Self-Repairing systems modeling and verification using AGG

Self-repairing (or healing) systems are systems equipped with a mechanism that monitors the system behaviour to determine whether it behaves within prefixed parameters. If a deviation exists, then the system itself is in charge of adapting its configuration. In this paper we show how to model self-repairing systems by means of dynamic software architectures (DSAs). DSAs are formalized as typed (hyper) graph grammars (TGGs) and this formalization enables verification of correctness and completeness of self-repairing systems. DSAs are modeled and verified by using the attributed graph grammar system (AGG). The overall approach is applied to a traffic light system case study.