The problem of localization is a challenging task, and yet extremely crucial for many applications. While GNSS (such as GPS, or Galileo) solves the problem of localization for wide area of applications, the problem persists in many other cases like if satellite coverage is not available (skyscraper, tunnel), or the accuracy level is not sufficient. In such case, Invoking other devices and combining information from multiple sources may be considered as a potential solution.

1. LOCALIZATION AS A SERVICE IN
INTELLIGENT TRANSPORT SYSTEM
Saber Ferjani
PhD Student
Hana Lab - ENSI - Manouba University
http://www.fastcoexist.com/1681562/solar-roads-charging-roads-and-the-future-of-transportation

2. Traffic congestion
MOTIVATION
 Problems:
 Waste of time
 Waste of energy
 Air pollution
 More accident
 Solutions
 Construct new roads
 ReduceTraffic
 Improve transport efficiency
2/40

3. MOTIVATION
ITS functional areas:
 ATMS:AdvancedTraffic Management Systems
 ATIS:AdvancedTraveler Information Systems
 CVO: CommercialVehicle Operations
 APTS:Advanced PublicTransportation Systems
 AVCS:AdvancedVehicle Control Systems
3/40

4. MOTIVATION
4/40

5. OUTLINE
1. Localization algorithms
2. Future internet trends
3. Web service composition
4. Conclusion
5/40

6. 1. LOCALIZATION ALGORITHMS
1. One hop localization
2. Multi-hop localization
3. Probabilistic localization
6/40

7. 1. LOCALIZATION ALGORITHMS
1. ONE HOP LOCALIZATION
Trilateration
 Determining the location of a point by
measuring distances, using the
geometry of circles (2D), or spheres
(3D).
Triangulation
 Determining the location of a point by
measuring angles to it from known
points at either end of a fixed
baseline.
Location?
Location?
𝑅 𝑎
𝑅 𝑏
𝑅 𝑐 𝛼
𝛽
7/40

8. 1. LOCALIZATION ALGORITHMS
1. ONE HOP LOCALIZATION
1. PROPAGATION MODEL
The RSS (received signal strength) is provided by most radio chips.
Known :
 The path loss model
 Transmission power 𝑃𝑡𝑥
 Path lost coefficient α
 Receiver can determine the distance d to the transmitter :
𝑃𝑡𝑥
𝑅𝑆𝑆
𝑅𝑆𝑆 = 𝑐
𝑃𝑡𝑥
𝑑 𝛼
8/40

9. 1. LOCALIZATION ALGORITHMS
1. ONE HOP LOCALIZATION
2.TIME BASED
1. Time of arrival
The distance can be estimated, using the transmission time.
• The speed of propagation is known.
• Receiver and sender are synchronized
2. Time Difference of arrival
Use two transmissions mediums of different propagation speeds to generate an implicit
synchronization
Tx TxRx Rx
TOA
TDOA
9/40

10. 1. LOCALIZATION ALGORITHMS
1. ONE HOP LOCALIZATION
3. ANGLE BASED
1. DOA: Direction of arrival is the direction that
maximizes the RSS of directional antenna
2. AOA: Angle of arrival is the angle between
DOA, and a conventional direction
Measured using:
Rotatable DirectionalAntennas
TOA, RSS Differences ofAntenna array
Smart Antenna:
ESPRIT: estimation of signal parameters via rotational
invariant techniques
MUSIC: MUltiple SIgnal Classification
10/40

11. 1. LOCALIZATION ALGORITHMS
2. MULTI-HOP LOCALIZATION
Centralized
Designed to run on a central machine with
powerful computational capabilities.
 Multi-Dimensional Scaling
 Semi-Definite Programming
Distributed
Designed to run in network, using massive
parallelism and internode communication.
 Beacon Based Localization
(DV-hop, DV distance, Iterative localization)
 Coordinate System Stitching
Higher Accuracy
Low propagation error
Low Computation cost
Low Communication cost
High Computation cost
High Communication cost
Lower Accuracy
High propagation error
11/40

12. 1. LOCALIZATION ALGORITHMS
3. PROBABILISTIC LOCALIZATION
12/40

13. 1. LOCALIZATION ALGORITHMS
3. PROBABILISTIC LOCALIZATION
In probabilistic localization, we
distinguish two update steps:
1) ACTION:
 Use proprioceptive sensors to estimate
location .
 During this step, uncertainty grows.
2) PERCEPTION:
 Combine data from exteroceptive
sensors with the belief before the
observation.
 In this step, location uncertainty shrinks.
http://www.asl.ethz.ch/education/master/mobile_robotics/year2010/5b_-_Probabilistic_Map_Based_Localization_and_Markov_printable.pdf
13/40

