10 March 2002 doc IEEE 802 15 02126

  • Slides: 23
Download presentation
10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Project: IEEE P

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Project: IEEE P 802. 15 Working Group for Wireless Personal Area Networks (WPANs) Submission Title: Security Suite Comparative Characteristics Date Submitted: 10 March, 2002 Source: Bob Huang Company: Sony Electronics Address: One Sony Drive TA 3 -12, Park Ridge, NJ 07656 Voice: 201 -358 -4409, FAX: 201 -9306397, E-Mail: robert. huang@am. sony. com Re: P 802. 15. 3 Security Suite Abstract: This presentation identifies and discusses some differences in the proposed security suites. The information on the suites was drawn from contributions presented at the Schaumburg ad hoc meeting (February ’ 02) and related at that meeting discussions. The specific contributions were 02106 -08, 0211112 and 02114 r 2. Also considered were the meeting minutes (02122), the Security Sub-committee Status Report (02121) and some personal notes of mine. This contribution provides a practical perspective on the differences in the security suite proposals. Purpose: For information and guidance to 802. 15. 3 prior to the Security Suite selection. Notice: This document has been prepared to assist the IEEE P 802. 15. It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor acknowledges and accepts that this contribution becomes the property of IEEE and may be made publicly available by P 802. 15. Submission Slide 1 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 802. 15 TG

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 802. 15 TG 3 Security Suite Comparative Characteristics A Perspective Submission Slide 2 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 (My) Baseline Thoughts

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 (My) Baseline Thoughts • The cryptographic algorithms proposed are hard to break – offer high security • Greater attention should be paid to the strength of the protocol • It will be difficult to thoroughly evaluate the protocol(s) in a short time • It is best to evaluate the fit of the security suites against the applications: ‘Where the rubber meets the road’ Submission Slide 3 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Presentation Outline •

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Presentation Outline • (My) Baseline Thoughts • Approach to Evaluation • What are the Major Differences? • Look at the Trust Models • Look at Security Topology • Summary & Conclusions Submission Slide 4 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 (My) Baseline Thoughts

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 (My) Baseline Thoughts (cont. ) • The MAC (draft) standard – Must have a mandatory security suite – May have an optional security suite – Message structure (public key object) to allow • The standards text to accommodate divergent architectures as mandatory and optional • Either architecture to be mandatory or optional – Based on the assertion that the security manager implementation was ‘basically free’, the security manager will be included in devices in the distributed system – Agreed to at ad hoc in Schaumburg (Feb ’ 02) Submission Slide 5 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Approach to Evaluation

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Approach to Evaluation Focus on differences • How well do the differences support the applications? • How do the differences impact the user? • How do the differences impact the manufacturer? First: Identify and understand the differences Submission Slide 6 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 What are the

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 What are the Major Differences? Basic premeses for this comparison: • Not considering differences in cryptographic algorithms or protocols • All approaches have flexibility in implementation to allow ‘changing’ the fundamental characteristics – Therefore, must compare ‘native mode*’ • Most used/natural mode • Most efficient mode * Assumption: operation outside of the ‘native mode’ will cost more Submission Slide 7 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 What are the

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 What are the Major Differences? Fundamental differences • Trust model: – Digital certificates (fixed model or infrastructure based) – Consumer/user trust model • Security topology – Centralized (with PNC) – Distributed Submission Slide 8 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Framework for Considering

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Framework for Considering Trust Models • • • When is the first trust decision made? Interoperability When activated Control/flexibility Typical use model Next: Consider trust models against this framework Submission Slide 9 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Trust model: Fixed

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Trust model: Fixed model digital certificates • Imbedded at manufacture (royalty fee applies) – Example given at Schaumburg ad hoc: $0. 05 each • Interoperability: Manufacturers must act in unison • Activated at manufacture (fixed trust) • Control/flexibility – Fixed by definition, therefore no flexibility after deployment – What to accept any device that can authenticate (not restricted to a particular device) – Consumer can not control (who is in the network) • Typical use: DVB, 5 C Submission Slide 10 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Trust model: Infrastructure

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Trust model: Infrastructure model digital certificates • Imbedded at manufacture (royalty) • Interoperability: typically closed system • Activated/deactivated ‘On-line’ – Implies communications infrastructure – Allows tracking of device use • Infrastructure model allows strong central control – Single controlling party: a service provider • not the consumer, • not the manufacture • Typical use: Cable system Submission Slide 11 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Trust model: Consumer/user

