1. Peter Kropf Ubiquitous Computing - Hiver 2006/20071 Models

2. Peter Kropf Ubiquitous Computing - Hiver 2006/20072 How to get there? ◆ Device miniaturization and wireless connectivity ◆ Integration of computing devices in our physical world ◆ Association, adaptation, integration, interaction and coordination of software and hardware components ◆ Interoperation in a volatile and context-aware environment

3. Peter Kropf Ubiquitous Computing - Hiver 2006/20073 Evolution to Ubiquitous Computing

4. Peter Kropf Ubiquitous Computing - Hiver 2006/20074 Distributed Systems ◆ Definitions ► “A system in which hardware or software components located at networked computers communicate and coordinate their actions only by message passing.” [Coulouris] ► “A system that consists of a collection of two or more independent computers which coordinate their processing through the exchange of synchronous or asynchronous message passing.” ► “A distributed system is a collection of independent computers that appear to the users of the system as a single computer.” [Tanenbaum] ► “A distributed system is a collection of autonomous computers linked by a network with software designed to produce an integrated computing facility.” ► “A distributed system is one in which the failure of a machine you have never heard of can cause your own machine to become unusable”. [L. Lamport] – They have in common: »Distributed hardware and/or »Distributed data and/or »Distributed control »Computers connected through some network

5. Peter Kropf Ubiquitous Computing - Hiver 2006/20075 Mobile Networking ◆ Mobile Networks in Distributed Systems ► Cellular phone systems (GSM, CDMA, UMTS): sharing frequencies, mobility ► Laptop computers and PDA’s: WLAN connection (home, campus, hot spots), Bluetooth, etc. ► Mobile Ad-hoc Networks © Addison-Wesley Publishers 2000

6. Peter Kropf Ubiquitous Computing - Hiver 2006/20076 Mobile Computing ◆ How to provide continuous connectivity for mobile devices? ◆ How to enable collections of devices to wirelessly communicate where there is no infrastructure? ◆ How to deal with competition for limited bandwidth (with large coverage)? ◆ How to operate devices with minimal energy? (remember : energy needed for transmission is proportional with the range)

7. Peter Kropf Ubiquitous Computing - Hiver 2006/20077 Volatility : a Ubicomp characteristic ◆ Ubicomp systems are volatile : highly dynamic, spontaneous, unpredictable in changes ◆ In the given context of physical integration these events occur frequently: ► Devices appearing and disappearing (and they are very diverse) ► Failure of devices and communication links ► Changes in communication characteristics (e.g. bandwidth) ► Creation and destruction of associations – logical communication relationships – between software components of the devices ◆ Volatile connectivity means : ► frequent disconnection ► Variable bandwidth and latency

8. Peter Kropf Ubiquitous Computing - Hiver 2006/20078 Smart spaces ◆ Mobility takes place in physical spaces – ubiquity is embedded in physical spaces ◆ A smart space is any physical space with embedded services ◆ Physical mobility : environments where devices enter and leave (includes failure and replacement) ◆ Logical mobility : mobile processes (agents) move in or out of smart spaces. ◆ Mobility results in changing associations with other components ► Frequency of change ? ► Security/privacy ?

9. Peter Kropf Ubiquitous Computing - Hiver 2006/20079 A room responding to a user wearing an active badge 2. Infrared sensor detects user ﾕ s ID Hello Roy 1. User enters room wearing active badge User ﾕ s ID 3. Display responds to user Infrared

10. Peter Kropf Ubiquitous Computing - Hiver 2006/200710 Device model ◆ Limited energy (batteries) ◆ Resource constraints (processor speed, storage capacity, bandwidth) ◆ Size and weight constraints ◆ Sensors and actuators for enabling integration with the physical world

11. Peter Kropf Ubiquitous Computing - Hiver 2006/200711 Spontaneous interoperation ◆ Interoperation preferably without user intervention ◆ Interoperation is interaction during association ◆ Association : ► Logical relationship ► Association does not mean connectivity ► Association can be pre-configured or spontaneous

12. Peter Kropf Ubiquitous Computing - Hiver 2006/200712 Examples of pre-configured versus spontaneous association Pre-configuredSpontaneous Service-driven: email client and server Human-driven: web browser and web servers Data-driven: P2P file-sharing applications Physically-driven: mobile and ubiquitous systems

13. Peter Kropf Ubiquitous Computing - Hiver 2006/200713 Smart space ◆ Arriving in a smart space requires: ► Network bootstrapping (for the device) ► Association (to services, software components) ◆ What are scope and scale of a smart space? ► Smart spaces need to have system boundaries (territorial and/or administrative) ► Devices need discovery services to find suitable association partners

14. Peter Kropf Ubiquitous Computing - Hiver 2006/200714 Discovery Services ◆ Discovery relies somehow on a directory ► Service registration/removal ► Definition of lookup attributes (matching exact or not?) ◆ There might be no infrastructure for a directory in a smart space ◆ Directory dynamically built at run-time as a function of the client’s context

15. Peter Kropf Ubiquitous Computing - Hiver 2006/200715 Discovery Services ◆ Interface for automatically (de-)registering services available for associations ◆ Interface for clients to lookup services for associations Methods for service de/registration Explanation lease := register(address, attributes) Register the service at the given address with the given attributes; a lease is returned refresh(lease)Refresh the lease returned at registration deregister(lease)Remove the service record registered under the given lease Method invoked to look up a service serviceSet := query(attributeSpecifica tion) Return a set of registered services whose attributes match the given specification

