Modeling, Mapping and Simulating Scalable Hierarchical Embedded Signal Processing Systems (MASSIVE) (LES.5028)

Designing a complex, large scale and distributed embedded signal processing system is hardly possible without a strategy to master the complexity. Such a strategy is necessary to ensure completeness and consistency of specifications, acceptable duration of the design process, and correct behavior of the resulting system. These are the objectives of the underlying proposal in which we propose to investigate methods for a structured design of a hierarchy of embedded subsystems. The SKA telescope concept will be used as the target example of a complex, large-scale and distributed signal processing system. The aim of the current Progress proposal is to develop methods and tools to master the complexity of large scale embedded signal processing systems. Mastering system complexity paves the way to reducing the system architecture design cycle. The challenge is to pioneer structured design of hierarchical embedded systems, that is, embedded systems that are hierarchies of embedded systems. The design of single embedded systems will be covered by various other Progress proposals. We propose to

1. model the system as a hierarchy of modules that are distributed in space,
2. to explore the overall architecture design space by simulating the behavior and measuring the performance of essential parts and their relations on a level high enough that we can master the complexity and - at the same time - preserve interpolation and extrapolation options at various levels in the hierarchy, and
3. to demonstrate the achievements by realizing a downscaled prototype demonstrator of the large-scale SKA system.

For results, see http://www.liacs.nl/research/LERC/index.html

Eight companies and one other university is involved in this project.