Location Systems. Petteri Nurmi

Samankaltaiset tiedostot
Location Systems Petteri Nurmi

Location Systems. Petteri Nurmi

Indoor Localization II Location Systems. Petteri Nurmi Autumn 2015

Capacity Utilization

Efficiency change over time

1. SIT. The handler and dog stop with the dog sitting at heel. When the dog is sitting, the handler cues the dog to heel forward.

LYTH-CONS CONSISTENCY TRANSMITTER

The CCR Model and Production Correspondence

Positioning Algorithms. Petteri Nurmi

AKKREDITOITU TESTAUSLABORATORIO ACCREDITED TESTING LABORATORY VERKOTAN OY VERKOTAN LTD.

16. Allocation Models

7.4 Variability management

Huom. tämä kulma on yhtä suuri kuin ohjauskulman muutos. lasketaan ajoneuvon keskipisteen ympyräkaaren jänteen pituus

Indoor Localization I Introduction and Positioning Algorithms Petteri Nurmi

Returns to Scale II. S ysteemianalyysin. Laboratorio. Esitelmä 8 Timo Salminen. Teknillinen korkeakoulu

AKKREDITOITU TESTAUSLABORATORIO ACCREDITED TESTING LABORATORY GRANT4COM OY

National Building Code of Finland, Part D1, Building Water Supply and Sewerage Systems, Regulations and guidelines 2007

AKKREDITOITU TESTAUSLABORATORIO ACCREDITED TESTING LABORATORY

Positioning Algorithms. Petteri Nurmi

Network to Get Work. Tehtäviä opiskelijoille Assignments for students.

AKKREDITOITU TESTAUSLABORATORIO ACCREDITED TESTING LABORATORY

On instrument costs in decentralized macroeconomic decision making (Helsingin Kauppakorkeakoulun julkaisuja ; D-31)

Characterization of clay using x-ray and neutron scattering at the University of Helsinki and ILL

Alternative DEA Models

E80. Data Uncertainty, Data Fitting, Error Propagation. Jan. 23, 2014 Jon Roberts. Experimental Engineering

C++11 seminaari, kevät Johannes Koskinen

Mobile Sensing V Motion Analysis. Spring 2015 Petteri Nurmi

Other approaches to restrict multipliers

HITSAUKSEN TUOTTAVUUSRATKAISUT

The Viking Battle - Part Version: Finnish

Uusi Ajatus Löytyy Luonnosta 4 (käsikirja) (Finnish Edition)

S Sähkön jakelu ja markkinat S Electricity Distribution and Markets

Metsälamminkankaan tuulivoimapuiston osayleiskaava

Kvanttilaskenta - 1. tehtävät

Gap-filling methods for CH 4 data

Bounds on non-surjective cellular automata

Results on the new polydrug use questions in the Finnish TDI data

Vaisala s New Global L ightning Lightning Dataset GLD360

Information on preparing Presentation

Travel Getting Around

On instrument costs in decentralized macroeconomic decision making (Helsingin Kauppakorkeakoulun julkaisuja ; D-31)

( ( OX2 Perkkiö. Rakennuskanta. Varjostus. 9 x N131 x HH145

Särmäystyökalut kuvasto Press brake tools catalogue

Tynnyrivaara, OX2 Tuulivoimahanke. ( Layout 9 x N131 x HH145. Rakennukset Asuinrakennus Lomarakennus 9 x N131 x HH145 Varjostus 1 h/a 8 h/a 20 h/a

KMTK lentoestetyöpaja - Osa 2


Kvanttilaskenta - 2. tehtävät

TM ETRS-TM35FIN-ETRS89 WTG

Valuation of Asian Quanto- Basket Options

Lämmitysjärjestelmät

AKKREDITOITU TESTAUSLABORATORIO ACCREDITED TESTING LABORATORY

ReFuel 70 % Emission Reduction Using Renewable High Cetane Number Paraffinic Diesel Fuel. Kalle Lehto, Aalto-yliopisto 5.5.

