single node architecture - notesinterpreter

Single Node ArchitectureCHAPTER-2

Hardware Components

Choosing the hardware components for awireless sensor node is dependent onapplication requirement.

Selection of hardware also depends onsize, cost and energy consumption of thenodes

In some extreme cases the entire sensornode should be◦ Smaller than 1cubic centimetre◦ Weigh less than 100 gm◦ Cheaper than $1◦ Dissipate less than 100 micro watt

Sensor node Hardware overview

Main components of a WSN node

Controller

Communication device(s)

Sensors/actuators

Memory

Power supply

Sensor node Hardware overview

Main components of a WSN node◦ Controller

Processes all relevant data Capable of executing arbitrary code

◦ Communication device(s) Device for sending and receiving data over a wireless channel

◦ Sensors/actuators The actual interface to the physical world; Devices that can observe or control physical parameters of the

environment.

◦ Memory Stores data and programs Often different types of memory are used for programs and data

◦ Power supply Some form of batteries to provide energy Sometime recharging by obtaining energy from the environment,

e.g. solar cells

Basic Node Architecture-

Controllers Controllers are core of a wireless sensor

node and known as CPU of the node

These◦ Collects data from the sensors

◦ Process the collected data

◦ Decides when and where to send the data

◦ Receives data from the other sensor nodes

◦ Decides on the actuator’s behaviour

Controller has to execute variousprograms ranging from time critical signalprocessing and communication protocol toapplication programs

Main Options for Controllers

Microcontroller◦ For the sensor nodes mostly microcontroller are used◦ It is a general purpose processor, optimized for embedded

applications, low power consumption

◦ Examples Intel StrongARM

High-end processor often used in PDAs SA-1100 model: 32-bit RISC core, running @206MHz

Texas Instruments MSP 430 Intended for usage in embedded applications 16-bit RISC core, up to 4 MHz, 2-10 kB RAM, several DACs, RT clock

Atmel ATMega ATMega 128L: Intended for usage in embedded applications 8-bit controller, larger memory than MSP430, but slower

Main options for controllers Digital signal processor(DSPs)

◦ These are familiar because of their architecture and instruction set forprocessing large amount of vector data

◦ In wireless sensor node, DSP could be used to process data comingfrom simple analog, wireless communication device to extract a digitaldata stream.

◦ Usually not used in WSN

FPGA(Field Programmable Gate Arrays)

◦ Flexible, can be reprogrammed to adapt to a changing set ofrequirements.

◦ But take time and energy.

◦ Hence not as feasible as microcontrollers.

ASIC(Application Specific Integrated Circuit)

◦ These are used only when peak performance is needed

◦ Ex: High speed routers and switches

◦ As it includes more h/w, it is costlier

Basic Node Architecture-Memory

RAM (Random Access Memory)◦ To store intermediate sensor readings, packets from other

nodes and so on.

◦ Is fast, but looses content if power supply is interrupted

ROM (Read-Only Memory)◦ To store fixed programs; not writeable

EEPROM or Flash Memory◦ Programs are stored

◦ Enables overwriting of data

◦ Can be used as intermediate storage if RAM is insufficient or if the RAM’s power supply should be turned-off

◦ BUT: long read/write access delays, high energy requirement


Communication Device Communication device is used to exchange

data between individual nodes

Choices of transmission medium

◦ In some applications wired communication can be used.(used in Profibus)

◦ The wireless communication is more considerable

◦ Wireless communication chooses any one of the following as transmission medium

Radio frequencies(most used in WSN)

Optical communication

Ultra sound

Magnetic inductance(less used)


Communication Device Among the above Radio Frequency(RF)

is best because it provides◦ Long range

◦ High data rates

◦ Acceptable error rates at reasonable energy expenditure

◦ Does not require line of sight between sender and receiver

For a practical wireless RF based system the carrier frequency has to be carefully chosen


Communication Device

Transceivers

◦ Both a transmitter and a receiver are required in a sensor node

◦ For practical purposes these two tasks are often combined in one entity, the so called transceiver

◦ It convert a bit stream coming from amicrocontroller(or a sequence of bytes or frames)and convert them to and from radio waves

◦ Usually half-duplex operation is used, because transmitting and receiving at the same time is impractical in the wireless medium.

Transceivers A range of low cost transceiver is commercially

available.

It includes all the circuitry required fortransmitting and receiving modulation,demodulation, amplifiers, mixers, filters and soon.

