Wireless Sensor Networks (WSN) - uni-potsdam.de · WSN has to adapt to changes, self- monitoring, adapt operation Incorporate possible additional resources, e.g., newly deployed nodes

Wireless Sensor Networks (WSN)

Introduction

M. Schölzel

Difference to existing wireless networks

Infrastructure-based networks e.g., GSM, UMTS, …

Base stations connected to a wired

backbone network

Mobile entities communicate wirelessly to these base stations

Traffic between different mobile entities is relayed by base stations and wired backbone

Infrastructure-free networks Try to construct a network without

infrastructure, using networking abilities of the participants

This is an ad hoc network – a network constructed “for a special purpose”

Simplest example: Laptops in a

conference room – a single-hop ad hoc network

IP backbone

Server Router

Gateways

A Wireless Sensor Network

Sensornet

Internet, LAN, GSM-Network, …

Network is embedded in environment

Nodes in the network are equipped with sensing and actuation to measure/influence environment

Nodes process information and communicate it wirelessly

wireless communication wired communication

Roles of Participants in WSN

sources of data: Measure data, report them “somewhere” Typically equip with different kinds of actual sensors sinks of data: Interested in receiving data from WSN May be part of the WSN or external entity, PDA, gateway, … actuators: Control some device based on data, usually also a sink

4

WSN Application Examples

5

Goal: Collecting data Environmental Monitoring Each node measures temperature Derive a “temperature map”

Intelligent buildings, bridges, or machines Measuring vibrations Control of leakages in chemical plants Monitor mechanical stress (e.g. during

earthquakes) Predictive maintenance

Example: Parking Space Management

Each parking space is equipped with a sensor node AMR sensor senses disturbances on the magnetic field of the earth

(resistance changes, if a car is above the sensor) Hardware:

− 8-bit microcontroller − RF transceiver: CC1120

with 4kbps data rate − 3.6V Li-ion battery

Rather simple network topology: Single hop from sensor node to sink.

Other WSN Application Domains

7

Precision agriculture Bring out fertilizer/pesticides/irrigation only where needed

Medicine and health care Wearable sensor nodes monitoring movement of patients Logistics Equip goods (parcels, containers) with a sensor node Track their whereabouts – total asset management E.g. Bosch already provides commercial solutions

Telematics Provide better traffic control by obtaining finer-grained information about traffic conditions Intelligent roadside Cars as the sensor nodes

Examples Car2Car and Car2X communication

Communication Standard WLAN-p already exists

Cars have on-bord-units, infrastructure has road-side-units

Cohda-Box commercially available − 32-bit ARM-like processor − runs a Linux system − up to 27 mbps

Cars and infrastructure can send various kinds of

messages − CAMs: position, direction, speed, etc. − DENMs: Detected hazards − SPATs: State of the traffic light − …

8

Cohda OBU

WSN Application Examples

Goal: Building control loops (sensing -> computing -> acting) Communication in industrial environments Communication between mobile machines

Requirement real time demands

− Wireless heart standard much higher reliability of communication required

− probability of packet error rate ~10-9

− typical bit error rate in wireless channels: 10-3 to 10-4

wireless ad-hoc networks: problems

10

Without a central infrastructure, things become much more difficult Problems are due to Lack of central entity for organization available Limited range of wireless communication Battery-operated entities Mobility of participants

Problem: Lack of Central Entity

Without a central entity (like a base station), participants must organize themselves into a network (self-organization) Pertains to (among others): Medium access control – no base station can assign

transmission resources, must be decided in a distributed fashion

Finding a route from one participant to another

11

Problem: Limited Range

For many scenarios, communication with peers outside immediate communication range is required Direct communication limited because of distance, obstacles,

… Solution: multi-hop network

12

?

Problem: Battery-Operated Entities

Often (not always!), participants in an ad hoc network draw energy from batteries Desirable: long run time for Individual devices Network as a whole

Required: Energy-efficient networking protocols E.g., use multi-hop routes with low energy consumption (energy/bit) E.g., take available battery capacity of devices into account How to resolve conflicts between different optimizations?

13

Problem: Mobility

In many (not all!) ad hoc network applications, participants move around (-> mobile ad hoc networks (MANET)) Car2Car-communication Moving robots in industrial environments In cellular network: simply hand

over to another base station

In MANET − Mobility changes neighborhood relationship − Routes needs adaptation − Complicated by scale

• Large number of such nodes difficult to support

14

Requirements for WSN (user perspective)

Quality of Service (QoS) Detect events and deliver data reliable

− events may be detected by multiple nodes − reliability of a single node doesn’t matter, system reliability is important!