WindPRO version joulu 2012 Printed/Page :47 / 1. SHADOW - Main Result

TM ETRS-TM35FIN-ETRS89 WTG

TM ETRS-TM35FIN-ETRS89 WTG

AKKREDITOITU TESTAUSLABORATORIO ACCREDITED TESTING LABORATORY

TM ETRS-TM35FIN-ETRS89 WTG

TM ETRS-TM35FIN-ETRS89 WTG

TM ETRS-TM35FIN-ETRS89 WTG

TM ETRS-TM35FIN-ETRS89 WTG

AKKREDITOITU TESTAUSLABORATORIO ACCREDITED TESTING LABORATORY

On instrument costs in decentralized macroeconomic decision making (Helsingin Kauppakorkeakoulun julkaisuja ; D-31)

TM ETRS-TM35FIN-ETRS89 WTG

Alternatives to the DFT

( ,5 1 1,5 2 km

WindPRO version joulu 2012 Printed/Page :42 / 1. SHADOW - Main Result

TM ETRS-TM35FIN-ETRS89 WTG

Microsoft Lync 2010 Attendee

TIEKE Verkottaja Service Tools for electronic data interchange utilizers. Heikki Laaksamo

Review Petteri Nurmi

T Statistical Natural Language Processing Answers 6 Collocations Version 1.0

TM ETRS-TM35FIN-ETRS89 WTG

WAMS 2010,Ylivieska Monitoring service of energy efficiency in housing Jan Nyman,

I. Principles of Pointer Year Analysis

,0 Yes ,0 120, ,8

GNSS-vastaanottimet. Havaintosuureet

Tietorakenteet ja algoritmit

Innovative and responsible public procurement Urban Agenda kumppanuusryhmä. public-procurement

FinFamily PostgreSQL installation ( ) FinFamily PostgreSQL

Exercise 1. (session: )

TM ETRS-TM35FIN-ETRS89 WTG

Mat Seminar on Optimization. Data Envelopment Analysis. Economies of Scope S ysteemianalyysin. Laboratorio. Teknillinen korkeakoulu

7. Product-line architectures

AYYE 9/ HOUSING POLICY

Infrastruktuurin asemoituminen kansalliseen ja kansainväliseen kenttään Outi Ala-Honkola Tiedeasiantuntija

FPGA-piirien käyttökohteet nyt ja tulevaisuudessa Tomi Norolampi

Copyright 2008 Pearson Education, Inc., publishing as Pearson Addison-Wesley.

Tarua vai totta: sähkön vähittäismarkkina ei toimi? Satu Viljainen Professori, sähkömarkkinat

Toppila/Kivistö Vastaa kaikkin neljään tehtävään, jotka kukin arvostellaan asteikolla 0-6 pistettä.

AFCEA PVTO2010 Taistelija / S4

anna minun kertoa let me tell you

Indoor Positioning: Technologies and Use Cases in Retail Context

Kysymys 5 Compared to the workload, the number of credits awarded was (1 credits equals 27 working hours): (4)

S SÄHKÖTEKNIIKKA JA ELEKTRONIIKKA

Sähköjärjestelmän käyttövarmuus & teknologia Käyttövarmuuspäivä

Group 2 - Dentego PTH Korvake. Peer Testing Report

Land-Use Model for the Helsinki Metropolitan Area

Salasanan vaihto uuteen / How to change password

Choose Finland-Helsinki Valitse Finland-Helsinki

Ensimmäinen välikoe. Kurssin voi suorittaa tentillä tai kahdella välikokeella

Transkriptio:

Location Systems Petteri Nurmi 19.11.2016 1

Questions Which dimensions can be used to characterize location systems? Which criteria can be used to evaluate location systems? What is proximity sensing and what kind of sensor technologies can be used with it? What is a pseudorange? What are Keplerian elements and what is an ephemeris? Which error sources affect GPS accuracy? How GSM positioning works? What is dead reckoning? 19.11.2016 2

Location Systems System that provides measurements that can be used to determine the position of an entity Possible dimensions for characterizing location systems: Location representation Scale Location system type Measurement Error sources 19.11.2016 3