Usually it uses half duplex mode becausetransmitting and receiving at the same time isimpractical in the wireless medium

Transceiver states Transceivers can be put into different operational

states:◦ Transmit

In this state transceiver is active and the antenna radiates energy

◦ Receive In this state receive part is active

◦ Idle In this state transceiver ready to receive but currently no

receiving

◦ Sleep Significant parts of the receiver are switched off

The sensor node’s protocol stack and operatingsoftware must decide into which state thetransceiver is switched, according to the currentand anticipated communications needs.

Transceiver Structure

A fairly common structure of transceivers is into theRadio Frequency (RF) front end and the baseband part:

The radio frequency front end performs analog signalprocessing in the actual radio frequency band

The baseband processor performs all signal processingin the digital domain and communicates with a sensornode’s processor or other digital circuitry.


Between these two parts, a frequency conversion takes place, either directly or via one or several Intermediate Frequency (IFs).

The boundary between the analog and the digital domain is constituted by Digital/Analog Converters (DACs) and Analog/Digital Converters (ADCs).

Transceiver Structure The RF front end performs analog signal processing in the actual

radio frequency band, for example in the 2.4 GHz Industrial, Scientific,and Medical (ISM) band;

It is the first stage of the interface between the electromagnetic wavesand the digital signal processing of the further transceiver stages.


Some important elements of an RF front ends architecture are:◦ The Power Amplifier (PA) accepts up converted signals from the

IF or baseband part and amplifies them for transmission over theantenna.

◦ The Low Noise Amplifier (LNA) amplifies incoming signals up tolevels suitable for further processing without significantly reducingthe SNR. Without management actions, the LNA is active all the timeand can consume a significant fraction of the transceiver’s energy.

◦ Elements like local or voltage-controlled oscillators and mixersare used for frequency conversion from the RF spectrum to IFs or tothe baseband.

◦ The incoming signal at RF frequencies fRF is multiplied in a mixerwith a fixed-frequency signal from the local oscillator (frequency fLO).The resulting intermediate-frequency signal has frequency fLO− fRF.

Transceiver tasks and characteristics

To select appropriate transceiver a number of characteristics should be considered

1.Service to upper layer◦ A receiver has to offer certain services to the upper layers

generally to MAC layer

◦ Service is packet oriented

◦ Transceiver provides byte interface or bit interface to the microcontroller

2.Power consumption and energy efficiency◦ The simplest interpretation of energy efficiency is the energy

required to transmit and receive a single bit.

◦ The idle power consumption in each of these states andduring switching between them is also very important.


3.Carrier frequency and multiple channels◦ Transceivers are available for different carrier frequencies and it

must match application requirements

◦ If transceiver provides several carrier frequencies it help insolving congestion problem

◦ Ex: MAC protocol (FDMA or multichannel CSMA/ALOHAtechnique)

4.State change times and energy◦ Transceiver can operate in different modes, sending or

receiving using different channels or different power safestates

◦ The time and energy required to change between twostates are important figures of merit

◦ The turnaround time between sending and receiving isimportant


5.Data rates◦ Carrier frequency along with bandwidth, modulation and

coding determine the gross data rate

◦ Typical values are bits per second(bps), kilo bits persecond(kbps), mega bits per second(mbps)

◦ Data rates are less in broadband and sufficient for WSN


6. Modulations◦ Transceivers typically support one or more on/off

ASK,FSK keying or similar modulations

◦ If several modulations are available it should beselected based on experiments

7.Coding◦ Some transceivers allow various coding schemes to be

selected

8.Transmission power control◦ Some transceivers can directly provide control over

the transmission power to be used

◦ Some require external circuitry for power control


9.Noise Figure(NF)◦ It is defined as the ratio of signal to noise ratio at the

input(SNRi) of the element to the SNR at the element’soutput(SNRo)

NF=SNRi/SNRo

◦ SNR is given in dB(decibel)

◦ NF dB = SNRi dB – SNRo dB

10.Gain◦ It is the ratio of the output signal power to the input

power and is given in dB

◦ Amplifiers with high gain are desirable to achieve good

energy efficiency


13.Range◦ The range of transmitter should be clear

◦ And it should be interference free

◦ It depends on

the maximum transmission power

Antenna characteristics

Attenuation caused by the environment

Carrier frequency

◦ Modulation/coding scheme

◦ Bit error rate

◦ Quality of the receiver

◦ Typical values are difficult to give but the ranges between

few meters and several hundreds of meters are available


14.Blocking performance

◦ The blocking performance of a receiver is its achieved bit error rate in the presence of an interferer

