Should be accessible through a simple
application programming interface

Should make the underlying network complexity
transparent to the end users

Should be able to integrate seamlessly with the
legacy fieldbus
35
SOFTWARE DEVELOPMENT: OPERATING
SYSTEM AND MIDDLEWARE DESIGN

Operating system should balance the tradeoff between
energy and QoS requirements


Tiny OS
 component-based development
 flexible platform for implementing new communication protocols
 supports communication, multitasking, and code modularity
Middleware should provide efficient network and system
management




abstracts the system as a collection of massively distributed objects
enables industrial sensor applications to originate queries and tasks,
gather responses and results,
monitors the changes within the network
36
SOFTWARE: SYSTEM INSTALLATION AND
COMMISSIONING

During installation, what and where a sensor will monitor,
should be indicated

Network management and commissioning tools should be
provided by software


for example: a graphical user display to show network connectivity
and help to set the operational parameters
Network performance analysis and other management
features

detecting failed nodes, assigning sensing tasks, monitoring
network health, upgrading firmware, and providing QoS
provisioning
37
NETWORK ARCHITECTURE

Network should be scalable

Flexible and hierarchical architectures

should accommodate the requirements of both
heterogeneous and homogeneous infrastructure

flat single-tier network of homogeneous sensor nodes

Multi-tier heterogeneous approaches
(clustering/partitioning)
resource-constrained low-power elements are in charge of
performing simpler tasks, such as detecting scalar physical
measurements
 resource-rich high-power devices (such as gateways)
perform more complex tasks

38
CROSS-LAYER DESIGN

IWSNs demands
 Cross layer optimization (physical, MAC, and routing
layers optimization) due to
Technical challenges caused by harsh industrial conditions
 Application specific QoS requirements


Methodologies to


Leverage potential improvements of exchanging information between
different layers of the communication stack
Some form of logical separation of these functionalities
should be kept to preserve modularity
39
STANDARDIZATION ACTIVITIES

ZigBee



Advantages



A mesh-networking standard based on IEEE 802.15.4 radio
technology
Targeted at industrial control and monitoring, building and home
automation, embedded sensing, and energy system automation
Extremely low energy consumption
Support different topologies
Disadvantage

Cannot serve the high number of nodes within the specified
cycle time
40
STANDARDIZATION ACTIVITIES

Wireless HART

Specifically designed for process monitoring and control

Employs IEEE 802.15.4-based radio, frequency hopping,
redundant data paths, and retry mechanism

Utilize mesh networking, both transmission and relay
41
STANDARDIZATION ACTIVITIES

UWB

Short-range transmission of very short impulses emitted in periodic
sequences

Used in Multimedia and personal area networking, now trying in industries
Advantages:
•
Good localization capabilities
•
Share previously allocated radio frequency bands by hiding signals under
noise floor
•
Transmit high data rates with low power
•
Good security characteristics
•
Ability to cope with multipath environments
42
STANDARDIZATION ACTIVITIES: UWB
(Cont.)
Disadvantage:

Not viable for longer distance communication or measuring data
from unsafe zone
Challenges:
 Hardware development
 Handling MAC and multipath interference
 Understanding propagation characteristics
43
STANDARDIZATION ACTIVITIES (CONT..)

IETF6LoWPAN

Aims for standard IP communication over low power wireless
IEEE 802.15.4 networks utilizing IPV6
Advantages :
•
Communicate directly with other IP in wireless sensor
devices
•
Established application level model and services (e.g., HTTP,
HTML, XML)
•
Established network-management tools
•
Transport protocols
44
•
Support for IP option
STANDARDIZATION ACTIVITIES (CONT..)

ISA100


Targeted for reliable communication system for monitoring
and control applications
Bluetooth and Bluetooth Low Energy

Ultralow-power technology address very low battery
capacity
45
OPEN ISSUES
 To

devise analytical models
to evaluate and predict IWSNs performance
characteristics, such as communication latency and
reliability and energy efficiency
 Optimal
sensor-node deployment
 localization,
security, and interoperability
between different IWSN manufacturers
46
OPEN ISSUES

To cope with RF interference and dynamic wireless
channel conditions in industrial environments

Porting a cognitive radio paradigm to a low power
industrial sensor node

Developing controlling mechanisms for channel hand-off

Because of the diverse industrial application
requirements and large scale of the network, several
technical problems still remain to be solved in analytical
IWSN models
47
POSSIBLE SOLUTIONS
???
48