Visions: a Disruptive Technology

Visions: a Disruptive Technology

Technologies and Issues

Components and Architecture

Integration of multiple ICT technologies

• Integration of multiple ICT technologies

• identification and tracking technologies
• sensor networks
• network protocols (cf. Future Internet
• autonomic, pervasive and ubiquitous computing
• AI, knowledge management, semantics

• Key issues:

• interoperability
• security/trust and privacy
• low resources (=> revisit protocols and algorithms
• implemented in Internet and Web)
• scalability

Network/Internet Point of View

• Network/Internet Point of View

• Things point of View

A Digital Ecosystem Point of View?

• A Digital Ecosystem Point of View?

Key issue: object identification and tracing

• Key issue: object identification and tracing

• A prominent killer app: Electronic Product Code

• basic goal: product/object traceability
• RFID tags attached to products

• RFID advantages

• low cost
• maturity
• no need of power (passive tags)
• lifetime
• strong support from supply chain and consumer goods industries

• Other basic things

• Mobile equipments (Near Field Communications (NFC), GSM…)
• Sensors and (Wireless) Sensor Networks ((W)SN)

Development of supply chain platforms

• Development of supply chain platforms

• Ex: WISP (Wireless Identification and Sensing Platforms) project

• “Philosophical” vision: spime (B. Sterling)

• object tracking through space and time
• autonomy and collaboration

• (Web) Semantic vision: The Web of Things

Key issue: object2object communication

• Key issue: object2object communication

• First approach: develop specific communication protocols

• adapted to each type of things and type of applications
• need for standardization
• which compatibility with Internet?

• Second approach: (re-)use IP

• integrate IP and IEEE 802.15.4 (6LoWPAN) (IP for Smart Objects (IPSO) Alliance)
• make IP lighter (Internet Ø)

• Third approach: (re-use) 802.15.4

• enrich 802.15.4 (Zigbee)

Visions: a Disruptive Technology

• Visions: a Disruptive Technology

• Technologies and Issues

• Components and Architecture

Identification (“sensing”)

• Identification (“sensing”)

• (passive, active) RFID tags
• sensor networks

• Communication

• see discussion above
• interface object/network
• embed the TCP/IP stack into the devices (TinyTCP, mIP, IwIP…)?

• Integration

• object and service discovery
• object and service cataloging
• service composition/orchestration

• Intelligence and Collaboration

• Security and Privacy

Ultimate goal: unique/universal Id for naming and addressing individual objects i.e., to attach an ego to each object, condition to develop ego-centric applications (cf. Jacob and the Angel (Genesis))

• Ultimate goal: unique/universal Id for naming and addressing individual objects i.e., to attach an ego to each object, condition to develop ego-centric applications (cf. Jacob and the Angel (Genesis))

• Naming is difficult!

• ONS: Object Name Service
• basically, RFID tag/EPC code → URI of a description file (Object Code Mapping Service-Direct Search (OCMS-DS)
• more complex Object Code Mapping Service-Reverse Search (OCMS-DS): description → EPC code(s)

• Addressing is difficult!

• stupid but tricky issue: RFID addresses are different from IPv6 addresses (64-96 bits vs 128 bits)
• addressing moving objects is even more difficult

From host2host to object2object

• From host2host to object2object

• TCP is not adapted

• designed for long-lasting connections while objects (like tags or sensors) exchange small pieces of data => handshake + congestion control/retransmit/recovery + flow control + buffering procedures too complex

• Very heterogeneous networks and traffic

• Scalability?

• Quality of service?

A definitive threat for privacy!

• A definitive threat for privacy!

• A security nightmare!

• Security

• IoT = a kind of unsupervised mobile/pervasive grids whose end-components are resource limited tiny objects = a security nightmare
• memory segments of tags are protected by (short) password
• physical attacks
• Man in the Middle attacks
• cryptographic techniques too CPU-intensive for low energy objects
• multiple administrative domains (cf. grids)

Privacy

• Privacy

• all your life can will be traced => possible monitoring, mining, analysis
• connection possible with Linked Open Data => worsen the threats
• open air connections => possibility of eavesdropping
• not only your digital life but also your “analogical” life
• you cannot even know what is sensed about you, when it is sensed, etc. Sensors do not ask for permission (cf. video surveillance)
• no “forget option”

Privacy (cont’d)

• Privacy (cont’d)

• Basic approach (e.g. EEXCESS EU project, W3C P3P (Platform for Privacy Preference)
• user defined policy
• privacy proxy
• negotiation protocol
• anonymization/pseudomization
• integration of reputation and trust mechanisms (cf. course on security and privacy)
• Issues
• cryptographic techniques are too complex
• scalability

RFID

• RFID

• EPCGlobal
• ISO, IEC (Int. Electrotechnical Commission), CEN, NAFTA…
• industrial consortia

• M2M

• ETSI (European Telecommunications Standards Institute )
• 7 standard bodies joined in 2012

• Communication

• 6LoWPAN
• ZigBee
• NFC
• … all communications standard bodies (ISO, IEEE…)

Yet another buzzword or a revolution?

• Yet another buzzword or a revolution?

• Strong support from the industry

• IoT platforms yet exist: xively (ex cosm, ex pachube), sen.se, etc.

• Close to reality in “closed” ecosystems

• Far from reality in “open” ecosystem”

• Need for an holistic vision → multi-scale digital ecosystem?