Location Systems - Representation and Scale Location representation Absolute: coordinates specified according to some reference system Relative: distance and orientation from a specific reference point Symbolic: semantic name, e.g., name of a room or street address Scale Worldwide: works everywhere in the world Local: works in a specific area / region Outdoors or indoors: works only indoors or outdoors 19.11.2016 4

Location Systems - Location System Type Active or passive Determines whether client transmits anything or not Client-based Client is responsible for estimating its own position without assistance from network infrastructure Network-based Network infrastructure is responsible for estimating the location of the client Network-assisted Client and network infrastructure both participate in determining the location 19.11.2016 5

Location Systems - Location System Type Active or passive Determines whether client transmits anything or not Client-based Client is responsible for estimating its own position without assistance from network infrastructure Network-based Network infrastructure is responsible for estimating the location of the client Network-assisted Client and network infrastructure both participate in determining the location 19.11.2016 6

Location Systems - Measurement Refers to the signal and type of measurement that is used for position estimation Types of signal: radio, infrared, ultrasound Measurement types ID: Identifier of a radio beacon RSS: Received signal strength of a radio beacon Time: Time of flight or time difference or arrival Angle: Orientation between transmitter and receiver 19.11.2016 7

Location Systems - Evaluation Criteria Accuracy: difference between estimate location and true location Kilometer, meter or centimeter level accuracy Consistency: how often a reported position accuracy can be obtained Typically reported in terms of percentiles For example, 95-percentile of 12cm means that system error is within 12cm 95% of times Latency Time period between a position request and delivery of position fix 19.11.2016 8

Accuracy and Consistency Consistency often represented using a table that contains accuracy at selected percentiles The i th percentile is the i th value when performance measures are sorted Common percentiles: median (50), and highorder percentiles (95, 99, 100) 100-percentile = worst case error Alternative is to use CDF plots Percentiles on the x-axis, error on the y-axis 19.11.2016 9

Accuracy and Consistency Example Ground truth Estimates Error (meters) 358.0 391.5 516.5 1398.5 268.0 535.5 358.0 822.5 516.5 235.0 569.5 966.5 569.5 391.5 358.0 1110.5 358.0 1110.5 93.0 1254.0 516.5 235.0 516.5 1254.5 310.5 535.5 516.5 822.5 569.5 235.0 410.5 966.5 569.5 391.5 226.0 1110.5 410.5 1110.5 226.0 1254.5 5.11 3.30 0.97 3.64 1.22 3.65 0.00 3.03 1.20 3.05 50-percentile: 3.04 95-percentile: 5.11 19.11.2016 10

Location Systems - Evaluation Criteria Overhead Signaling overhead: additional burden on communication infrastructure Computation overhead: burden on client/infrastructure resulting from necessary calculations Power overhead: how much power is required from the terminal Roll-out and operating cost Roll-out costs refer to expenditure required to make the system operational Operating costs refer to expenditure required to maintain the system operational 19.11.2016 11

Location Systems Error Sources As discussed during previous lecture, measurements provided by location systems are inaccurate Noise in measurements causes errors in distance / angle values Geometry of reference points amplifies effects of noise Errors in coordinates of reference points can cause further errors Dilution of Precision (DoP) characterizes how range and geometry affect position estimates 19.11.2016 12

Error Sources Geometry Reference point geometry has a strong influence on positioning accuracy When reference points orthogonal, the size of the region of uncertainty smallest As reference points move closer to each other, size of the region of uncertainty increases 19.11.2016 13

Error Sources Clock Errors Clocks have two main components Oscillator: creates a consistent frequency Counter: translates frequency cycle s of oscillator into a common time unit Oscillators inaccurate and susceptible to variations (i.e., instable) è need for clock synchronization Time offset: difference in time between two clocks Clock drift: error in absolute time over a period of time Clock models used to reduce effect of clock errors 19.11.2016 14

