DSP Architecture Design Essentials

Book Preface

The advancement of semiconductor industry over the past few decades has made significant social and economic impacts by providing inexpensive computing and communication technologies. Our ability to access and process increasing amounts of data has created a major shift in information technology towards parallel data processing. Today’s microprocessors are deploying multiple processor cores on a single chip to increase performance; radios are starting to use multiple antennas to transmit data faster and farther; new technologies are needed for processing large records of data in biomedical applications. The fundamental challenge in all these applications is how to map data processing algorithms onto the underlying hardware while meeting application constraints for power, performance, and area. Digital signal processing (DSP) architecture design is the key for successful realization of many diverse applications in hardware.

The tradeoff of various types of architectures to implement DSP algorithms has been a topic of research since the initial development of the theory. Recently, the application of these DSP algorithms to systems that require low cost and the lowest possible energy consumption has placed a new emphasis on defining the most appropriate solutions. The flexibility consideration has become a new dimension in the algorithm/architecture design. Traditional approach to provide flexibility has been through software programming a Von Neumann architecture. This approach was based on technology assumptions that hardware was expensive and the power consumption was not critical so time multiplexing was used to provide maximum sharing of the hardware resources. The situation now for highly integrated system-on-a-chip implementations is fundamentally different: hardware is cheap with potentially 1000’s of multipliers and adders on a chip and the energy consumption is a critical design constraint in portable applications. Even in the case of applications that have an unlimited energy source, we have moved into an era of power-constrained performance since heat removal requires the processor to operate at lower clock rates than dictated by the logic delays.