14. 1. LOCALIZATION ALGORITHMS
3. PROBABILISTIC LOCALIZATION
1. KALMAN FILTER APPROACH
uses a single, well-defined Gaussian
probability density function.
Updating the parameters of the
Gaussian distribution is all that is
required.
For highly nonlinear systems:
 Extended Kalman filter
 Unscented Kalman filters
14/40

15. 1. LOCALIZATION ALGORITHMS
3. PROBABILISTIC LOCALIZATION
2. MARKOV CHAIN APPROACH
The location state is usually
represented as separate probability
assignments for every possible position
in its map.
 Complete Sampling
 Randomized Sampling
 particle filter algorithms
 condensation algorithms
 Monte Carlo algorithms.
15/40

16. 1. LOCALIZATION ALGORITHMS
4. CONCLUSION
0.5
12.5
25
50
6.3 6.8 7.2 7.6
2003 2010 2015 2020
Connected devices (Billion) World population (Billion)
Cisco IBSG white paper “The Internet of Things”, April 2011
16/40

17. 2. FUTURE INTERNET TRENDS
1. Current challenges
2. Semantic web
3. Dbpedia
4. SPARQL Query Language
17/40

18. 2. FUTURE INTERNETTRENDS
2. CURRENT CHALLENGES
How can I
travel to
Japan?
18/40

19. 2. FUTURE INTERNETTRENDS
2. CURRENT CHALLENGES
The web is extremely big!
And still getting bigger every minute!
The meaning of web pages can be understood
only by humans!
19/40

20. ”The Web was designed as an information
space, with the goal that it should be useful
not only for human-human communication,
but also that machines would be able to
participate and help…”
Tim Berners-Lee, SemanticWeb Roadmap, Sept 1998
2. FUTURE INTERNETTRENDS
3.SEMANTIC WEB
20/40

21. 2. FUTURE INTERNETTRENDS
3.SEMANTIC WEB
Ontology definition:
 It is a compound word, composed of
 ontos- (ὄντος) meaning “being”
 -logia (λογία) interpreted as "science, study,
theory".
 Ontology, is the science or study of being
Ontology In computer science:
 “An ontology is an explicit, formal
specification of a shared conceptualization.”
 (Thomas R. Gruber, 1993)
Top-Level
Ontology
Task Ontology
Domain Ontology
Application
Ontology
Dublin Core
General Formal Ontology
OpenCyc
Railway Domain Ontology
soccer ontology
music ontology
21/40

22.  XML was created to structure, store,
and transport information.
 XML Separates Data from HTML:
 Simplifies Data Sharing
 Simplifies DataTransport
 Simplifies Platform Changes
 Used to Create New Internet
Languages
2. FUTURE INTERNETTRENDS
3.SEMANTIC WEB
1.XML
22/40

23. Resource
 can be everything
 must be uniquely identified and be
referencable
Description
 via representing properties and relationships
among resources
 relationships can be represented as graphs
Framework
 Combination of web based protocols (URI,
HTTP, XML,...)
 based on formal model
 defines all allowed relationships among
resources
2. FUTURE INTERNETTRENDS
3.SEMANTIC WEB
2.RDF
http://about.me/ferjani +216 22 94 20 94
<Subject> <PREDICATE> <Object>
hasPhoneNumber
23/40

24. 2. FUTURE INTERNETTRENDS
3.SEMANTIC WEB
3.RDF Schema
24/40

25. 2. FUTURE INTERNETTRENDS
3.SEMANTIC WEB
3.RDF Schema
https://openhpi.de/course/semanticweb
25/40

26. Controlled
Vocabulary
Glossarie
Thesauri
Informal
IS-A
Formal
Is-A
Formal
instance
Frames
Value
restrictions
General
logical
constraints
Disjunctness,
Inversiveness,
Part-of…
Informal Formal
2. FUTURE INTERNETTRENDS
3.SEMANTIC WEB
4. Ontology engineering
heavyweight
lightweight
First Order Logics
Description Logics
Logic Programming
FormalTaxonomies
Folksonomies
Data Dictionaries
Terms
Expressivity
26/40