Error Sources Clock Errors: Example Signals propagate (approximately) with the speed of light c = 299792458 m/s è small errors in time can cause HUGE errors Example: Delay of 1ms causes a distance error of 299792 meters Normal clock errors in the magnitude of 10-6 or smaller For long distances and accurate position estimation, clock accuracy essential è GPS uses high precision atomic clocks Errors in the magnitude of 10-16 19.11.2016 15

Error Sources Atmospheric Effects Source: http://en.wikipedia.org/ wiki/refractive_index Source: http://en.wikipedia.org /wiki/ionosphere Propagation follows speed of light only in a vacuum Refractive index: speed of light in a particular medium Refraction: change in direction of a wave, occurs when a wave enters from one medium to another Ionosphere Electronically charged outer layer of the Earth s atmosphere, ionic intensity varies according to time of day and solar radiation Refraction in ionosphere can cause errors of tens of meters Troposphere: lowest part of the atmosphere Refraction in troposphere can cause errors in distance estimates that are up to 250cm in magnitude Satellite positioning systems use models of atmospheric refraction to reduce the influence of atmospheric effects 19.11.2016 16

Error Sources Multipath Propagation Source: http://en.wikipedia.org/ wiki/multipath_propaga tion Propagation phenomenon caused by signal reaching receiver along various paths Reflection: when wavelength of signal much shorter than size of obstacles, signals reflect Diffraction: bending of a signal as it hits on irregularities of an obstacle Scattering: signal multiplied at obstacles whose size is less or the same magnitude as the signal s wavelength Reflection Diffraction Scattering 19.11.2016 17

Location Systems Proximity Sensing First indoor positioning systems were based on proximity, i.e., closeness of a reference point Essentially a variation of trilateration where only a single reference point can be observed Uncertainty determined by the range of the transmitters The higher the positioning accuracy requirement, the higher the cost for infrastructure Limitations in terms of Range Latency Receiving angle 19.11.2016 18

Location Systems ActiveBadge Source: http://koo.corpus.cam.ac.uk/projects/badge s/index.html One of the first indoor location tracking systems Introduced by Olivetti Research Laboratory (Cambridge, UK) in 1992 Badge transmits IR signal every 10-15 seconds Receivers that detected the signal store information about badge location in a central database Location of receivers known, position of client can be estimated from badge observations Designed to operate up to 6m away from a sensor However, infrared generally weak and are only slightly reflected and scattered è line-of-sight usually needed 19.11.2016 19

Location Systems Wireless Indoor Positioning System Variant of ActiveBadge, position estimated on the terminal instead of the network infrastructure Transmitters installed in the infrastructure When badge receives signal from transmitter, it sends information using WLAN to a central server 19.11.2016 20

Proximity Sensing Other Techniques Radio Frequency Identification (RFID) Radio signals that are exchanged between a reader and tags (or transponders) Active tags: use own power supply Passive tags: draw energy from signals emitted by reader Essentially a proximity sensing technique Bluetooth Wireless technology for transmitting data over short distances Proximity sensing or trilateration can be used for position estimation Main issue latency due to Bluetooth scanning and difficulty in obtaining suitable transmitters 19.11.2016 21

Proximity Sensing Beacon Scan Request Instead of requiring dedicated badges, recently several systems that identify mobile devices Overall principle same as in Active Badge Devices detected from beacon broadcast messages, which are performed while scanning for devices Position estimated as the location within infrastructure where scan observed ibeacon: Bluetooth Low Energy (BLE) WalkBase: WiFi 19.11.2016 22

ibeacon http://www.youtube.com/v/lwnmx-5mblm 19.11.2016 23

Walkbase http://www.youtube.com/v/z57awrko9ec 19.11.2016 24

Estimating Distances Recall (from previous lecture) that trilateration can be used to estimate position with distance measurements Two reference points required Four main methods (3 time, 1 signal-based): Time of arrival: travel time from transmitter to receiver requires synchronization between them Time difference of arrival: difference in arrival times between multiple receivers Requires synchronization between receivers Round-trip time: time from transmitter to receiver and back (or vice versa) Signal strength: Propagation modeling based on signal decay 19.11.2016 25

