10/1/2008 1 Communication Theory as Communication Theory as Applied to Wireless Sensor Networks muse Objectives • Understand the constraints of WSN and how i ti th hi i fl d communication theory choices are influenced by them • Understand the choice of digital over analog schemes • Understand the choice of digital phase Understand the choice of digital phase modulation methods over frequency or amplitude schemes muse
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
10/1/2008
1
Communication Theory asCommunication Theory as Applied to Wireless Sensor
Networks
muse
Objectives
• Understand the constraints of WSN and how i ti th h i i fl dcommunication theory choices are influenced
by them
• Understand the choice of digital over analog schemes
• Understand the choice of digital phaseUnderstand the choice of digital phase modulation methods over frequency or amplitude schemes
muse
10/1/2008
2
Objectives (cont.)
• Understand the cost/benefits of implementing source and channel coding for sensorsource and channel coding for sensor networks
• Understand fundamental MAC concepts• Grasp the importance of node synchronization• Synthesize through examples these concepts to understand impact on energy and bandwidth requirements
muse
Outline
• Sensor network constraints
• Digital modulation
• Source coding and Channel coding
• MAC
• Synchronization
• Synthesis: Energy and bandwidth requirements
10/1/2008
3
WSN Communication Constraints
• Energy!
Communication constraints
Data Collection Costs
• Sensors
• Activation
• Conditioning
• A/D
Communication constraints
10/1/2008
4
Computation Costs
• Node life support
• Simple data processing
• Censoring and Aggregation
• Source/Channel coding
Communication constraints
Communication Costs
current ~= 1.0313 * rf_power + 20.618R2 = 0.9572
30
35
n (m
A)
Communication constraints
15
20
25
-10 -5 0 5 10
RF Transmit Power (dBm)
Cur
rent
Con
sum
ptio
n
10/1/2008
5
Putting it all together
Communication constraints
Outline
• Sensor network constraints
• Digital modulation
• Source coding and Channel coding
• MAC
• Synchronization
• Synthesis: Energy and bandwidth requirements
10/1/2008
6
Modulation
• Review
• Motivation for Digital
Modulation
The Carrier
Modulation
10/1/2008
7
Amplitude Modulation (AM)
DSB‐SC (double sideband – suppressed carrier)
Modulation
Frequency representation for DSB‐SC (the math)
Modulation
10/1/2008
8
Frequency representation for DSB‐SC (the cartoon)
Modulation
Demodulation – coherent receiver
Modulation
10/1/2008
9
DSB‐LC (or AM as we know it)
Modulation
Frequency representationof DSB‐LC
Modulation
10/1/2008
10
Amplitude Modulation
Modulation
Frequency Modulation (FM)
Modulation
10/1/2008
11
Frequency Modulation
Modulation
Phase Modulation
Modulation
10/1/2008
12
SNR Performance
ModulationFig. Lathi
Digital Methods
Digital Modulation
10/1/2008
13
Quadrature Modulation
Digital Modulation
BPSK
Digital Modulation
10/1/2008
14
QPSK
Digital Modulation
Constellation Plots
Digital Modulation
10/1/2008
15
BER Performance vs. Modulation Method
Digital ModulationFig. Lathi
BER Performance vs. Number of Symbols
Digital ModulationFig. Lathi
10/1/2008
16
Outline
• Sensor network constraints
• Digital modulation
• Source coding and Channel coding
• MAC
• Synchronization
• Synthesis: Energy and bandwidth requirements
Source Coding
• Motivation
• Lossless
• Lossy
Source Coding
10/1/2008
17
Lossless Compression
• Zip files
• Entropy coding (e.g., Huffman code)
Source Coding
Lossless Compression Approaches for Sensor Networks
• Constraints
• Run length coding
• Sending only changes in data
Source Coding
10/1/2008
18
Lossy Compression
• Rate distortion theory (general principles)
• JPEG
Source Coding
Example of Lossy Compression ‐JPEG
10/1/2008
19
Another comparison
Lossy Compression Approachesfor Sensor Networks