◦ Blocking performance can be improved by interposing a filter between antenna and transceiver

15.Out of band emission◦ The inverse to adjacent channel suppression is the out of

band emission of a transmitter

◦ The transmitter should produce little transmission power outside of its prescribed bandwidth


16. Carrier sense and RSSI(Received Signal Strength Indicator)

◦ In many MAC protocols sensing whether the wireless channel is busy or not is critical

◦ The receiver should able to provide that information

◦ The carrier sense signal depends on the implementation

◦ The signal strength at which an incoming data packet has been received can provide useful information


17. Frequency stability◦ It denotes the degree of variation from nominal center frequencies

when environmental conditions of oscillators like temperature orpressure change.

◦ In extreme cases, poor frequency stability can break downcommunication links.

◦ For example, when one node is placed in sunlight whereas itsneighbor is currently in the shade.

18.Voltage Range◦ Transceivers should operate reliably over a range of supply voltages.

◦ Otherwise, inefficient voltage stabilization circuitry is required.

Example radio transceivers RFM TR1000 family

◦ 916 or 868 MHz

◦ 400 kHz bandwidth

◦ Up to 115,2 kbps

◦ On/off keying or ASK

◦ Dynamically tuneable output power

◦ Maximum power about 1.4 mW

◦ Low power consumption

Chipcon CC1000◦ Range 300 to 1000 MHz, programmable in 250

Hz steps

◦ FSK modulation

◦ Provides RSSI

Example radio transceivers Chipcon CC 2400

◦ Implements 802.15.4

◦ 2.4 GHz, DSSS modem

◦ 250 kbps

◦ Higher power consumption than above transceivers

Infineon TDA 525x family

◦ E.g., 5250: 868 MHz

◦ ASK or FSK modulation

◦ RSSI, highly efficient power amplifier

◦ Intelligent power down, “self-polling” mechanism

◦ Excellent blocking performance

Basic Node Architecture-Sensors Sensors can be roughly categorized into three categories:

◦ Passive, omnidirectional sensors

These sensors can measure a physical quantity at the pointof the sensor node without actually manipulating theenvironment by active probing -In this sense, they arepassive.

Moreover, some of these sensors actually are self-poweredin the sense that they obtain the energy they need from theenvironment

Energy is only needed to amplify their analog signal. Thereis no notion of “direction” involved in these measurements.

Typical examples for such sensors include: Thermometer, light sensors, microphones

Humidity, mechanical stress or tension in materials, chemical sensorssensitive for given substances, smoke detectors, air pressure, and soon.

Basic Node Architecture-Sensors Sensors can be roughly categorized into three categories

◦ Passive, narrow-beam sensors These sensors are passive as well, but have a well-defined notion

of direction of measurement.

A typical example is a camera, which can “take measurements” ina given direction, but has to be rotated if need be.

◦ Active sensors This last group of sensors actively probes the environment.

For example, a sonar or radar sensor or some types of seismic sensors, which generate shock waves by small explosions.

These are quite specific – triggering an explosion is certainly not a lightly undertaken action – and require quite special attention.

Basic Node Architecture-Actuators

Sensor:◦ a device that converts a physical parameter to an electrical output.

Actuator: ◦ a device that converts an electrical signal to a physical output

◦ to open or close a switch or a relay or to set a value in some way. Whether this controls a motor, a light bulb, or some other

◦ physical object is not really of concern to the way communication protocols are designed.

Power supply of sensor nodes

There are essentially two aspects:

◦ First, storing energy and providing power in the required form;

◦ Second, attempting to replenish(refill) consumed energy by “scavenging” it from some node-external power source over time.

Storing power is conventionally done using batteries.

Traditional batteries

The power source of a sensor node is a battery, either non rechargeable (“primary batteries”) or,

if an energy scavenging device is present on the node, also rechargeable (“secondary batteries”).

Batteries are electro-chemical stores for energy

The chemicals being the main determining factor of battery technology.

Batteries-Requirements

1. Capacity ➢ They should have high capacity at a small weight, small

volume, and low price.

➢ The main metric is energy per volume, J/cm3.


2. Capacity under load

➢They should withstand various usage patterns as asensor node can consume quite different levels ofpower over time and actually draw high current incertain operation modes.