Source: http://www.cl.cam.ac.uk/r esearch/dtg/research/wiki /BatSystem/BatPhotos Location Systems Ultrasound I: ActiveBat Ultrasonic location tracking system Receivers dispersed in environment Tags emit short ultrasound pulses Position estimated using trilateration Receivers use radio frequency to cue for badges When badge receivers cue, it emits an ultrasound pulse Receiver uses time difference between sending the cue and receiving the ultrasound pulse to estimate distance Acoustic signals travel significantly slower than radio signals è delay from the polling signal negligible 19.11.2016 26

Location Systems Ultrasound II: Cricket Source: http://nms.lcs.mit.edu/pr ojects/cricket/ Beacons affixed to the environment emit RF message together with ultrasound pulse Ultrasound pulse used to estimate distance from beacons RF messages contain information about beacon positions Trilateration can be used to estimate the position of the client given a set of beacon observations Ultrasound does not penetrate obstacles è lineof-sight required Trilateration è at least three beacons need to be observed 19.11.2016 27

Location Systems Ultrasound III: Relate Source: http://eis.comp.lancs.ac.uk/hardware/ Source: http://eis.comp.lancs. ac.uk/software/ Peer-to-peer ultrasound positioning sensing system that provides relative position information Estimates distance and angle between devices Non-linear regression used to estimate relative 2D positions of objects Ad hoc radio communications used to exchange information between objects Provides also time synchronization Enables refining estimates of objects 19.11.2016 28

Location Systems: Ultra-wide band Low-energy, high-bandwidth, short-range Wide-band = uses large portion of radio spectrum (> 500 MHz) In positioning, between 3.1 and 10.6 GHz (regulated) Very high accuracy, but short range implies high installation costs Positioning can be based either on distance (trilateration) or angular direction (triangulation) Distance: time of arrival (TOA), time difference of arrival (TDOA), round trip time (RRT), propagation modeling 19.11.2016 29

Ubisense (Ultrawide Band UWB) http://www.youtube.com/v/diitfrx3hxo 19.11.2016 30

Satellite Positioning Trilateration-based positioning approach Reference points are satellites on a specific orbit Distances from satellites measured using one way time-offlight measurements Position of a satellite specified by its orbit Regular and repeating path around the Earth Orbit plane: orbit position with respect to equatorial plane Orbit shape: circle or ellipse specifying how the satellites moves around the earth Orbit altitude: height from surface of Earth, determines period needed to circulate Earth Orbit specified by six parameters known as Keplerian elements 19.11.2016 31

Satellite Positioning Keplerian Elements Source: http://en.wikipedia.org /wiki/keplerian_elem ents Shape of orbit specified by Eccentricity: the flattening of a circle needed to transform it into an ellipse Semi-major axis: diameter/size of the ellipse See Lecture I Orientation of orbital plane specified by: Inclination: tilt of the reference plane Right ascension of ascending node: point where the satellite crosses equatorial plane Position of the satellite specified by: Argument of perigree: orientation in which an ellipse is flattened compared to a circle True anomaly: angle between perigree (2 in image) and satellite s position 19.11.2016 32

Global Positioning System (GPS) Some History Satellite navigation system Initiated in the 1970s Fully operational in 1995 Initially designed for the needs of tactical bombers: Accurate 3D position worldwide Passive receivers (to not reveal position to enemy) 19.11.2016 33

Global Positioning System (GPS) Position Estimation Positioning based on trilateration 24+ satellites orbiting the earth (currently 31) Satellites broadcast messages that contain Orbital position of the satellite System time of the satellite Receivers on earth listen for the broadcasts and estimate their distance from the satellites Messages weak è GPS does not work indoors Position can be solved from (at least three) range equations: 19.11.2016 34

Global Positioning System (GPS) Position Estimation Range estimates valid only if satellite and clock receivers synchronized In practice not the case Solution: consider receiver clock offset as an unknown variable Estimates called pseudoranges (biased range) Include a term for other error sources 19.11.2016 35

