Computer Networks II Advanced Features (T-110.5111)

(1)

Computer Networks II

Advanced Features (T-110.5111)

Wireless Sensor Networks Mario Di Francesco, PhD

Assistant Professor – DCS Research Group

(2)

Wireless sensor networks:

an introduction

 Network architecture

 Wireless sensor nodes

 Approaches to energy conservation

G. Anastasi, M. Conti, M. Di Francesco, A. Passarella, “Energy conservation in wireless sensor networks: A

survey”, Ad Hoc Networks, 7(3):537–568, May 2009 (http://dx.doi.org/10.1016/j.adhoc.2008.06.003)

(3)

Wireless sensor network

Architecture and components

Sensing field

Sensor Node Sink

(Base station) Remote

user

Internet

(4)

Computer networks II – Advanced topics

T-110.5111 – Wireless sensor networks (09.10.2013)

Mario Di Francesco

http://www.uta.edu/faculty/mariodf

Wireless sensor node

Architecture and components

ADC

Sensors Radio

Memory MCU DC-DC

Battery

Mobilizer Location Finding System

Power Generator

Power Supply Subsystem Sensing Subsystem Processing Subsystem Communication Subsystem

(5)

Wireless sensor node

Architecture and components

ADC

Battery

Power Generator

Data acquisition

from the environment

(6)

Wireless sensor node

Architecture and components

ADC

Battery

Power Generator

Local data processing

and data storage

(7)

Wireless sensor node

Architecture and components

ADC

Battery

Power Generator

Short range wireless communication

Radio is the most power hungry component

(8)

Wireless sensor node

Architecture and components

ADC

Battery

Power Generator

Battery powered devices

Batteries cannot be changed nor recharged

(9)

Examples of sensor nodes: UCB Motes

(10)

Examples of sensors (i.e., transducers)

Name Producer Type Power

consumption

STCN75 STM Temperature 0.4 mW

ADXL330 Analog Devices Accel. (3 axis) 0.2 mW SHTx Sensirion Temperature/humidity 3 mW iMEMS Analog Devices Accel. (3 axis) 30 mW 2200 and 2600

series

GEMS Pressure 50 mW

CP18, VL18, GM60, GLV30

VISOLUX Proximity 350 mW

FCS-GL1/2A4- AP8X-H1141

TURCK Flow control 1250 mW

(11)

Telos node: board and integrated circuits

(12)

Wireless sensor node

Breakdown of energy consumption

0 2 4 6 8 10 12 14 16

Power (mW)

SENSORS CPU TX RX IDLE SLEEP

Sending 1 bit of information is equivalent to process

~1000 instructions from as for energy consumption

RADIO

(13)

Wireless sensor node

Breakdown of energy consumption

0 2 4 6 8 10 12 14 16

Power (mW)

SENSORS CPU TX RX IDLE SLEEP

The power consumption of the sensor (transducer)

is not always negligible

(14)

Wireless sensor networks

Application scenarios and goals

 Data collection

– Long-term network lifetime – Self organization

 Dense networks

– Multi-hop routes

– Interference

(15)

Energy Conservation Schemes for WSNs

Duty Cycling Data-driven Mobility-based

Energy conservation in WSNs

 Mostly targeted to the radio

and the sensing (data acquisition) subsystems

(16)

Taxonomy of approaches based on duty cycling

Duty Cycling

Topology Control

Location-driven Connectivity- driven

Power Management

Sleep/Wakeup Protocols

Low-Duty Cycle

MAC Protocols

(17)

Taxonomy of

(general) sleep/wakeup protocols

Sleep/wakeup Protocols

On-demand Scheduled

Rendez-vous Asynchronous

 On demand: low-power radios, radio-triggered wakeup

 Scheduled rendez-vous: synchronized wakeup

 Asynchronous: wakeup at any time

(18)

Taxonomy of

MAC protocols with a low duty cycle

Low-Duty Cycle MAC Protocols

Time Division

Multiple Access Contention-based Hybrid

 Time Division Multiple Access: Bluetooth, TRAMA

 Contention-based: IEEE 802.15.4, B-MAC, S-MAC, T-MAC

 Hybrid: Z-MAC, Crankshaft

(19)

Channel access

based on long preambles

 Low-power listening

– Exploit transmit mode as it consumes less than receive mode – Use a duty cycle for further energy savings

– Implemented by B-MAC and derived solutions (e.g., X-MAC)

Preamble Msg

(20)

Taxonomy of data-driven approaches

Data-driven

Data Reduction

In-network Processing

Data Compression

Data Prediction

Energy-efficient Data Acquisition

Adaptive Sampling

Hierarchical Sampling

Model-based

Sensing

(21)

Example of data prediction:

differential sending strategy

 Only send messages if values differ more than 𝛿

𝒇(𝒙)

