Dynamically reconfigurable systems have the potential of realising efficient systems as well as providing adaptability to changing system requirements. Such systems are suitable for future mobile multimedia systems that have limited battery resources, must handle diverse data types, and must operate in dynamic application and communication environments. We propose an approach in which reconfiguration is applied dynamically at various levels of a mobile system, whereas traditionally, reconfigurable systems mainly focus at the gate level only. The research performed in the CHAMELEON project 1 aims at designing such a heterogeneous reconfigurable mobile system. The two main motivations for the system are 1) to have an energy-efficient system, while 2) achieving an adequate Quality of Service for applications.

In this paper

This paper introduces a method which can be used to map applications written in a high level source language program, like C, to a coarse grain reconfigurable architecture, MONTIUM. The source code is first translated into a control dataflow graph. Then after applying graph clustering, scheduling and allocation on this control dataflow graph, it can be mapped onto the target architecture. The clustering and allocation algorithm are presented in detail. High performance and low power consumption are achieved by exploiting maximum parallelism and locality of reference respectively. Using our mapping method, the flexibility of the MONTIUM architecture can be exploited.

This paper introduces a transformational design method which can be used to map code written in a high level source language, like C, to a coarse grain reconfigurable architecture. The source code is first translated into a Control Dataflow graph (CDFG), which is minimized using a set of behaviour preserving transformations such as dependency analysis, common subexpression elimination, etc. After applying graph clustering, scheduling and allocation transformations on this minimized graph, it can be mapped onto the target architecture.