ADVANCED SYNTHESIS OF DSP ALGORITHMS
IN MODERN PROGRAMMABLE ARCHITECTURES
Tadeusz Łuba, Mariusz Rawski, Paweł Tomaszewicz
Institute of Telecommunications,
Warsaw University of Technology,
Nowowiejska 15/19, 00-665 Warsaw, Poland,
e-mail: { luba, rawski, p.tomaszewicz }@tele.pw.edu.pl
Abstract: In this paper, using FIR filters as an example, the discussion on efficiency of
different implementation methodologies used in DSP application targeted modern FPGA
architectures is presented. Nowadays programmable technology provides possibility to
implement digital system with use of specialized embedded DSP blocks. On the first
place, however, this technology gives the designer the possibility to increase efficiency of
designed system by exploitation of parallelisms of implemented algorithms. Moreover, it
is possible to apply special techniques, such as distributed arithmetic (DA). Since in this
approach general purpose multipliers are replaced by combinational LUT blocks, it
allows constructing digital filters of very high performance. Additionally, application of
the functional decomposition based method to LUT blocks optimisation and mapping has
been investigated. The paper presents results of comparison of different design
approaches. Copyright © 2006 IFAC
Keywords: Digital signal processors, FPGA, Logic synthesis, Functional decomposition.
1. INTRODUCTION
Digital-signal processing (DSP), thanks to explosive
growth in wired and wireless networks and in
multimedia, represents one of the hottest areas in
electronics. Within the next few years we can
anticipate more wireless and digital communication
standards, pervasive computing and widespread
availability of data on the move. Therefore the
growing requirements for processing speeds of the
order of 10-100 billions of operations per second, the
need for rapid prototyping and software definable
architectures will further penetration of FPGAs into
the DSP communication.
In recent years digital filters have been recognised as
primary digital signal processing operation. With
advances in digital technology they are rapidly
replacing analogue filters, which were implemented
with RLC components. Digital filters are typically
implemented as multiply-accumulate (MAC)
algorithms with use of special DSP chips (Lapsley, et
al., 1997; Lee, 1988; Lee, 1989). When one speaks of
a DSP chip, one perceives of an off-the-shelf chip,
which has an architecture designed to efficiently
execute math-intensive DSP algorithms. Such devices
are based on concept of RISC processors with an
architecture consisting fast array multipliers. By using
pipeline architecture the speed of such
implementation is limited by the speed of array
multiplier. But there are several other chip types that
are better choices for specific DSP implementations.
Recent reductions in the cost of high-density FPGAs,
combined with advances in software-oriented FPGA
design tools, have led to a corresponding rise in the
use of these devices to handle functions that are
traditionally the domain of DSP processors, while at
same time dramatically reducing the risks and up-
front costs of custom ASIC solutions. Nowadays, by
taking advantage of the opportunities provided by
programmable digital devices, we are able to build
very complex digital circuits and systems at relatively
low cost in a single programmable structure. Field
programmable gate arrays now possess sufficient
performance and logic capacity to implement a
number of digital signal processing algorithms
effectively.