Global Positioning System (GPS) Error Sources With pseudoranges, position estimation requires at least 4 satellites GPS errors consist of two components Pseudorange error Less than 10 meters with modern receivers Dilution of precision GDOP: Geometric Dilution of Precision; overall accuracy PDOP: Positional Dilution of Precision; 3D position HDOP: Horizontal Dilution of Precision; latitude and longitude VDOP: Vertical Dilution of Precision; altitude TDOP: Temporal Dilution of Precision; clock offset 19.11.2016 36

Global Positioning System (GPS) How to calculate DOP values? Differentiate the pseudorange equations with respect to the unknown variables Define Q = (A T A) -1 GDOP = trace(q) TDOP = (Q 44 ) PDOP = (Q 11 + Q 22 + Q 33 ) HDOP = (Q 11 + Q 22 ) VDOP = (Q 33 ) 19.11.2016 37

Global Positioning System (GPS) It can be shown that the differentiated pseudorange matrix is given by Where 19.11.2016 38

Global Positioning System (GPS) Consider the following measurements: Substituting the values gives: 19.11.2016 39

Global Positioning System (GPS) The DoP matrix is then given by: From which we can calculate the DoP values: GDOP = trace(q) = 17.6 TDOP = (Q 44 ) = 9.3 PDOP = (Q 11 + Q 22 + Q 33 ) = 14.9 HDOP = (Q 11 + Q 22 ) = 3.7 VDOP = (Q 33 ) = 14.5 19.11.2016 40

Global Positioning System - Dual Frequency Ranging GPS navigation messages transmitted along two carrier frequencies L1 (154 f 0 ) and L2 (120 f 0 ) f 0 equals the fundamental frequency of the satellite s clocks (10.23 MHz) Dual-frequency ranging considers messages transmitted on both carrier frequencies Allows differentiating for estimating effect of ionospheric refraction è improvements in accuracy (UERE) Access restricted to military use 19.11.2016 41

Global Positioning System Differential GPS (D-GPS) Navigation messages from satellites observed at a reference station located at a known position Orbital information can be used to estimate the true distance between the reference station and satellite This can be compared against estimated distance given by the navigation messages Difference in distances specifies effect of atmospheric, ephemeric and clock errors Reference stations transmit correction parameters that receivers can use for adjusting range estimates 19.11.2016 42

Global Positioning System: Signal (Re-)Acquisition Two types of satellite information: Ephemeris: precise orbital position Almanac: coarse-grained information that is used to acquire satellite fix Cold start: no information about satellite locations ( 1 minute) Warm start: almanac and last position If position changed, can cause outliers Hot start: ephemeris data available 19.11.2016 43

GSM Global System for Mobile Communications Worldwide cellular telephone standard First deployed in 1992 Network divided to base stations and cells Each cell linked with one base station Each base station serves multiple cells Cells grouped into clusters, clusters have unique LAI (location area identifier) Client knows cell ID, Network knows LAI 19.11.2016 44

GSM Positioning Positioning can be performed either on the client side (handset-based) or on the network-side Handset-based techniques typically more accurate, but require installing software on the client Fingerprinting or beacon-based positioning Network-based techniques can be used to position all handsets without modifications to them Cell identifier positioning, timing advance, TDOA, Network-assisted: estimation on the client, but network infrastructure used to help Assisted GPS 19.11.2016 45

GSM Positioning Components Location Measurement Unit (LMU) Responsible for performing timing (or angle) measurements from a client Essential for infrastructure-based positioning Service Mobile Location Center (SMLC) Controls positioning process, including LMUs, and calculates position fixes Cell Broadcast Center (CBC) Broadcasts positioning assistance data to terminals, used in E-OTD positioning 19.11.2016 46

GSM Cell Identifier Positioning Simplest technique is to use proximity sensing Client knows cell ID è coordinates of the corresponding GSM tower can be used as estimate When cell tower locations not known, they can be estimated from empirical measurements E.g., average of the locations where the RSS is strongest Problem GSM cell size varies and can be up to 30km Around 150 meters in dense urban areas, 500-800 meters in other urban areas and up to several kilometers in rural areas Accuracy depends on GSM cell density 19.11.2016 47