3. Self-discharge

➢ Their self-discharge should be low; they mightalso have to last for a long time.

➢ Zinc-air batteries, for example, have only a veryshort lifetime (on the order of weeks), whichoffsets their attractively high energy density.


4. Efficient recharging

➢ Recharging should be efficient even at low and intermittentlyavailable recharge power; consequently, the battery shouldalso not exhibit any “memory effect”.

➢Some of the energy-scavenging techniques are only able toproduce current in the μA region (but possibly sustained) atonly a few volts at best.


5. Relaxation

➢ Their relaxation effect – the seeming self-recharging of an empty or almost empty battery when no current is drawn from it, based on chemical diffusion processes within the cell .

➢ Battery lifetime and usable capacity is considerably extended if this effect is leveraged.

➢ As but one example, it is possible to use multiple batteries in parallel and “schedule” the discharge from one battery to another, depending on relaxation properties and power requirements of the operations to be supported


6. Unconventional energy stores

➢Apart from traditional batteries, there are also other forms ofenergy reservoirs that can be contemplated.

➢ In a wider sense, fuel cells also qualify as an electro-chemicalstorage of energy, directly producing electrical energy byoxidizing hydrogen or hydrocarbon fuels.

➢Fuel cells actually have excellent energy densities (e.g. methanolas a fuel stores 17.6 kJ/cm3)

➢But currently available systems still require a non negligibleminimum size for pumps, valves, and so on.

➢Use miniature versions of heat engines, for example, a turbine.

➢ Another option are so-called “gold caps”, high-quality and high-capacity capacitors, which can store relatively large amounts ofenergy, can be easily and quickly recharged, and do not wear outover time.


7. DC-DC conversion◦ Unfortunately, batteries alone are not sufficient as a direct

power source for a sensor node.

◦ One typical problem is the reduction of a battery’s voltageas its capacity drops. Consequently, less power is deliveredto the sensor node’s circuits, – a node on a weak batterywill have a smaller transmission range than one with a fullbattery.

◦ A DC – DC converter can be used to overcome this problemby regulating the voltage delivered to the node’s circuitry.

◦ The DC – DC converter does consume energy for its ownoperation, reducing overall efficiency.

◦ But the advantages of predictable operation during theentire life cycle can outweigh these disadvantages.

Energy scavenging

Some of the unconventional energy storesdescribed above – fuel cells, micro heatengines, radioactivity– once the fuel supply isexhausted, the node fails.

To ensure truly long-lasting nodes andwireless sensor networks, such a limitedenergy store is unacceptable.

Rather, energy from a node’s environmentmust be tapped into and made available tothe node – energy scavenging should takeplace.

Energy scavenging-Approaches1. Photovoltaics

◦ The well-known solar cells can be used to power sensor nodes.

◦ The available power depends on whether nodes are used outdoors orindoors, and on time of day and whether for outdoor usage.

◦ The resulting power is somewhere between 10 μW/cm2indoors and 15mW/cm2outdoors.

◦ Single cells achieve a fairly stable output voltage of about 0.6 V and havetherefore to be used in series.

2. Temperature gradients◦ Differences in temperature can be directly converted to electrical

energy.

◦ Theoretically, even small difference of, for example, 5 K can produceconsiderable power, but practical devices fall very short of theoreticalupper limits

◦ Seebeck effect-based thermoelectric generators are commonlyconsidered;

- one example is a generator, that achieves about 80 μW/cm,at about

1 V from a 5 Kelvin temperature difference.

Energy scavenging-Approaches

3. Vibrations◦ One almost pervasive form of mechanical energy is

vibrations: walls or windows in buildings are resonating with cars or

trucks passing in the streets.

◦ The available energy depends on both amplitude andfrequency of the vibration and ranges from about0.1 μW/cm3 up to 10, 000 μW/cm3 for some extremecases.

◦ Converting vibrations to electrical energy can beundertaken by various means, based onelectromagnetic, electrostatic, or piezoelectricprinciples.

Energy scavenging-Approaches Figure shows, as an example, a generator based on a variable

capacitor. Practical devices of 1 cm3 can produce about 200μW/cm3 from 2.25 m/s2, 120 Hz vibration sources, actuallysufficient to power simple wireless transmitters

Energy scavenging-Approaches

4. Pressure variations

◦ Somewhat akin to vibrations, a variation of pressurecan also be used as a power source. Such piezoelectricgenerators are in fact used already.

