Software application acceleration, using parallelization techniques and dedicated
hardware components, is often an optimization compromise in a cost-benefit relationship, during the migration of software processes to hardware-intellectual property dedicated cores or accelerators. In real-time applications extra care is needed
in dealing with these issues, so that the real-time requirements of the application
aren't compromised. Through linear programming mathematical tools, it is possible to establish a good compromise, minimizing the desired system metric, like power consumption, temporal latency or silicon area. In order to characterize the temporal evolution on a system level, an integrated co-simulation environment where all metric costs are contemplated should be used. Such an environment would not only be useful by providing measurements of the system metrics, so the linear mathematical tools can aid in design optimizations, but also in system modelling.
Usually, hardware-software co-design is a long process of iterative nature. If the system is modelled and simulated on separate application domains and later implemented and debugged physically, the development time can grow exponentially. An isolated validation, as far as application domains are concerned, does not guarantee integral system functionality, and with an integrated co-simulation environment, system problems can be detected early before moving to the physical implementation phase.  By following a design flow aided by co-simulation, not only is the development process sped up, but also provided with platform independency, since the system can be developed in its entirety in a host platform without being bound to a target platform. Platform independence in development has some obvious benefits, as the system can be modelled first, optimized later for certain metrics, and only afterwards is a target platform is selected. Also, design teams with limited available prototypes are not slowed down by limited platform resources.
The aim of this dissertation is the development of such an integrated co-simulation environment for domain specific applications, namely in the field of power systems. As such, it will be composed of software simulation tools, hardware simulation tools and power system simulation tools synchronized and interacting to provide a qualitative and temporal integral simulation of the system being designed. As means of validation, power systems test scenarios will be developed following a design flow aided by the designed co-simulation environment.