Transcript Node Localization in Sensor Networks

MAC, Physical Layer, Energy
Consumpion and IEEE 802.15.4
Lecture 8
September 28, 2004
EENG 460a / CPSC 436 / ENAS 960
Networked Embedded Systems &
Sensor Networks
Andreas Savvides
[email protected]
Office: AKW 212
Tel 432-1275
Course Website
http://www.eng.yale.edu/enalab/courses/eeng460a
Announcements
 Appointment schedule for projects
 Student presenter for Oct 12 – Diffusion routing
 Project proposal
• 1 page description of your project (including references)
• Should include:
o What is the problem your solving and what is the new feature that you are
adding to the problem
– Narrow down the problem you will be working on, be very precise with what you
are going to do
o Give an initial list of paper references on which your paper will be based
o A list of resources that you will need for the project (any additional HW, SW
and sensors)
• Do not exceed 1-page!!!!
• Email to [email protected]
o Filename: name1_and_name2_proposal
o Email Subject: EENG460 Project Proposal
Frequency Bands and Data Rates
In 2.4GHz band 62.5 ksymbols/second
• 1 symbol is 4 bits
• 1 symbol is encoded into a 32-bit pseudorandom sequence the chip
chip rate = 62.5 x 32 = 2000 kchips/s
Raw data rate = Symbol rate * chips per symbol
= 62.5 * 4 = 250kb/s
• In 868/915 MHz bands
1 bit symbol (0 or 1) is represented by a 15-chip sequence
Physical Layer Transmission Process
Binary Data from
PPDU
Bit to Symbol
Conversion
Symbol to Chip
Conversion
O-QPSK
Modulator
RF Signal
Radio Characteristics
 Power output
• The standard does not specify a power output limit.
• Devices should be able to transmit -3dBm
o In US 1Watt limit in Europe 10mW for 2.4GHz band
 Receiver should be able to decode a packet with
receive power of
• -85dBm in 2.4GHz and -92dBm in the lower frequency
bands
 What does that mean in terms of range?
Going from Watts to dBm
P(in mW)
P(in dBm)  10log
1mW
+20dBm=100mW
+10dBm=10mW
+7dBm=5mW
+6dBm = 4mW
+4dBm=2.5mW
+3dBm=2mW
0dBm=1mW
-3dBm=.5mW
-10dBm=.1mW
Friss Free Space Propagation Model
PR
  
 c 
 GT GR 
  GT GR 

PT
 4d 
 4d 
2
2
PT and PR - power values at thereceivingand transmitting antennas(in watts)
GT and GR are thepower gains for thetransmitting and receivingantenna
 - wavelength in meters
c - speed of light
d - distancebetween receiverand transmitter
Same formula in dB path loss form (with Gain constants filled in):
LB (dB)  32.44  20log10 f MHz  20log10 dkm
How much is the range for a 0dBm transmitter 2.4 GHz band transmitter
and pathloss of 92dBm?
Friss Free Space Propagation Model
PR
  
