ECE734    Project proposal                          
                         
Anne Pratoomtong

A survey on Reconfigurable Computing for Signal Processing Appplicaitons

Signal processing system has to meet real-time requirements and provide very large throughput in the magnitude of Giga-operation per second. These real time requirements put the pressure on the hardware running signal-processing application.  Therefore, custom hardware such as ASIC, dedicated DSP processor or scalable parallel hardware architecture is frequency used as an implementation mechanism for many DSP applications to gain high performance. Configurable computing has emerged in the past decade as a method for dynamically adapting computing has emerged in the past decade as a method for dynamically adapting computer algorithms, architecture and hardware to track the variations in the computation and its environment. Signal processing application has a small and well define structure, which can be represented by signal flow graph. The signal-processing algorithm is based on the available of a priori information about the signal source and the noise in the environment. The performance of the implemented system largely depends on the environmental conditions as well as addition factors, such as the computing resources available. These major characteristics of signal processing application make the reconfigurable and structure adaptive as an attractive approach to design a signal processing system. 

This project surveys academic research in reconfigurable and structure adaptive computing for DSP systems in the past fifteen years. The main focus of this project is to survey systems in the past fifteen years. The main focus of this project is to survey design methodology and tools for dynamic runtime reconfigurable system. The survey presents definition and history of reconfigurable computing for DSP, various design tools and methodologies for reconfigurable and structure adaptive signal processing system, and some application of reconfigurable and structure adaptive computing in high performance signal processing system.