27. Acronym Language Characteristics
XML Extensible Markup Language Extensions for arbitrary domains and specific tasks.
RDF Resource Description Framework
Syntactic conventions and simple data models to represent semantics.
It supports interoperability aspects with object - attribute - value
relationships.
RDFS
Resource Description Framework
Schema
Primitives to model basic ontologies with RDF.
OWL Web Ontology Language
W3C Recommandation
2004-02-10; 2009-10-27; 2010-06-22;
http://www.emc-eu.de/index-Dateien/3_ONTOLOGY_UK.html
2. FUTURE INTERNETTRENDS
3.SEMANTIC WEB
4. Ontology engineering
27/40

28.  Road classification (Motorways, dual
carriageway, etc.)
 Vehicle classification (Truck, car, etc.)
 Location (Area, Point, Section, etc.)
 Geography (Towns, Countries, etc.)
 Events (Accidents, Incidents,
Measures, etc.)
 People (Driver, Passenger, etc.)
 Routes (Urban, interurban, etc.).
http://dx.doi.org/10.1109/ITSC.2006.1707385
2. FUTURE INTERNETTRENDS
3.SEMANTIC WEB
4. Ontology engineering
28/40

29. 2. FUTURE INTERNETTRENDS
4.DBPEDIA
29/40

30. Find all schools within a
5km radius around a
specific location, and for
each school find
coffeeshops that are closer
than 1km.
2. FUTURE INTERNETTRENDS
5. SPARQL QUERY LANGUAGE
30/40

31. 2. FUTURE INTERNETTRENDS
5. SPARQL QUERY LANGUAGE
Query Result
31/40

32. 3. WEB SERVICE COMPOSITION
1. Web services & RESTful services
2. Web process lifecycle
3. Composition methods
32/40

33. 3. WEB SERVICE COMPOSITION
1.WEB SERVICES & RESTFUL SERVICES
Client app code Client service code
Proxy/stub
Encoding
Protocol
Transport
Skeleton
Encoding
Protocol
Transport
attachementdataheader
Jaxb, direct XML
XML, JSON
HTTP
TCP
WADL
SSL HTTP session
Client app code Client service code
Proxy/stub
Encoding
Protocol
Transport
Skeleton
Encoding
Protocol
Transport
attachementdataheader
WSDL
Jaxb, direct XML
XML, Fast-infoset
HTTP, SIP, SMTP
TCP, UDP
WS-Trust, WS-Security,
WS-SecureConversation
WS-ReliableMessaging,
WS-AtomicTransactions
SOAP REST 33/40

34. 3. WEB SERVICE COMPOSITION
1.WEB SERVICES & RESTFUL SERVICES
Eg: Ethernet link
IP
TCP
HTTP
Payload
Constrained link
IP
UDP
CoAP
Payload
Shelby, Z. - Embedded web services - Wireless Communications, IEEE, 2010
34/40

35. 3. WEB SERVICE COMPOSITION
2.WEB PROCESS LIFECYCLE
Annotation
Advertisement
Discovery
Selection
Composition
Execution
35/40

36. 3. WEB SERVICE COMPOSITION
3. COMPOSITION METHODSWebService
Composition
Static
Orchestration WS-BPEL
Choreography WS-CDL, CHOReOS
Dynamic
SemanticWeb
Service
RDF, DAML, OWL-S
36/40

37. 3. WEB SERVICE COMPOSITION
3. COMPOSITION METHODS
1. STATIC COMPOSITION
Orchestration
a central process takes control of the
involvedWeb services and coordinates
the execution of different operations.
Choreography
 it is a collaborative effort focusing on
the exchange of messages in public
business processes.
coordinator
Web
service 1
Web
service 2
1
2
Web
service 33
4
5
Web
service 1
Web
service 3
Web
service 2
1
2
34
http://www.oracle.com/technetwork/articles/matjaz-bpel1-090575.html
37/40

38. 3. WEB SERVICE COMPOSITION
3. COMPOSITION METHODS
2. DYNAMIC COMPOSITION
Specification
Matchmaking
Algorithms
Generation
CSL language
Composabilty
Model
Composition
plans
Web
service
registries
Ontological
organization and
description of
WS
High level
description
of desired
composition
Composite
Service
QoC
parametersComposition
plan cost
Orchestration
Service Composition for the Semantic Web - DOI 10.1007/978-1-4419-8465-4
38/40

39. 5. CONCLUSION
39/40

40. 5. CONCLUSION
 Context-aware systems uses context to provide relevant information and services
 Location is one of the most important context information
 Proliferation of mobile devices improve real time information sharing
 Semantic web allow autonomic web service composition
 ITS functional areas use location with different accuracy level
40/40

41. Thank you for your attention!
41