 c 
 GT GR 
  GT GR 

PT
 4d 
 4d 
2
2
PT and PR - power values at thereceivingand transmitting antennas(in watts)
GT and GR are thepower gains for thetransmitting and receivingantenna
Highlyh idealized
 - wavelengt
in meters model. It assumes:
• Free space, Isotropic antennas
c - speed
light
• of Perfect
power match & no interference
• Represent
the theoretical
max
d - distance
between receiver
and transmitt
er transmission range
Same formula in dB path loss form:
LB (dB)  32.44  20log10 f MHz  20log10 dkm
How much is the range for a 0dBm transmitter 2.4 GHz band transmitter
and pathloss of 92dBm?
Propagation Mechanisms in Space with Objects
 Reflection
• Radio wave impinges on an object >> λ (30 cm @1 GHz)
• Earth surface, walls, buildings, atmospheric layers
 Diffraction
• Radio path is obstructed by an impenetrable surface with sharp
irregularities (edges)
• Secondary waves “bend” arounf the obstacle
• Explains how RF energy can travel without LOS
 Scattering
• When medium has large number of objects < λ (30cm @1 GHz)
• Similar principles as diffraction, energy reradiated in many directions
• Rough surfaces, small objects (e.g foliage, lamp posts, street signs)
 Other: Fading and multipath
A more realistic model: Log-Normal Shadowing
Model
• Model typically derived from measurements
LB (dB)  32.44  10n log10 f MHz  10n log10 dkm  X 
X  is zero- mean Gaussian r.v (in dB) with standard
deviation (in dB)
• Statistically describes random shadowing effects
• values of n and σ are computed from measured data using linear regression
• Log normal model found to be valid in indoor environments!!!
Transmit Power Levels in Chipcon CC2420
Radio supply voltage= 2.5V
And Power = I*V
= 1mW
= 43.5mW
Budgeting Battery Power
 Assuming power drain is the same for Transmitting and
Receiving = 43.5mW
 We need to power the device from a 750mAh battery for 1
year
 What is the duty cycle we need to operate at?
Budgeting Battery Power
 Assuming power drain is the same for Transmitting and Receiving =
43.5mW
 We need to power the device from a 750mAh battery for 1 year
 What is the duty cycle we need to operate at?
1 year has 365 x 24 = 8760 hours
The average current drain from the battery should be
Pavg  2.5V *86A
Average power drain
I avg  750mAh/ 8760h  86A
Computing Duty Cycle
I avg  Ton * I on  (1- Ton ) * Istby
Where
Ton  Fractionof timeeit herreceiveror transmitteris on
I on  Current drain from thebattery when eit her thereceiveror transmitteris on
Istby  Current drain from thebattery when both transmitterand receiverare off
Assuming I on  17.5mA, I stby  20A, I avg  86mA
Ton 
I avg  I stby
I on  I stby
86  20

 0.0038 0.38%
17400 20
Energy Implication
 Active transceiver power consumption more
related to symbol rate rather than raw data rate
 To minimize power consumption:
• Minimize Ton - maximize data rate
• Also minimize Ion by minimizing symbol rate
 Conclusion: Multilevel or M-ary signalling should
be employed in the physical layer of sensor
networks
• i.e need to send more than 1-bit per symbol
Radio Energy Model: the Deeper Story….
Tx: Sender
Rx: Receiver
Incoming
information
Outgoing
information
Channel
Tx
elec
Transmit
electronics
E
ERF
Power
amplifier
Rx
elec
Receive
electronics
E
 Wireless communication subsystem consists of three
components with substantially different characteristics
 Their relative importance depends on the
transmission range of the radio
Radio Energy Cost for Transmitting 1-bit of
Information in a Packet
The choice of modulation scheme is important for energy vs.
fidelity and energy tradeoff
Estart Pelec  PRF ( M )  H 
Ebit 

* 1  
L
RS * log2 M 
L
Estart  energy asssociated with radio startup
L  packetpayloadlength
H  packetheader length
Pelec  power consumption of elect roniccircuit ry
for frequencysynthesis
Rs  Symbol ratefor an M - ary modulationscheme
M  Modulationlevel
Examples
nJ/bit
nJ/bit
8000
6000
Medusa Sensor
Node (UCLA)
Nokia C021
Wireless LAN
GSM
nJ/bit
300
600
200
400
100
200
0
0
4000
2000
0
Tx
Rx
ERF Eelec
Eelec
~ 1 km
Tx
Rx
ERF Eelec
Eelec
~ 50 m
Tx
Rx
ERF Eelec
Eelec
~ 10 m
 The RF energy increases with transmission range
 The electronics energy for transmit and receive are typically
comparable
Power Breakdowns and Trends
Radiated power
63 mW (18 dBm)
Intersil PRISM II
(Nokia C021 wireless LAN)
Power amplifier
600 mW
(~11% efficiency)
Analog electronics
240 mW
Digital electronics
170 mW

Trends:
 Move functionality from the analog to the digital electronics
 Digital electronics benefit most from technology improvements

