Model checking of adaptive programs with mode-extended linear temporal logic

Increasingly, software needs to dynamically adapt its structure and behavior at runtime in response to changing conditions in the supporting computing, network infrastructure, and in the surrounding physical environments. By high complexity, adaptive programs are generally difficult to specify, verify, and validate. Assurance of high dependability of these programs is a great challenge. Efficiently and precisely specifying requirements and flexible model checking for adaptation are the key issues for developing dependably adaptive software. This paper introduces a formal model for adaptive programs which have different behavioral modes. We consider that adaptive programs have two behavioral level, functional behavior and adaptation. State machine is used to describe functional behavior in different modes and mode automata is proposed for adaptations. Specifications of adaptive programs are classified into three categories, local, adaptation and global properties from their different scope of dynamic adaptation. To specify and verify specifications on our model, We propose the Mode-extended Linear Temporal Logic (mLTL) and its model checking approach. mLTL extends Linear Temporal Logic (LTL) by adding mode related element and enables describing properties on different modes. Our formal model and mLTL formulae are translated to SMV language and verified in NuSMV model checker.