◦ One well-known example is the inclusion of apiezoelectric generator in the heel of a shoe, togenerate power as a human walks about . This devicecan produce, on average, 330 μW/cm2.

5. Flow of air/liquid◦ Another often-used power source is the flow of air or

liquid in wind mills or turbines.

◦ The challenge here is again the miniaturization, butsome of the work on millimeterscale

Energy Supply - Comparison

Energy consumption of sensor

nodes

Rough estimation

◦ Number of instructions

Energy per instruction: 1 nJ

Small battery (“smart dust”): 1 J = 1 Ws

Corresponds: 109 instructions!

◦ Lifetime

Or: Require a single day operational lifetime = 24x60x60 = 86400 s

1 Ws / 86400s 11.5 W as max sustained power consumption!

Multiple Power Consumption Modes

Do not run sensor node at full operation all the

time

◦ If nothing to do, switch to power safe mode

Typical modes

◦ Controller: Active, idle, sleep

◦ Radio mode: Turn on/off transmitter/receiver, both

◦ Strongly depends on hardware

Questions:

◦ When to throttle down?

◦ How to wake up again?

Energy Consumption Figures

TI MSP 430 (@ 1 MHz, 3V):

◦ Fully operation 1.2 mW

◦ One fully operational mode + four sleep modes

◦ Deepest sleep mode 0.3 W – only woken up by

external interrupts (not even timer is running any

more)

Atmel ATMega

◦ Operational mode: 15 mW active, 6 mW idle

◦ Six modes of operations

◦ Sleep mode: 75 W

Switching Between Modes

Simplest idea: Greedily switch to lower

mode whenever possible

Problem: Time and power consumption

required to reach higher modes not

negligible

Pactive

Psleep

timeteventt1

Esaved

tdown tup

Eoverhead

Should We Switch?

Switching modes is beneficial if

which is equivalent to

−

++− up

sleepactive

sleepactive

downeventPP

PPtt tt

2

1)( 1

Eoverhead < Esaved

Computation vs. Communication

Energy Cost

Sending one bit vs. running one instruction

◦ Energy ratio up to 2900:1

◦ I.e., send & receive one KB = running three million instruction

So, try to compute instead of communicate whenever possible

Key technique – in-network processing

◦ Exploit compression schemes, intelligent coding schemes, aggregate data, …

Operating System Challenges

Usual operating system goals◦ Make access to device resources abstract

(virtualization)

◦ Protect resources from concurrent access

Usual means ◦ Protected operation modes of the CPU

◦ Process with separate address spaces

These are not available in microcontrollers◦ No separate protection modes, no MMU

◦ Would make devices more expensive, more power-hungry

Possible OS Options

Try to implement “as close to an operating system” on WSN nodes

◦ Support for processes!

◦ Possible, but relatively high overhead

Stay away with operating system

◦ There is only a single “application” running on a WSN node

◦ No need to protect malicious software parts from each other

◦ Direct hardware control by application might improve efficiency

Possible OS Options

Currently popular approach

No OS, just a simple run-time environment

◦ Enough to abstract away hardware access details

◦ Biggest impact: Unusual programming model

Concurrency Support

Simplest option: No

concurrency, sequential

processing of tasks

◦ Risk of missing data

◦ Should support

interrupts/asynchronous

operations

Poll sensor

Process

sensor

data

Poll transceiver

Process received

packet

Processes/Threads

Based on interrupts,

context switching

Difficulties

◦ Too many context

switches

Most tasks are short anyway

◦ Each process required its

own stack

Handle sensor

process

Handle packet

process

OS-mediated

process switching

Event-Based Concurrency Event-based programming model

◦ Perform regular processing or be idle

◦ React to events when they happen immediately

◦ Basically: interrupt handler

Must not remain in interrupt handler too long

◦ Danger of loosing events

Idle/regular

processingRadio event handler

Sensor event

handler

Components Instead of Processes

An abstraction to group functionality

Typically fulfill only a single, well-defined function

◦ E.g., individual functions of a networking protocol

Main difference to processes: Component does not have an execution

Components access same address space, no protection against each other

Event-based Protocol Stack

Usual networking API: sockets

◦ Issue: blocking calls to receive data

◦ Not match to event-based OS

API is therefore also event-based

◦ E.g., Tell some component that some other component wants to be informed if and when data has arrived

◦ Component will be posted an event once this condition is met

single node architecture - notesinterpreter

Documents