DISPS-3.1

A New Scalable DSP Architecture for System on Chip (SoC) Domains
Matthias H Weiss, Frank Engel, Gerhard P Fettweis (Mobile Communications Systems Chair, EE, UoT Dresden)

The ongoing advances in semiconductor technology are the enabler for complete System on Chip (SoC) solutions. In this SoC domain Digital Signal Processors (DSPs) are employed to carry out software driven digital signal processing tasks. Although DSPs could still be modified in the SoC domain, they are mainly employed as fixed DSP cores. Possible adaptations to the embedding system cannot be carried out. Thus, our work is targeted to design expandable DSP architectures. To achieve this expandability, we designed a sliced DSP architecture. Here, the number of slices can be adapted towards system needs. Specific system requirements can be achieved by adding dedicated datapaths to these slices. With this approach one magnitude of order in performance boost can be achieved, which creates new demands for I/O processing. Thus, within our DSP architecture we integrated a dedicated I/O processor. In this paper we present this new scalable DSP architecture, tools to map algorithms onto this DSP architecture, and the concept of our new I/O controller. These technologies allow to easily adapt our DSP architecture to different system requirements.

DISPS-3.2

A New Flexible Architecture for Variable Length DCT Targeting Shape-Adaptive Transform
Thuyen Le, Manfred Glesner (Darmstadt University of Technology, Germany)

Shape-assisted block-based texture coding methodologies such as the shape-adaptive DCT raise the need for an architecture which can perform efficiently the transform of variable length N. This paper presents an 1D DCT architecture satisfying the given requirement in terms of scalability and modularity for 2 <= N <=8. The architecture employs a Canonical-Signed-Digit serial multiplication to reduce hardware resources and requires only one multiplier to perform the final scaling. A proposed algorithm searching for an optimal assignment of cosine factors leads to a resource saving of about 12% for the multiplication blocks if Carry-Ripple-Adders are assumed to be used. Different area and speed requirements are possible since only feed-forward paths are involved and easily pipelined. The architecture represents a trade-off between time-recursive, fully modular but operation non-efficient structure and multiplication efficient but irregular and fixed-length implementation.

DISPS-3.3

An adaptive block-matching algorithm for motion estimation
Vasily G. Moshnyaga (Fukuoka University)

A new adaptive algorithm for the block matching motion estimation is presented. The algorithm works in the full-search fashion but unlike the FSBMA it adjusts the number of computations dynamically to picture variation. Due to incorporated mechanism of data-driven thresholding, the proposed approach performs as four times as less operations comparing to the FSBMA while maintaining the same quality of results. Its hardware implementation is simple and compact. A supportive hardware design as well as simulation results on benchmarks are outlined.

DISPS-3.4

Hardware Architecture for Real-Time Distance Transform
Jarmo H Takala (Tampere University of Technology), Jouko O Viitanen (VTT Automation), Jukka P.P Saarinen (Tampere University of Technology)

A distance transform (DT) converts a binary image consisting of foreground (feature) and background (non-feature) pixels into a gray level image where each pixel contains the distance from the corresponding pixel to the nearest foreground pixel. The computation of the exact Euclidean DT is computationally complex task and, therefore, approximations are typically utilized. In this paper, an area-efficient architecture for computing a DT approximation is presented. The architecture utilizes order-based encoded distance representation allowing simple bitwise operations to be used for determining the distance to the nearest foreground pixel in the constrained neighborhood. Tabulated distance values are used thus cumulative errors are avoided. Due to the simple operations real-time operation can be expected.

DISPS-3.5

AN EFFICIENT VLC DECOMPRESSION SCHEME FOR USER-DEFINED CODING TABLES
Bai_Jue Shieh, Chen-Yi Lee (Department of Electronics Engineering, National Chiao Tung University)

With the increase of information and data types, high-throughput and flexible memory-based VLC decoder is required for user- defined coding tables to achieve higher compression ratio. In this paper, we present a memory-based VLC decoder which is quite suitable for the applications with user-defined tables. By parallel loading data into memories, the coding tables can be changed with much less time. The codeword-boundary prediction algorithm breaks the recursive dependency of decoding procedures. As a result, the VLC decoder can be realized on multi-processor architecture and hence the decoding throughput is enhanced significantly. Additionally, the INDEX- OFFSET symbols that can recover all data with pure VLC codeword and smaller table size are presented. Simulation results show that the combination of the proposed VLC decoder and user-defined table can achieve high decompression rate. As a result, it is quite suitable for high data rate applications with user-defined coding tables, such as MPEG-4.

DISPS-3.6

AN ANALOG ASSOCIATIVE MEMORY CHIP FOR VQ IMAGE COMPRESSION
Loris Navoni, Monica Besana, Pier Luigi Rolandi (STMicroelectronics)

This paper presents a hardware implementation of Full-Search Vector Quantization Image Compression using an associative memory chip based on analog flash technology. Taking advantage of the features of this architecture, that performs a parallel search on 4K 64-elements codebook in 4.6 micro sec., encouraging results have been obtained in terms of perceived image quality and computation speed.

DISPS-3.7

A VLSI IMPLEMENTATION OF A REVERSIBLE VARIABLE LENGTH ENCODER/DECODER
Mario Novell (Electrical Engineering Department, University of California, Los Angeles), Steve Molloy (Luxxon Corporation)

Variable Length Codes (VLCs) are known for their efficient compression, but are susceptible to noisy environments due to synchronization losses that can occur from bit error propagation. Recent interest in Reversible Variable Length Codes (RVLCs) has come about due to the growing need for wireless exchange of compressed image and video signals over noisy channels and the ability for RVLCs to provide greater error robustness than their non-reversible counterparts (VLCs). With the current ITU H.263+ and ISO MPEG-4 standards already using RVLCs, low power implementations of the RVLC are essential in providing error robustness in real-time systems, while minimizing power consumption. This paper will present the first published VLSI architectures of a low power reversible variable length encoder and decoder. Results show power consumption of less than 1 mW for both encoder and decoder, with an additional 65% increase in area for the decoder over that of a conventional VLD design.

DISPS-3.8

Parametric Spectral Estimation on a Single FPGA
Stephen J Bellis, William P Marnane (National Microelectronics Research Centre, University College Cork), Peter J Fish (School of Electronic Engineering & Computer Systems, University of Wales, Bangor)

Parametric, model based, spectral estimation techniques can offer increased frequency resolution over conventional short-term fast Fourier transform methods, overcoming limitations caused by the windowing of sampled, time domain, input data. However, parametric techniques are significantly more computationally demanding than the Fourier based methods and require a wider range of arithmetic functionality; for example, operations such as division and square-root are often necessary. These arithmetic processes exhibit communication bottleneck and their hardware implementation can be inefficient when used in conjunction with multipliers. A programmable, bit-serial, multiplier/divider, which overcomes the bottleneck problems by using a data interleaving scheme,is introduced in this paper. This interleaved processor is used to show how the parametric Modified Covariance spectral estimator can be efficiently routed on a field programmable gate array for real-time applications.

< DISPS-2 DISPS-4 >

Last Update: February 4, 1999 Ingo Höntsch