The Role-Based Access Control (RBAC) model, under which the users are assigned to certain roles while the roles are associated with prescribed permissions, remains one of the most popular authorization control mechanisms. Workflow is a business flow composed of several related tasks. These tasks are interrelated and context-dependent during their execution. Execution context can introduce uncertainty in authorization decisions for tasks. This paper investigates the role-based authorization model with the context constraints and dynamic cardinality constraints. The Petri-net is used to model the authorization process and the formal expression of the model is presented. The general stochastic Petri-net simplifying method is used to analyze and calculate the authorization workload of a role in the system. With this work, given the workflow load, context and role authorization system parameters, we can predict the performance of the constraints role-based authorization system on mobile devices. Based on these performance metrics, the mobile system parameters can be adjusted to achieve the optimal system performance and meet the user demand best.