• Constraints
• Transformations / Mathematical Operations
• Predictive coding / Modeling
Source Coding
10/1/2008
20
Example Actions by Nodes
• Adaptive Sampling
• Censoring
Source Coding
In‐Network Processing
• Data Aggregation
Source Coding
10/1/2008
21
Outline
• Sensor network contraints
• Digital modulation
• Source coding and Channel coding
• MAC
• Synchronization
• Synthesis: Energy and bandwidth requirements
Channel Coding (FEC)
• Motivation
• Block codes
• Convolution codes
Channel Coding
10/1/2008
22
Channel Coding Approachesfor Sensor Networks
• Coding constraints
• Block coding
Channel Coding
Example: Systematic Block Code
Channel Coding
10/1/2008
23
Alternative: Error Detection
• Motivation
• CRC
Channel Coding
Performance
• Benefits
• Costs
Channel CodingFig. Lathi
10/1/2008
24
Outline
• Sensor network contraints
• Digital modulation
• Source coding and Channel coding
• MAC
• Synchronization
• Synthesis: Energy and bandwidth requirements
Sharing Spectrum
MACFig. Frolik (2007)
10/1/2008
25
MAC
• Motivation
• Contention‐based
• Contention‐free
MAC
ALOHA (ultimate in contention)
• Method
• Advantages
• Disadvantages
MAC
10/1/2008
26
CSMA (contentious but polite)
• Method
• Advantages
• Disadvantages
MAC
Throughput comparison
MAC
10/1/2008
27
Contention Free Approaches
• RTS/CTS
• Reservations
MAC
MAC for Sensor Networks: 802.15.4
• Beacon enabled mode for star networks
MAC
10/1/2008
28
Bandwidth details: 802.15.4
• 2.4 GHz band (2.40‐2.48 GHz)
• Sixteen channels spaced at 5 MHz (CH 11 – 26)
• Data rate – 250 kbps
• Direct sequence spread spectrum (DSSS)
• 4 bits → symbol → 32 chip sequence
• Chip rate of 2 Mcps
• Modulation – O‐QPSK
• Total bandwidth requirement: ~3 MHzMAC
DSSS
• Motivation
• Operation
MAC
10/1/2008
29
Outline
• Sensor network contraints
• Digital modulation
• Source coding and Channel coding
• MAC
• Synchronization
• Synthesis: Energy and bandwidth requirements
Synchronization
• Motivation
• Categories
Synchronization
10/1/2008
30
Node Scheduling
• Sleep
• Listening
• Transmitting
Synchronization
Sleep Scheduling for Sensor Networks: S‐MAC
Synchronization
10/1/2008
31
Synchronizing for Effective Communications
• Carrier
• Bit/Symbol
• Frame
Synchronization
Outline
• Sensor network contraints
• Digital modulation
• Source coding and Channel coding
• MAC
• Synchronization
• Synthesis: Energy and bandwidth requirements
10/1/2008
32
Putting the Pieces Together
Synthesis: Energy and Bandwidth
• M‐ary Signaling
• Channel Coding
Energy & Bandwidth
10/1/2008
33
Sensor Network Example 1: Single vs. Multihop
• Multihop
• Single hop
Energy & Bandwidth
Sensor Network Example 2: Polling vs. Pushing
• Polling
• Pushing
Energy & Bandwidth
10/1/2008
34
Conclusions
• A digital communications approach to WSN h d t i b t dhas advantages in robustness, energy, and bandwidth performance
• Source coding reduces overall system level energy requirements
• Simple channel coding schemes improve dataSimple channel coding schemes improve data reliability minimizing the need for retransmissions
muse
Conclusions ‐ 2
• MAC and routing strategies should be chosen ith t d t k hit twith an eye towards network architecture –
cross‐layer design
• Node synchronization must occur regularly due to clock drift between nodes
• Simple digital communication techniquesSimple digital communication techniques enable low‐energy, low‐bandwidth WSN system requirements
muse
10/1/2008
35
What to know more?
• B. Lathi, Modern Analog and Digital C i ti S t 3rd d O f d 1998Communication Systems, 3rd ed., Oxford, 1998.
• B. Krishnamachari, Networking Wireless Sensors, Cambridge Press, 2005.
• J. Frolik, “Implementation Handheld, RF Test Equipment in the Classroom and the Field ”Equipment in the Classroom and the Field,IEEE Trans. Education, Vol. 50, No. 3, August 2007.