Borderline between ‘long’ and ‘short’-range moves towards shorter
transmit distances
What is wrong with this model?
 Does not include many parameters
• DC-DC converter inefficiencies
• Overhead for transitioning from on to standby
modes
• Different power consumptions for receiver and
transmitter
• Battery discharge properties
• Does not include the processor power and any
additional peripherals
Where does the Power Go?
Peripherals
Disk
Display
Processing
Programmable
Ps & DSPs
ASICs
(apps, protocols etc.)
Memory
Battery
DC-DC
Converter
Radio
Modem
Power Supply
RF
Transceiver
Communication
DC-DC Converter Inefficiency
Current drawn from the battery
Current delivered to the node
Battery Capacity
from [Powers95]
 Current in “C” rating: load current
normalized to battery’s capacity
o e.g. a discharge current of 1C for a capacity of 500
mA-hrs is 500 mA
Microprocessor Power Consumption
CMOS Circuits
(Used in most microprocessors)
Static Component
Bias and leakage currents
O(1mW)
Dynamic Component
Digital circuit switching inside
the processor
2
l dd clk
P  IstandbyVdd  Ileakage Vdd  IscVdd  C V f
Static
Dynamic
Power Consumption in Digital CMOS Circuits
2
Power  Istandby Vdd  Ileakage Vdd  IscVdd  ClVdd fclk
Istandby
- current constantly drawn from the power supply
Ileakage
- determined by fabrication technology
Isc
- short circuit current due to the DC path between the
supply rails during output transitions
Cl
fclk
- load capacitance at the output node
- clock frequency
Vdd
- power supply voltage
DVS on Low Power Processor
Number of gates
M
P   Ck  f  V
Dynamic Power Component
k 1
2
dd
Load capacitance of gate k
Propagation delay
Transistor gain factor
τ
VDD
 (VDD  VT )2
CMOS transistor threshold voltage
Maximum gain when voltage is lowered BUT
lower voltage increases circuit delay
Now Back to IEEE 802.15.4 MAC
 MAC supports 2 topology setups: star and peer-to-peer
 Star topology supports beacon and no-beacon structure
• All communication done through PAN coordinator
Star: Optional Beacon Structure
Generic Superframe Structure
Beacon packet transmitted by PAN Coordinator to help
Synchronization of network devices. It includes:
Network identifier, beacon periodicity and superframe structure
GTS: Guaranteed time
Slots assigned by PAN
coordinator
Star Network: Communicating with a Coordinator
Star Network: Communicating from a Coordinator
Beacon packet indicates that there
is data pending for a network device
Device sends request on a data slot
Network device has to ask coordinator
if there is data pending.
If there is no data pending the
Coordinator will respond with a zero
Length data packet
Peer-to-Peer Data Transfer
 Peer-to-peer data transfer governed by the network
layer – not specified by the standard
 Four types of frames the standard can use
•
•
•
•
Beacon frame – only needed by a coordinator
Data frame – used for all data transfers
ACK frame – confirm successful frame reception
A MAC Command Frame – MAC peer entity
controltransfers
Beacon Frame
ACK & Data Frames
ACK Frame
Data Frame
MAC Command Frame
Wrap-up Low Power MAC
 You now have enough information to do a more
detailed power consumption analysis for IEEE
802.15.4
 Need to factor in different packet structures header
and MAC overheads
 What are the issues related with low power MAC
protocols?
 Design of low power schemes for peer-to-peer
networking…
Concept of Primitives
Request: To initiate a service
Indication: Indicate an N-layer event that is significant to the used
Response: to complete a procedure previously invoked by an
indication primitive
Confirm: conveys the results of one or more associated previous
service requests
Next Lecture
 Time Synchronization
 Read the paper
[Elson02] Fine-Grained Network Time Synchronization
using Reference Broadcasts, Jeremy Elson, Lewis Girod
and Deborah Estrin, Proceedings of the Fifth Symposium
on Operating Systems Design and Implementation (OSDI
2002), Boston, MA. December 2002. UCLA Technical
Report 020008.