Abstract
Critical Software systems must recover when they experience degradation, either through external actors or internal system failures. There is currently no accepted generic methodology used by the software engineering community to design self-healing systems. Such systems identify when they require healing resources, and then change their own behavior to acquire and utilize these same resources. This study investigates using a design pattern to build such a system. It uses simulated robot tank combat to represent a challenge faced by an adaptive self-healing system. It also investigates how an adaptive system chooses different behaviors balancing its actions between healing activities, movement activities, and combat activities.The results of this study demonstrate how an adaptive self-healing system utilizes behavior selection within a contested environment where other external actors attempt to deny resources to it. It demonstrates how a multi-system architecture inspired by cognitive science its behavior to maximize its ability to both win matches, and survive. This study investigates system characteristics such as how behaviors are organized and how computer memory is utilized. The performance of the adaptive system is compared with the performance of 840 non-adapting systems that compete within this same environment.
Highlights
This paper investigates building a self-healing adaptive system that uses a design pattern inspired by the neocortex of the mammalian brain
This study demonstrates a self-adaptive system that shares an environment with another system, where these two systems compete for shared resources while attempting to degrade or destroy each other
The Hypothesis H1 stated that the performance of the self-adaptive system would exceed the average performance of the non-adaptive system
Summary
This paper investigates building a self-healing adaptive system that uses a design pattern inspired by the neocortex of the mammalian brain. This pattern uses two separate behavior selection systems to choose behaviors based on the availability of data and the needed speed of the behavior change. The research team utilizes an extended RRobots simulation to provide an experimentation environment [1] This is used to investigate the efficacy of different adaptation strategies based on machine learning techniques and a two-system behavior selection process. The Neocortex Adaptive Systems Pattern (NASP) [21] provides a useful starting point for building adaptive systems This design pattern codifies an architectural approach used to build a decision system that uses short term and longer-term behavior selection approaches. The cognitive science aspect loosely models the organization of a two-system decision system documented by Kahneman [23] and a hierarchical behavioral selection architecture
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