GSM Timing Advance (TA) TA characterizes the time that is needed for a signal to reach the base station from a mobile client Network parameter that controls adjustments needed to prevent collisions between users Divided into 64 steps, each step representing an advance of one bit period, or approximately 553.5m TA value together with cell identifier can be used to estimate locations TA value determines a circular region within which the client is positioned 19.11.2016 48

GSM E-OTD Enhanced Observed Time Difference Handset measures arrival times or differences in arrival times between base station pairs Trilateration or multilateration can then be used to estimate position of the client Time synchronization needed for estimation GSM standard does not cover this è separate location measurement units (LMUs) used to measure time offsets and achieve a posteriori synchronization Details out of scope for the course Also defined for UMTS 19.11.2016 49

GSM U-TDoA Uplink Time Difference of Arrival Network infrastructure based multilateration approach Base station and additional LMUs measure differences in the time when a handset transmission is received At least three LMUs needed to estimate the position Network can only measure transmission time when mobile device is in busy state Cell handover or active transmission (call, SMS, data) Network generates a pseudo-handover signal which triggers the client into busy mode 19.11.2016 50

Assisted GPS Solution that facilitates obtaining a satellite fix Designed for mobile phones that have integrated GPS receiver Assisting server collects satellite data and provides information about satellites to client Client can find satellites more rapidly Network can also provide correction data to improve position estimation accuracy 19.11.2016 51

Estimating Angular Directions Recall (from previous lecture) that triangulation can be used to estimate position with angular measurements Two reference points required Two main methods for estimating angles of arrival Beamforming: fixed number of beams used to scan plane, angle corresponds to beam with highest power Multiarray systems: multiple receivers organized into an array and used to estimate angle Two main signal sources: Acoustic: ultrasound, speech, or other audible sounds Radio signal: delays in observed signal patterns 19.11.2016 52

Estimating Angular Directions: Multiarray Systems P 0 a T x P 1 Based on variations of time difference of arrival Basic equation: TDOA = cos(a) (P 1 P 0 ) / c a is the angle the source makes to the plane connecting two receivers P 1 and P 0 are the locations of two receivers c is propagation speed Relative time delay can be estimated from (discrete) Fourier transform of the delayed signal Classical methods: least squares, Bartlett, Subspace methods: Multiple Signal Classification (MUSIC) Performance depends on SNR (signal-to-noise ratio) Highly varying computational complexity between different approaches 19.11.2016 53

Example: Quuppa HAIP (High Accuracy Indoor Positioning) Triangulation based on single reference point (2D) Ceiling of indoor space mounted with locators Angular direction of tags estimated within the locator If the location and height of the locator, and an approximate height of the tag are known, 2D position can be estimated 3D position requires two observable locators Submeter accuracy: typically 0.5 1m Depends on density and configuration of locators Works best in open ceiling areas 19.11.2016 54

Quuppa http://www.youtube.com/v/o1fffioyixk 19.11.2016 55

Dead (or deduced) Reckoning Source: http://en.wikipedia. org/wiki/dead_rec koning Estimate position using extrapolation from last known position Also referred to as inertial navigation Requires information about Direction of motion E.g., compass or gyroscope Velocity of motion or distance travelled since last known position E.g., accelerometers or odometers Errors in motion measurements amplify over time è accuracy drifts and decreases over time 19.11.2016 56

Dead Reckoning Let (x 0,y 0 ) denote the current position of the target New position (x 1,y 1 ) given by x 1 = x 0 + L cos α, y 1 = y 0 + L sin α Where L is the distance traveled If distance is not known, but velocity is, we have L = vδt y 1 y 0 x 0 x 1 19.11.2016 57