16. Peter Kropf Ubiquitous Computing - Hiver 2006/200716 Directories ◆ Discovery : server based or serverless by collaboration : ► Push model ► Pull model ◆ Uses broadcast, multicast, lease policies, etc… ◆ Problems : ► often bound to subnets ► Scope : RF reach (not controllable), exact match ◆ Physical association : human input, sensing the context, temporal correlation, direct association

17. Peter Kropf Ubiquitous Computing - Hiver 2006/200717 Example: Service discovery in Jini Printing service Lookup service Lookup Printing service admin admin, finance finance Client Corporate infoservice 2. Here I am:..... 4. Use printing service Network 3. Request ‘printing’ 1. ‘finance’ lookup service

18. Peter Kropf Ubiquitous Computing - Hiver 2006/200718 Interoperation ◆ Components should have a reasonable chance of interoperating with a functionally compatible component even in a different type of smart space than it was developed for. ◆ Global agreement between SW-developers, as minimal as possible ◆ Possibilities : 1. Allow interfaces to be heterogeneous : this requires adapters 2. Constrain interfaces to be identical in syntax. Examples : UNIX pipes, HTTP “GET” and “POST” ◆ Data-orientation : a component can be invoked by any other one that knows the fixed interface ◆ Object-orientation : varying set of interfaces. Can only be invoked by those that know these interfaces (e.g. Corba)

19. Peter Kropf Ubiquitous Computing - Hiver 2006/200719 Interoperation ◆ Models between indirectly associated components : ► Event systems : publish and subscribe principle ► Tuple spaces : fixed generic interface is used for adding and retrieving structured data (tuples) to/from a tuple space (Example : Linda, LIME) ◆ Direct device interoperation is an alternative, but merely only for data transfer or mobile code (security problems)

20. Peter Kropf Ubiquitous Computing - Hiver 2006/200720 Sensing and Context Awareness ◆ Context is typically the location, identity and state of people, groups and computational and physical objects.” (Dey, Salber, & Abowd, 2001) ◆ Being integrated with the physical world is a ubicomp characteristic ► Processing data collected from sensors ► context/-awareness to respond to physical (sensed) environment ◆ “Problem 1: Always hard to interpret human intensions and needs based on simple sensors ◆ Problem 2: How do human agents know how their actions will be interpreted and used by the system?

21. Peter Kropf Ubiquitous Computing - Hiver 2006/200721 Sensors ◆ Combination of HW-SW to measure contextual values : ► Location, velocity, orientation ► Ambient conditions (temperature, sound, etc.) ► Presence (RFID, infrared, microphone, biometric sensors, etc.) ◆ Problem : matching physical context with semantic context : what is the meaning of contextual attributes? ◆ Error model : accuracy, confidence? ◆ Sensors may be part of static or dynamic architectures

22. Peter Kropf Ubiquitous Computing - Hiver 2006/200722 Location sensing ◆ Location: ► User/object determines its own location ► Something else determines the user’s/object’s location ◆ Requirements : ► Generality : types of sensors, data delivered. ►Example : location stack with 4 layers : sensor, measurement, fusion, arrangement (relationship between objects) ► Scalability : number of objects to locate (may be large), rate of updating, mobility, real-time responsiveness, GIS integration

23. Peter Kropf Ubiquitous Computing - Hiver 2006/200723 Locating an active bat within a room 3. Ultrasound receivers 2. Active bat 1. Base station sends timing signal to ultrasound receivers and radio signal to bat simultaneously emits ultrasound signal on receipt of radio signal report times of flight of ultrasound pulse 4. Base station computes distances to ultrasound receivers from times of flight, and thus position of bat

24. Peter Kropf Ubiquitous Computing - Hiver 2006/200724 Some location-sensing technologies TypeMechanismLimitationsAccuracyType of location dataPrivacy GPSMultilateration from satellite radio sources Outdoors only (satellite visibility) 1 – 10m Absolute geographic coordinates (latitude, longitude, altitude) Yes Radio beaconing Broadcasts from wireless base stations (GSM, 802.11, Bluetooth) Areas with wireless coverage 10m – 1km Proximity to known entity (usually semantic) Yes Active BatMultilateration from radio and ultrasound Ceiling sensors 10cmRelative (room) coordinates. Bat identity disclosed Ultra Wide Band Multilateration from reception of radio pulses Receiver in stallations 15cmRelative (room) coordinates Tag identity disclosed Active badge Infrared sensingSunlight or fluorescent light Room sizeProximity to known entity (usually semantic) Badge identity disclosed Automatic identification tag RFID, Near Field Communication, visual tag (e.g. barcode) Reader installations 1cm – 10m Proximity to known entity (usually semantic) Tag identity disclosed Easy LivingVision, triangulation Camera installations VariableRelative (room) coordinates No mounted

25. Peter Kropf Ubiquitous Computing - Hiver 2006/200725 Sensors : challenges ◆ Abstraction from sensor data ◆ Combination of sensor output (sensor data fusion) ◆ Dynamicity of the context ◆ Static or dynamic collection of sensors ◆ Non-functional requirements (e.g. energy saving operation, screen size for output, etc.) ◆ How to produce semantically rich information accurately from raw sensor data?

26. Peter Kropf Ubiquitous Computing - Hiver 2006/200726 Relation to Other Areas