smartdata para uma internet confi vel de coisas ciber f
play

SmartData para uma Internet Confivel de Coisas Ciber-Fsicas Antnio - PowerPoint PPT Presentation

SmartData para uma Internet Confivel de Coisas Ciber-Fsicas Antnio Augusto Frhlich Summary The IoT Smart Cities, the Smart Grid, and Smart Things CPS @ IoT (vs Smart Phones and Social Networks) Limitations of Current


  1. SmartData para uma Internet Confiável de Coisas Ciber-Físicas Antônio Augusto Fröhlich

  2. Summary  The IoT ● Smart Cities, the Smart Grid, and Smart Things ● CPS @ IoT (vs Smart Phones and Social Networks)  Limitations of Current Protocols ● TCP/IP, SSL, SOAP, CoAP, IPv6, MQTT ● Security, space, time, and trustfulness  IoT Gateways and Future Protocols ● SmartData: semantics, space, time, and trust ● FT-TSTP: Fault-Tolerant Trustful Space-Time Protocol  CPS @ IoT ● UFSC Hydrology Sensor Network ● UFSC Smart Solar Building ● UFSC e-Bus

  3. IoT and Smart Cities source: IBM

  4. Smart Cities on the Smart Grid source: Energy Collective

  5. Smart Cities on the Cloud source: Oppino

  6. Smart Things Everywhere: the Internet of Things source: Google

  7. Are we Building Killer Robots? source: Sony Pictures

  8. Actually, Yes! source: Google, Volvo and Thames

  9. Actually, Yes! source: Google, Volvo and Thames

  10. Internet of PC-like Things  PC-like gadgets ● Android@Linux OS ● Cloud-connected apps ● Ordinary TCP/IP stack  Susceptible to the same problems of the ordinary Internet  Initial push!

  11. We Cannot Build the IoT on old Internet Technology!  Smart light bulb leaks Wi-Fi passwords http://www.bbc.com/news/technology-28208905  Hacker shows off lethal attack on wireless Medical Device http://go.bloomberg.com/tech-blog/2012-02-29-hacker-shows-off-lethal-attack-by-controlling-wireless -medical-device/  “Thingbot”-net http://www.proofpoint.com/products/targeted-attack-protection/internet-of-things.php  Router-based Botnet http://www.darkreading.com/attacks-breaches/router-based-botnet-on-the-loose/d/d-id/1130625  SmartTV hack https://securityledger.com/2013/08/samsung-smart-tv-like-a-web-app-riddled-with-vulnerabilities/  “Hacking Medical Devices for fun and insulin” http://media.blackhat.com/bh-us-11/Radcliffe/BH_US_11_Radcliffe_Hacking_Medical_Devices_WP. pdf

  12. Internet of Cyber-Physical Things  Zillions of smart objects  Embedded Systems ● Resource-constrained (energy, processing, storage, networking) Internet ● Event-driven (parallel and asynchronous) ● Aware of time and space ● Dependable, Secure, and Trustful

  13. IoT ↔ Internet Gateway Do Things have to talk TCP/IP? And be subjected to the same problems of the ordinary Internet?

  14. SmartData for IoT

  15. IoT↔Internet with SmartData

  16. SmartData  SI Units (inspired by IEEE 1451)  Data Origin (x,y,z,t) Unit Value Expiry MAC

  17. SmartData  API Transducer SmartData +SmartData(dev, expiry, mode) +SmartData(region, expiry, period, fuser) +SmartData(controller, period, smartdata...) +operator Value() : Value +operator=(value) +location(): Coordinates +time(): Time +wait()  Messages Region (x,y,z,r,t 0 ,t f ) Unit Mode Error Period Interest Response Origin (x,y,z,t) Unit Value Expiry Error MAC Command Region (x,y,z,r,t 0 ,t f ) Unit Period

  18. Speculative Precision Time Protocol (SPTP) Header

  19. Heuristic Cooperative Calibration Positioning System (HeCoPS) Header

  20. TSTP Security Bootstraping  Elliptic Curve Diffie–Hellman  Poly1305-AES  Synchronous clocks

  21. SmartData Trickling  Machine learning on time-series of real P sensors in the field = P = P =  Sensor and sink run D X the same predictor P = ● Sensor only transmits P = what is miss-predicted D X P = P = P = D X

  22. SmartData Trustfulness  Machine learning on time-series of real il pleut! C’est -5°C sensors in the field il a -3 -2 -1 0 1 2 3 le soleil! ● Feature selection 3 3 Vous êtes 2 2 tous fous, ● Bayesian classification il neige! C’est 30°C les gars! 1 1 ● Offline learning 0 0 0 0 ● NN optimization with GA -1 -1 -2 -2 ● Online prediction with NN -3 -3 0 1 2 3 -3 -2 -1  Data tagged with confidence based on correlation  Security issues!

  23. FT-TSTP -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 7 7 6 6 5 5 4 4 3 3 2 2 1 1 0 0 0 0 -1 -1 -2 -2 -3 -3 -4 -4 -5 -5 -6 -6 -7 -7 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13

  24. Secure SmartData Gateway Enclaved Gateway Application Trustfulness from SmartData Time from SmartData Location from SmartData Data Application JSON SmartData HTTPS TSTP TCP SSL / DTLS Fusion Trickle Energy Auth Crypt Timing IPv4 / IPv6 MAC Routing Location MAC PHY IEEE 802.15.4 / BLE / 5G Internet IoT

  25. SmartData Integrity: Blockchain IoT Gateways Internet IoT Devices Encrypted Ledger o t p y Signed Ledger r s C e h s a H Geolocated Data Cloud Storage Timed Ledger SmartData TPC/IP TSL

  26. SmartData Integrity: Untrusted GW Security Agent public private Hash 1 Data 2 Log 1 Log 1 Hash 2 Encrypted Data 1 Hashes and Log 2 Sampled Data Log n Log 2 Hash n Data n-1 Log n SmartData Cloud Storage

  27. LISHA’s IoT Platform IoT Gateways IoT Devices Internet TPC/IP TSL Learning Machine Analytics SmartData@TSTP Timed Geolocated Signed Encrypted

  28. AI with LISHA’s IoT Platform

  29. AI with LISHA’s IoT Platform source: Universal Pictures

  30. Smart with a Brain, sort of... r e f e r e n c e error input output + Controller Plant measurement Sensors

  31. Smart with a Brain, sort of... Model Model Model Model mode Ruler setpoints r e f e r e n c e error input output + Controller Plant measurement Sensors

  32. UFSC Environmental Monitoring

  33. UFSC Smart Solar Building TOTAL POWER 50 kWp

  34. UFSC Smart Solar Building TOTAL POWER 50 kWp

  35. UFSC Smart eBus

  36. LISHA  The Software/Hardware Integration Lab at the Federal University of Santa Catarina has been established in 1984 to promote research on the frontier between software and hardware  Come to work with us in beautiful Floripa!

  37. Credits  Davi Resner ● Consolidated 10 years of WSN research ● TSTP main developer ● Protocol performance evaluation  César Huegel Richa ● Consolidated the cloud part of the platform ● IoT and Vehicles ● Predictors  Lucas Wanner ● Former PostDoc, now Professor at UNICAMP ● WSN and CPSoC ● AI enthusiast that pushed LISHA into it

  38. Obrigado! Antônio Augusto Fröhlich lisha.ufsc.br

Recommend


More recommend