8.2 KiB
| title | tags | ||
|---|---|---|---|
| 07-location-sensors |
|
Terminology
-
Precision vs. Accuracy
-
How practical are systems that are:
-
Degrees of Freedom (DoF):
-
How many translational and rotational movements are possible?
-
How many degrees of freedom in 3D space?
- 2 DoF?
- 3 DoF?
- 6 DoF?
- 9 DoF?
- Unlimited DoF?
-
(Tri)angulation:
[!INFO] if you have two reference points and the angle towards them, you can get the location
- (Tri)lateration
[!INFO] siimilra to triangulation but you need three points and its calculated using distancees not angles. the distances can be calculated by time of arrival of a signal
Cell-based
- Positioning based on GSM cell ID
- Mobile Positioning System (MPS) by Ericsson to improve accuracy
- Few changes of infrastructure
- No changes of user devices
- Cell Global Identity (CGI) for identification of the cell
[!INFO] compute your position based on which cell tower you are connected to. each tower has a unique ID. used by emergency providers, they know where you last logged in
- GSM based on FDMA (Frequency-division multiple access) and TDMA (Time-division multiple access)
[!INFO] "towers" use FDMA or TDMA 21-data-link-layer#multiple access links and protocols Timing advance used for multiple access can be used to calculate time
- Uplink-Time Difference of Arrival (UL-TOA)
[!INFO]
-
Integrated approach based on existing infrastructure
-
Low costs (for user)
-
Low power draw
-
High availability
-
Works indoor and outdoor (no direct line of sight needed)
-
Information not always available to the user (often only network provider)
- Used for “Enhanced 911”
-
Other cell-based location approaches: WiFi/WLan-cells
-
Based on already existing WLAN infrastructure, primarily installed for communication purposes
-
Cell ID requires database with mapping WiFi cell ID to GPS or …. (e.g. created by Apple, Google)
-
Lateration
- Requires accurate information about access point positions
- Measurement of signal strength of various access points
- At least three access points have to be available
- Measurements are influenced by obstacles like walls
-
Improved approach: Fingerprinting
-
Fingerprinting:
- 1st Phase: Radio Map Creation (Offline):
- Measurements of fingerprints at reference points -> stored in fingerprint DB
- Definition of reference points according to accuracy needs and building structure
- 1st Phase: Radio Map Creation (Offline):
-
2nd Phase: Radio Map Usage (Online):
- Fingerprint of current position is taken
- Search for closest matching reference point on radio map Position taken from that reference point or interpolated
-
Fingerprint
- Location: lat,lon,floor
- VisibleAccessPoints{
- Fingerprint DB
- AP1:
- SSID: eduroam
- MAC: 08:17:35:33:5f:80
- signal strength: -64
- AP2
- SSID: VPN/WEB
- MAC: 08:17:35:33:5f:81
- signal strength: -61
- APn o...
Fingerprinting services using Wifi
- ~99% of queries use Google location database (Google Maps Geolocation API)
- https://developers.google.com/maps/documentation/geolocation
- Access is logged
- Used to improve DB
- Privacy issues (e.g. private routers)
- Alternatives (less coverage & accuracy)
- OpenCellID
- Mozilla Location DB
- Not available when offline, cell DB too large
- WLAN cells can move/change - frequent update required
- Cell structure: circular (theory), hexagonal (planning), irregular (reality)
- Neighbouring cells use different frequencies to minimise interference
- Some overlap inevitable
- Typical cell sizes:
- WiFi: 10m - 100m
- 3G (UMTS): 100m - 5km
- 2G (GSM): 100m - 35km
Proximity Sensors & Near Field Communication
- Different technologies for sensing proximity or exchanging data (often dual purpose)
- Stand-alone infrastructure approaches
- Not widely accepted due to special hardware requirements infrastructure costs (for tracking)
- Conceptually often similar to cell-based approaches but require extra infrastructure
- WIPS (Infrared)
- Beacons installed in the rooms sending unique ID
- User’s badges receive signals of local beacons
- Received beacon ID is sent to location server via WLAN
- Server maps received beacon ID to semantic location which is sent back to the user
- Active Badge (Infrared)
- Users carry badge sending specific user ID
- IR-receivers in the rooms receive those signals
- Position of user is tracked by central server
- Energy-efficient badges (sending short signals of 0,1s each 15s)
- Bluetooth Low Energy LE (Bluetooth specification 4.0)
- Not to confuse with “classic” Bluetooth (e.g. used in headsets or for I/O devices)
- Optimized for battery powered sensors
- Months or years of battery liftetime
- Lower data rate (1MBIT/sec)
- Backwards compatible (same frequency and modulation)
-
Two classes for peripherals:
- Beacons (pure static broadcasts:
- iBeacon simply broadcasts UUID
- Additional data possible on request
- Sensors (broad/unicast with sensor data)
- Different profiles temperature, gravity, hear rate, pressure, …
- Notification possible to avoid polling
- Bidirectional communication
- Not primarily focus, mostly for setting parameters
- Beacons (pure static broadcasts:
-
Location and Proximity Sensing using Bluetooth LE
- One beacon (e.g. per room)
- Pure proximity sensing, looking for strongest signal
- E.g. Region monitoring to detect Beacon presence
- Many beacons (e.g. per room)
- Range calculation using Received Signal Strength Indicator (RSSI) and calibrated transmitter power (txPower = RSSI at 1m)
- Trilateration using several beacons but not very accurate
- One beacon (e.g. per room)
-
Range different for beacons but typically between 20-50m (rarely 100m)
- Different protocols on top of Bluetooth LE
- iBeacon (Apple)
- Broadcasts a UUID
- ID is used with database integrated in the app
- Further information on request (e.g. range information)
- Eddystone (Google)
- Beacons broacasts information about the beacon (telemetry frame e.g. battery or sensor information)
- Beacons broadcasts and redirects to an URL (physical web)
- iBeacon (Apple)


















