Embedded systems continue to execute computational- and memory-intensive applications\nwith vast data sets, dynamic workloads, and dynamic execution characteristics. Adaptive distributed\nand heterogeneous embedded systems are increasingly critical in supporting dynamic execution\nrequirements. With pervasive network access within these systems, security is a critical design\nconcern that must be considered and optimized within such dynamically adaptive systems.\nThis paper presents a modeling and optimization framework for distributed, heterogeneous\nembedded systems. A dataflow-based modeling framework for adaptive streaming applications\nintegrates models for computational latency, mixed cryptographic implementations for inter-task\nand intra-task communication, security levels, communication latency, and power consumption.\nFor the security model, we present a level-based modeling of cryptographic algorithms using mixed\ncryptographic implementations. This level-based security model enables the development of an\nefficient, multi-objective genetic optimization algorithm to optimize security and energy consumption\nsubject to current application requirements and security policy constraints. The presented\nmethodology is evaluated using a video-based object detection and tracking application and several\nsynthetic benchmarks representing various application types and dynamic execution characteristics.\nExperimental results demonstrate the benefits of a mixed cryptographic algorithm security model\ncompared to using a single, fixed cryptographic algorithm. Results also highlight how security\npolicy constraints can yield increased security strength and cryptographic diversity for the same\nenergy constraint.
Loading....