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Trust model: Consumer/user • Interoperability: – Dependent on P 802. 15. 3, not on implementation or industry agreements – Immediate interoperability • Activated by user/consumer – At startup – At addition of new devices • Control/flexibility – Consumer has control • Typical use: modeled on user controlled wired security Submission Slide 12 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Trust models: Conclusions

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Trust models: Conclusions • Digital certificate trust: fixed or infrastructure based – Some additional cost at manufacture – Control • By infrastructure • No consumer control in typical model (native mode*) • Consumer/user trust – Similar to physical wired security: consumer controlled your • Which model applies to 802. 15. 3 applications? * Assumption: operation outside of the ‘native mode’ will cost more Submission Slide 13 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 What are the

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 What are the Major Differences? Fundamental differences • Trust model: – Digital certificates (fixed model or infrastructure based) – Consumer/user trust model • Security topology – Centralized (on PNC) – Distributed Submission Slide 14 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Security Topology •

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Security Topology • Security manager implementation is – More complex (costly) than simple device security – More responsibility than simple device • Different security topologies – Centralized (with PNC) – Distributed (each device is security manager) Submission Slide 15 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Centralized Security Topology

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Centralized Security Topology • Security manager located with PNC – Complexity centralized in one powerful device (the PNC) – Simple to activate by user – Some critical tasks performed infrequently (n times), therefore process power may be ‘borrowed’ from other PNC activities • Authentication? • Key provisioning/re-provisioning • Provides unified network wide security • Allows additional (individual) link security to be applied Submission Slide 16 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Distributed Security Topology

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Distributed Security Topology Security manager (capability) located in every device • Piconet is divided into logical security networks How? • Each device decides who it will connect to • Less exposure than centralized security manager – Compartmentalized: Not higher security, more complex • Each device has security manager complexity – Memory, processing power • Critical tasks performed multiple times – up to n x n times (n x n relationships) – Authentication – Key provisioning/re-provisioning Submission Slide 17 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Security Topology Conclusions

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Security Topology Conclusions • Centralized topology – Less complex/lower cost – Easy to use (less user action) – Fewer authentications (better use of radio resources) • Distributed topology – Each device makes trust decision – Compartmentalizes security Submission Slide 18 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Security Topology Number

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Security Topology Number of sequential authentications for 5 devices: Display 1 DVD 2 Player Display & SM 1 Receiver & 0 PNC SM 4 3 DVD 2 Player & SM DTR Camcorder Submission 4 3 Camcorder & SM • Centralized = 4 SM: Security Manager Receiver 0 PNC & SM DTR & SM 38 1 • Distributed = 5 Each authentication consumes channel time, takes CPU cycles and adds delay. Slide 19 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Security Topology How

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Security Topology How does the security topology overlay the MAC control structure? Display 1 DVD 2 Player Display & SM 1 Receiver & 0 PNC SM 4 3 DVD 2 Player & SM DTR Camcorder Receiver 0 PNC & SM 4 3 Camcorder & SM • Centralized = 4 DTR & SM 38 1 • Distributed = 5 SM: Security Manager Submission Slide 20 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 What are the

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 What are the Major Differences? Fundamental differences • Trust model: – Digital certificates (fixed model or infrastructure based) – Consumer/user trust model • Security topology – Centralized (on PNC) – Distributed Submission Slide 21 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Conclusion on Major

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Conclusion on Major Differences Framework success: 1. Different approaches have different advantages 2. If the first big application of 802. 15. 3 fails, it will be harder for the second application to succeed. 3. Therefore consider the first applications that will make 802. 15. 3 a success. 4. Later consider the second applications that will make 802. 15. 3 a success. 5. Choose an appropriate security suite for overall success. Closing note Submission Go with your longest and your strongest ! Slide 22 Bob Huang, Sony Electronics

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Conclusion on Major

10 March 2002 doc. : IEEE 802. 15 -02/126 r 0 Conclusion on Major Differences? Framework success: 1. Different approaches have different advantages 2. If the first big application of 802. 15. 3 fails, it will be harder for the second application to succeed. 3. Therefore consider the first applications that will make 802. 15. 3 a success. 4. Later consider the second applications that will make 802. 15. 3 a success. 5. Choose an appropriate security suite for overall success. Closing note Submission Go with your longest and your strongest ! Slide 23 Bob Huang, Sony Electronics