Wireless 3D Positioning

1
Wireless 3D Positioning

• Outdoor Positioning using
• GPS or
• Operator driven telephone networks.
• Indoor Positioning based on Wireless Local Area
  Network using
• Geometrical modelling or
• Classification.
• Positioning Architecture and Some Research
  Questions.
• 
  John Aa. Sørensen, lektor
• 
  Dept. of Innovation, ITU
  
  
  
  

2
Systems for Outdoor Positioning I

• Outdoor Positioning based on
• Global Positioning System (GPS) Satellites or
• Operator driven Mobile Phone Networks.
• Position estimation is based on triangulation.
• The position is determined based on knowledge of
• distances to at least 3 known positions.
• The distances are estimated from time
  measurements.
• 
  

3
Systems for Outdoor Positioning II

• Outdoor positioning by Global Positioning System
  (GPS)
• Developed by US Department of Defense (DoD).
• Basic architecture was approved in 1973.
• First satellite was launched in 1978.
• System was declared operational in 1995.
• Cost of development is approx. 10 billion.
• Annual operation and maintenance approx. 400
  mill.
• Ref.
• 1 Special Issue on "GPS The Global Positioning
  System".
• Proceedings of the IEEE, January 1999.
• 
  

4
Systems for Outdoor Positioning III

• Outdoor positioning by Global Positioning System
  (GPS)
• Satellites move 4 km/sec. Their positions in
  space are estimated within
• a few meters, based on predictions made 24 hours
  earlier.
• 24 satellites in near circular orbits with radius
  26.560 Km.
• Frequency bands L1 1575.42 MHz, L2 1227.6 MHz
• Precision is approx. 10 m RMS, cf. ref. 1
• Ref.
• 1 Special Issue on "GPS The Global Positioning
  System".
• Proceedings of the IEEE, January 1999.
• 
  

5
Systems for Outdoor Positioning IV

• Using operator driven Mobile Phone Networks.
• GSM (Global System for Mobile Communication)
  system.
• Frequency bands at 900 MHz or 1800 MHz.
• Measure the time differences between base
  stations.
• Use that in the triangularization, knowing the
  positions of the base stations.
• 
  

6
Systems for Indoor Positioning I

• GPS is not sufficient for indoor positioning.
• The indoor environment of a building using WLAN
• Multipath propagation of the electromagnetic
  waves.
• Using a Wireless Local Area Network (WLAN) there
  are two
• fundamentally different approaches, based on
• Geometrical Modelling, or on
• Classification.
• 2 Kaveh Pahlavan et al. "Indoor Geolocation
  Science
• and Technology".
• IEEE Communications Magazine, Feb. 2002.
  
  

7
Systems for Indoor Positioning II

• Classification based approaches using IEEE
  802.11b WLAN.
• (11Mbit/sec). Carrier frequency in the ISM band
  at 2.4 GHz.
• Ekahau Positioning System www.ekahau.com ref.
  3.
• ROVER at CMU ref. 4.
• RADAR at Microsoft ref. 5.
• 3 URL www.ekahau.com
• 4 Suman Banerjee et al. "Rover Scalable
  Location-Avare Computing"
• IEEE Computer, Oct. 2002.
• 5 Paramvir Bahl et al. "RADAR An RF-Based
  In-Building User
• Location and Tracking System". Proc. of
  IEEE INFOCOM, March 2000.

8
What is needed for an Indoor Positioning
Infrastructure?
AP1
AP 2
Access Points for WLAN, in fixed positions.

AP3
Mobile unit
AP4
Direct propagation paths and multipaths
9
Systems for Indoor Positioning III

• Architecture of Indoor Positioning system based
  on classification.
• Use the small Ekahau installation at ITU as an
  example.
• Client-Server system with positioning engine at
  the server.
• 6 Access Points (5 AP at building level 3 and 1
  at level 2) covering positioning
• within approx. 600 m2 at level 3.
• Manually calibration procedure needed, because
  the system classification based.
• Initial estimation of precision approx. 2 m.

10
ITU 3th Floor Initial test of commercial
system from
Ekahau (Finland)
5 Access points at 3th floor
1 access point at 2th floor

11
What initial Platfrom will be available for
Projects and Experiments?

• Place ITU Glentevej.
• PDA (iPAQ) with WLAN interface.
• Java (if possible also C).
• Positioning client in Java, calibrated on a
  part of 3. Floor.
• Initial 3D datastructure for experiments.

12
Research Objectives in Indoor Positioning I

• Establishing a WLAN based open infrastructure
• for research within enhanced positioning and
• tracking algorithms, targeted at applications
  with a
• tightly coupling of position and multimedia
  information.
• Keeping the system as open as possible, with
  respect to
• new and improved WLAN standards.

13
Research Objectives in Indoor Positioning II

• Improved positioning precision based on
  continuously
• calibration of known points, covering the Volume
  of
• Interest, at the building site.
• Platform independent positioning systems, where
• the primary parameters for positioning can be
• targeted for a mixture of geometrically based
• models and classification based models, denoted
• hybrid models.