Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab -...
Transcript of Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab -...
![Page 1: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/1.jpg)
Wireless Systems Lab - 2014
Lesson 317 November 2014
![Page 2: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/2.jpg)
Wireless Systems Lab - 2014
Indoor Localization
![Page 3: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/3.jpg)
Wireless Systems Lab - 2014
WIFI Fingerprinting
![Page 4: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/4.jpg)
Wireless Systems Lab - 2014
Signal instability
![Page 5: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/5.jpg)
Wireless Systems Lab - 2014
Granularity
![Page 6: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/6.jpg)
Wireless Systems Lab - 2014
Number of samples
![Page 7: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/7.jpg)
Wireless Systems Lab - 2014
User orientation
![Page 8: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/8.jpg)
Wireless Systems Lab - 2014
Algorithms
● Training■ Fingerprinting algorithm
● Matching■ Range based algorithm■ Euclidean distance based algorithm
![Page 9: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/9.jpg)
Wireless Systems Lab - 2014
Fingerprinting algorithm
![Page 10: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/10.jpg)
Wireless Systems Lab - 2014
Determine Group Leader
![Page 11: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/11.jpg)
Wireless Systems Lab - 2014
Fingerprinting algorithm
![Page 12: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/12.jpg)
Wireless Systems Lab - 2014
Fingerprinting algorithm
![Page 13: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/13.jpg)
Wireless Systems Lab - 2014
Calculate filtered range
![Page 14: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/14.jpg)
Wireless Systems Lab - 2014
Calculate filtered range
![Page 15: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/15.jpg)
Wireless Systems Lab - 2014
Fingerprinting algorithm
![Page 16: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/16.jpg)
Wireless Systems Lab - 2014
Matching algorithm
Two algorithms• Range based algorithm• Euclidean distance based algorithm
![Page 17: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/17.jpg)
Wireless Systems Lab - 2014
Range based
Given the observed signal vector:MAC: A7C8, Signal: 37
Room 1 is more likely to be the correct location
![Page 18: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/18.jpg)
Wireless Systems Lab - 2014
Euclidean distance based
![Page 19: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/19.jpg)
Wireless Systems Lab - 2014
Euclidean distance based
Given the observed signal vector:MAC: A7C8, Signal: 37
Room 1 with reference point 2 is more likely to be the correct location, because it has the minimal Euclidean distance from the observed signal vector.
![Page 20: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/20.jpg)
Wireless Systems Lab - 2014
Results
The system is able to detect user's location correctly within 3-4 meters accuracy
• 70% of the time, using Euclidean based• 60% of the time, using range based
![Page 21: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/21.jpg)
Wireless Systems Lab - 2014
Basic notions of trilateration
• For e.g., GPS use trilateration
• Two steps are involved:
• The distance (range)
to anchors is calculated
• The location of the
nodes is estimated
based on the distance
Lateration is the most common method for deriving the location of a wireless device.
![Page 22: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/22.jpg)
Wireless Systems Lab - 2014
Main algorithms for trilateration positioning
• The most used approaches for trilateration positioning:• Non-Linear Least Square (NLS)
• Bayes Filters as Extended Kalman Filters (EKF) methods.
• There are three unknowns:• Coordinate of the receiver
• NLS is largely used to calculate these unknowns
• EKF for tracking and sensor fusion
![Page 23: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/23.jpg)
Wireless Systems Lab - 2014
Positioning: the problem of data fitting
The problem of positioning is one of data fitting. • Given a set of input measurements {ρi },
what parameters {xˆ, yˆ, zˆ } provide a best fit?
• Regression analysis estimates the parameters.• The best fit must provide a solution which is close (in
some sense) to the true location p = {x, y, z }
![Page 24: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/24.jpg)
Wireless Systems Lab - 2014
Non-linear regression
• Current position estimate p = {xˆ, yˆ, zˆ }• Associated static source s_i = {x_i , y_i , z_i}
ρ_i = || s_i - p || + v = √ { (xˆ-x_i)^2 + (yˆ-y_i)^2 + (zˆ-z_i)^2 +v }
• This equation is a non-linear function of the parameters • Non Linear regression
• No closed-form expression for the best-fitting parameters• Numerical optimization
![Page 25: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/25.jpg)
Wireless Systems Lab - 2014
When least-square (LS) works
It gives a position that agrees with most of the data.
