An architectural trade capability using the Ptolemy kernel

A perennial problem in the process of developing signal processing systems is identifying an architecture which meets the computational and memory needs of the algorithm yet is still affordable in terms of cost, size, and complexity. To aid in making this decision, an architecture trade tool, using the Ptolemy kernel, has been developed. The purpose of this tool is to provide the user with an easy mechanism of specifying a mapping of algorithmic functional blocks onto an architecture, and simulating its performance. The goal is to allow the user to quickly evaluate many architectures and mappings at a high-level to determine whether or not they satisfy basic criteria, and warrant further investigation. These criteria include computational and memory requirements as well as other system metrics (e.g. size, power, and reliability, etc.) which are often important in selecting an architecture.