Deliver data in within a specified delay Lifetime The network should fulfill its task as long as possible – definition depends

on application Lifetime of individual nodes relatively unimportant

Fault tolerance Be robust against faults

− node failures (often permanent): running out of energy, physical destruction − link failures (often temporarily): weather, moving obstacles, …

15

Requirements for WSN (administrator perspective)

Deployment Simple deployment even for thousands of nodes (self-organization) Scalability Support large number of nodes (several thousands)

− in existing practical applications < 100

Maintainability WSN has to adapt to changes, self-monitoring, adapt operation Incorporate possible additional resources, e.g., newly deployed nodes Programmability

− Re-programming of nodes in the field might be necessary, improve flexibility, fixing bugs

16

Mechanisms to Overcome Problems and to Meet Requirements

Multi-hop wireless communication to overcome limited communication range requires self-organization mechanisms Energy-efficient operation For communication, computation, sensing, actuating Very typical: computation far less energy hungry than communication

− Collaboration & in-network processing • Nodes in the network collaborate towards a joint goal • Pre-processing data in network (as opposed to at the edge) can greatly improve efficiency

− Locality • Do things locally (on node or among nearby neighbors) as far as possible

− Data centric networking • Focusing network design on data, not on node identifies (id-centric networking) • To improve efficiency

17

Mechanisms to meet requirements

Auto-configuration Manual configuration just not an option Right after deployment But also periodically during operation to overcome

− link and node failures − mobility problems

Exploit tradeoffs during the design phase E.g., between invested energy and accuracy Design-space-exploration

18

Design Space

Due to the many options for tackling problems, WSNs offer a very large design space

Trade-offs between many design space parameters must be found in order to meet the requirements

Design space parameters:

Form Factor

Heterogeneity

Communication Modality

Sensor Coverage

Radio Coverage

Network Topology

Connectivity

Design Space

Form factor of the mote classes: brick, matchbox, grain, dust Depends on application (microscopic to shoebox) Costs may range from cents to hundreds of euro's Impacts

− life time (e.g. size of battery), processing resources, costs Heterogeneity of the WSN classes: homogeneous/heterogeneous First approach: identical nodes only In practice: a variety of nodes can be very useful

− e.g. cluster heads with more resources − special capabilities only required for some nodes (e.g. GPS) − gateways to external networks (GSM, satellite, Internet)

Impacts − Complexity of software

20

Design Space

Communication Modality classes: radio (169MHZ, 434MHz, 868MHz, …), visual light communication (VLC), ultrasound, … How do nodes communicate ? most common: radio waves, usually sub-gigahertz bands light beams or laser: smaller, more energy efficient (cf. Smart Dust) Impacts

− data rate, life time, communication range

Connectivity classes: connected / intermittent / sporadic Nodes always connected or only sometimes (regularly or sporadic)? Impacts

− Protocols, data gathering, life time

21

Design Space

Sensor-Coverage classes: sparse / dense / redundant sensors could cover only part of area of interest, or all, or the same area is

covered multiple times Impacts:

− observational accuracy, number of nodes, costs

Radio-Coverage classes: from sparse to dense / static /dynamically How many other nodes will be in the communication range? May be adapted at runtime Impacts:

− reliability, energy consumption due to collisions, transmission power, etc., number of nodes

22

Design Space

Network Topology classes: infrastructure or ad-hoc (single-hop / star / networked stars / tree / graph) infrastructure-based: motes communicate via base stations only

− often costly

ad-hoc: direct communication between nodes − no expensive infrastructure

diameter is the max number of hops between any two nodes − single hop (d=1), infrastructure based (d=2), multi-hop (d big for ad-hoc networks)

Impacts: − Communication delay, protocol complexity

23

Development Flow and Design Space Exploration

Models − used for an early evaluation of possible design

parameters

Simulations: − Evaluate the prediction of the model in the full

dynamic of the network − Test and debug to find software bugs − Not all aspects of the hardware are accurately

reflected

Testbed uses real hardware − Test the software on real hardware − Test and Debug in a distributed system difficult − Systematic test of fault-tolerance techniques not easy

(often the code is instrumented) − real sensors and actors maybe not available →

virtualization

Deployment test in a real setup is required − Not always possible − Test-bed infrastructure not available anymore

• Over-the-air update • power measurement

24

Implementation (+Software)

Simulation

Test/Debug Testbed (+Hardware)

Start with requirements

Deployment (+Topology)

Models Evaluation

Test/Debug

Required Knowledge for Developing WSN Applications

Knowledge of Models and Simulators

Microprocessor architecture and programming − Communication with peripherals − Usage of operating systems − Power saving techniques

Architecture and Design of Protocols

− Power saving techniques − Fault Tolerance

• Forward- and backward-error correction