J. Kaiser University of Ulm Dept. Of Comp. Structures Jörg Kaiser Dept. Of Computer Structures University of Ulm And gets a lot harder !!! The Computer disappears- The Problem remains ! Challenges in Providing Security in Large Scale Co-Operative Embedded Systems or
J. Kaiser University of Ulm Dept. Of Comp. Structures Sea of SoCs A SoC may encapsulate sensors and/or actuators. A SoC may include hardware/software and mechanical elements (mechatronics). Co-operating Virtual Sensor and Actuator Networks: Large heterogeneous networks of smart sensors and actuators Co-operation of distributed sensors to augment perception of the environement Co-operation of actuators to increase speed, power and quality of actuation High Degree of Interaction and complex mesh of communication System Anonymous, Content-based Communication Models Communication is driven by the active generation of information Messages are routed by content rather than by address Model of a shared information space:
J. Kaiser University of Ulm Dept. Of Comp. Structures interaction structure in co-operative systems: dynamic evolution of interaction topology managebility as: point-to-point connections or client-server structure
J. Kaiser University of Ulm Dept. Of Comp. Structures Network Topology anonymous communication Source of Information is not necessarily known ! How to use its public key? messages may be routed by content rather than by address Routing nodes must read msg content ! Encryption a problem ! producer consumer
J. Kaiser University of Ulm Dept. Of Comp. Structures acceleration speed priority Co-operating vehicles speed position distance direction position warning speed
J. Kaiser University of Ulm Dept. Of Comp. Structures Availability and Integrity of Data become the crucial issues ! How to develop criteria for trusting anonymous information sources ? Based on continuous suspicion ? Authentification of anonymous data ? Certification of anonymous data ? SoCs may provide good conditions to embed security mechanisms ! Can we learn from fault-tolerant computing ? How to develop security in a system where we cannot relate information to individual entities? (in a quickly changing environment)