• Least-square methods are appropriate when:• The noise is Gaussian with equal variance ( v ).• The input data are reliable.• Only one or two datums are multipathed.
• Covered by physical layer protocol
![Page 26: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/26.jpg)
Wireless Systems Lab - 2014
Linearization: LLS algorithm
• The Linear Least Square (LLS) approach linearizes the NLS problem• by introducing a constraint in the formulation
• It obtains a closed form expression of the estimated location.
![Page 27: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/27.jpg)
Wireless Systems Lab - 2014
Example
Start with N > 1 equations to estimate position
and subtracting the constraint :
from both sides of each equation, the above can be rewritten as A p = b.
Solution is : p = (AT A)−1 AT b
![Page 28: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/28.jpg)
Wireless Systems Lab - 2014
Inertial Tracking
• Use electronic accelerometers, compasses. • Sense movement and direction in 2D and 3D• Dead-reckoning → relative positions
• (w.r.t. given fixed point)
• Position error grows with time and distance in the absence of:• Position fixes• Fusion with other sensors (EKF)• Context information (e.g., map)
![Page 29: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/29.jpg)
Wireless Systems Lab - 2014
Good and bad results with DR
• http://roodin.crs4.it/ • Implements dead-reckoning in android using pedometer • Code Libraries are available • Good starting point ( 2 publications + MS thesis reference (MS is in italian) )
![Page 30: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/30.jpg)
Wireless Systems Lab - 2014
Acoustic Localization
• Sound propagation speed is much slower than radio.
• ≈ 343.2 m/s (air) versus ≈ 300 m/μs, 10^6 times slower than light!!
• No need for high clock rates, but :• Clock drift 1 ms error in TOA estimation will translate
to more than 30 centimeters error in the ranging result
• Android system inaccurate to guarantee exact time of received packet and transmitted packet
![Page 31: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/31.jpg)
Wireless Systems Lab - 2014
BeepBeep
BeepBeep: A High Accuracy Acoustic Ranging System using COTS Mobile Devices
![Page 32: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/32.jpg)
Wireless Systems Lab - 2014
Details
• Off the shield implementation (only software)• No time-synchronization• Works directly on received signal,offline
computation• Accuracy on several cm• Easy to be implemented on Android :)
![Page 33: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/33.jpg)
Wireless Systems Lab - 2014
From land to water ...
• Underwater Acoustic Wireless Sensor Networks (UAWSN)
![Page 34: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/34.jpg)
Wireless Systems Lab - 2014
UAWSNs Communication Issues
• Propagation Delay ( 1500 m/s )• Attenuation ( depends on frequency )• High Variablity ( asymmetric links )• Low Data Rate ( 1 kb/s )• Noise
No mathematical UW channel model
•
Protocols designed for terrestrial sensor networks are unsuitable.
![Page 35: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/35.jpg)
Wireless Systems Lab - 2014
http://reti.dsi.uniroma1.it/UWSN_Group/
SUNSET Framework
● MAC & Routing protocol design, simulate and validate at sea
● Path planning algorithms for AUV
● Mathematical modelling of UWSN
● Acoustic transmission & Signal Processing
● Cooperation and communication Modems + Sensors + AUV
● Underwater IoT
![Page 36: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/36.jpg)
Wireless Systems Lab - 2014
Underwater Tic Tac Toe :)
![Page 37: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/37.jpg)
Wireless Systems Lab - 2014
Evolution of M2M
![Page 38: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/38.jpg)
Wireless Systems Lab - 2014
The Web of Things
![Page 39: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/39.jpg)
Wireless Systems Lab - 2014
IoT Standards
![Page 40: Lesson 3 - TWikitwiki.di.uniroma1.it/pub/Wireless/WebHome/Lesson_3a.pdf · Wireless Systems Lab - 2014 Range based Given the observed signal vector: MAC: A7C8, Signal: 37 Room 1 is](https://reader034.fdocuments.in/reader034/viewer/2022042804/5f5223655aebc5552463e0a0/html5/thumbnails/40.jpg)
Wireless Systems Lab - 2014
CoRE Requirements