𝒕 _𝟎 𝒕 𝒚 _𝟏

𝒚 _𝟎 ^send skip skip skip skip

skip

send

skip

send skip skip skip skip

𝒕 _𝟎 + 𝟐∆𝑻 ⋯

𝜹

(22)

Example of data prediction:

model-based strategy

 Send the description (representation) of the signal

𝒇(𝒙)

𝒕 𝒚 _𝟎

𝒕 _𝟎 𝒕 _𝟏

send all messages

build and send the model

no messages

signal differs from model,

start over

(23)

Example of hierarchical sampling:

triggered sensing in smart environments

 Event-triggered image capture

– Fall detection algorithm running at an ordinary sensor – Tiered architecture with a multimedia sensor node

Ordinary Sensor (Sun SPOT)

Multimedia Sensor Prototype

Sun SPOT

(gateway) BeagleBoard

Logitech

C905

IEEE 802.15.4

(24)

Wireless sensor networks with mobile elements

 Definition and taxonomy

 Sparse wireless sensor networks

 Discovery of mobile elements

M. Di Francesco, S. K. Das, G. Anastasi, “Data Collection in Wireless Sensor Networks with Mobile Elements: A Survey”, ACM Transactions on Sensor Networks, 8(1):7, August 2011

(http://dx.doi.org/10.1145/1993042.1993049)

(25)

WSNs with Mobile Elements

 Main components

– (Regular) sensor nodes

 Perform sensing as their main task

 Sources of data – Sinks (base stations)

 Collect messages and use them or make them available

 Destination of data – Support nodes

 Special nodes performing a specific task

 They exploit mobility to support network operation

 A network where at least one of them is mobile

(26)

Mobile Data Collectors

 Mobile elements that visit the network to gather data from source nodes

 Classification

– Mobile sinks

 Both dense and sparse WSNs

– Mobile relays

 Support nodes that provide a relay (forwarding) service between source nodes and the sink

 Gather data from sensors, store them and carry them to the base station

 (Rather) sparse WSNs

(27)

Mobile sinks

Mobile Sink Mobile

Sink

(28)

Mobile relays

Mobile Relay

Sink (Base station) Mobile

Relay

(29)

Relocatable nodes

Sink (Base station)

Relocatable

node

(30)

Mobile peers

Sink

(Base

station)

(31)

Overview of data collection

in WSNs with mobile elements

 Data collection

– Exploits contacts between nodes

 Three main phases

– Discovery – Data transfer

Mobile element

Start of contact

Communication range of the node reached

by the MS

End of contact

Nodes reachable

through multi-hop

paths

(32)

Sparse wireless sensor networks

Reference scenario and sensor states

 MDC in contact with at most one sensor at any time

 Additional sleeping phase

Mobile data collector

timeout

Data Transfer Discovery

Sleeping

MDC out of reach or communication over

timeout

MDC

discovered

(33)

Communication in sparse WNS

 Nodes wait for the MDC to approach and

then transfer data

 Pros

– Decreased message loss – Nodes do not have

to relay messages – Tight synchronization

is not required

 Cons

– Increased latency

– Cost of MDCs

(34)

Discovery phase

Asynchronous protocol with duty cycle

 Mobile data collector

– Emits beacon messages periodically

 Static sensor node

– Wakes up periodically to listen for incoming beacons

Node

MDC

...

T

D

T

ON

Active

T

OFF

T

B

𝑇 _𝑂𝑁 = 𝑇 _𝐵 + 𝑇 _𝐵𝐷

𝛿 = 𝑇 _𝑂𝑁

𝑇 _𝑂𝑁 + 𝑇 _𝑂𝐹𝐹

(35)

Evolution of sensing scenarios:

from sensors to phones and things

 From sensors to smartphones

 People-centric sensing applications

 Internet of Things

(36)

Wireless sensor networks

with mobile elements revisited

Mobile Sink Mobile

Sink

(37)

From sensor devices to smartphones

 Smartphones as sensing platforms

– Abundance of sensors

 Acceleration

 Location, orientation

 Sound, video

 Proximity

– Rich in processing and storage resources

 Enabling even computationally intensive applications – Several wireless technologies

 WiFi, Bluetooth (Low Energy)

 Long range cellular radio

 Near Field Communication (NFC)

(38)

People-centric sensing scenarios

 Passive sensing scenarios

– People and communities are characterized

through data sampled by the phone during everyday life

 Can be seen as a special case of WSNs with MEs or DTNs – Also referred to as people-based sensing

 Active sensing scenarios

– People involved in sensing campaigns

– Participants instructed to actively sense the environment

 Sample applications: traffic/accidents monitoring, well being

 Incentives for participation

– Also known as participatory sensing

(39)

The Internet of Things

 Networked objects (devices)

– Deployed worldwide

– Connected over the Internet

 IoT devices

– Individually addressable

– Interconnected and accessed through the standards of the web – Not only sensors but also actuators (i.e., power switches)

 Major issues

– Heterogeneity

– Scale

(40)