Pedestrian Dead Reckoning (PDR) PDR special case of dead reckoning targeted at pedestrians Displacement (i.e., elapsed distance) estimated by analysing walking patterns of the user Distance: d = i=1 n l i n = number of steps l i = length of step i Accordingly, requires mechanisms for (i) counting steps and (ii) estimating their length Direction of motion estimated through multisensor fusion Accelerometer, gyroscope, and magnetomer Discussed in detail during the (separate) course Mobile Sensing (see: https://www.cs.helsinki.fi/en/courses/582730/2015/k/k/1) Widely studied in specialized domains using wearable sensors at fixed orientations 19.11.2016 58

Other Location Systems Source: http://ubicomplab.cs. washington.edu/wiki/ PLP FM Radio FM signals are radio waves è propagation models or fingerprinting can be used to provide location Pressure floors Load sensors embedded within floor tiles PowerLine positioning Specialized modules emit signals along powerlines which tags carried by users sense Airbus positioning Determines room transitions and gross movement by sensing differential air pressure in a home Camera tracking 19.11.2016 59

Summary Location systems can be characterized according to different dimensions Location representation, scale, type, measurement type, error sources Location systems can be evaluated according to various criteria Accuracy and consistency Latency and overhead Roll-out and operation cost 19.11.2016 60

Summary Numerous different error sources possible Clock errors Geometry errors Noise in measurements Errors in reference point information Atmospheric effects Multipath effects (reflection, diffraction and scattering) Proximity sensing: infrared, RFID, Bluetooth Ultrasound Speed of sound slower than speed of radio waves (light) è RF based synchronization for pulses 19.11.2016 61

Summary GPS Satellite positioning system providing 3D position and time worldwide GPS error consists of errors in pseudorange estimates and geometry errors (DOP) GSM positioning Cell Identifier positioning and Timing Advance E-OTD and U-TDoA Assisted GPS Dead reckoning Position extrapolation from previous estimates 19.11.2016 62

Literature Küpper, A., Location-Based Services : Fundamentals and Operation, Wiley, 2005 Varshavsky, A. & Patel, S., Krumm, J. (Ed.), Location in Ubiquitous Computing Ubiquitous Computing Fundamentals, Chapman and Hall/CRC, 2010, 285-320 Langley, R. B., Dilution of Precision, GPS World, 1999, 10, 52-59 Want, R., The Active Badge Location System, ACM Transactions on Information Systems, 1992, 91-102 Ward, A.; Jones, A. & Hopper, A., A new location technique for the active office IEEE Personal Communications, 1997, 4, 42-47 Priyantha, N. B.; Chakraborty, A. & Balakrishnan, H., The Cricket location-support system, Proceedings of the International Conference on Mobile Computing and Networking (Mobicom), 2000, 32-43 Hazas, M.; Kray, C.; Gellersen, H.; Agbota, H.; Kortuem, G. & Krohn, A., A relative positioning system for co-located mobile device, Proceedings of the 3rd international conference on Mobile systems, applications, and services (MobiSys), ACM, 2005, 177-190 19.11.2016 63

Literature GPS and GSM Dixon, T. H., An Introduction to the Global Positioning System and Some Geological Applications, Reviews of Geophysics, 1991, 29, 249-276 Enge, P. K., The Global Positioning System: Signals, Measurements, and Performance, International Journal of Wireless Information Networks, 1994, 1, 83-105 Enge, P. & Misra, P., Special Issue on Global Positioning System, Proceedings of the IEEE, 1999, 87, 3-15 Getting, I. A., Perspective/Navigation - The Global Positioning System, IEEE Spectrum, 1993, 30, 43-47 McNeff, J. G., The Global Positioning System, IEEE Transactions on Microwave Theory and Techniques, 2002, 50, 645-652 Drane, C.; Macnaughtan, M. & Scott, C., Positioning GSM telephones, Communications Magazine, IEEE, 1998, 36, 46-54, 59 Silventoinen, M. I. & Rantalainen, T., Mobile station emergency locating in GSM Proceedings of the IEEE International Conference on Personal Wireless Communications, IEEE, 1996, 232-238 19.11.2016 64