• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.861-864
• /
• 2001

많은 멀티미디어와 DSP 응용에서 입력과 출력 데이터 길이가 같은 고정 길이 곱셈기가 요구된다. 고정 길이 곱셈기는 확률적인 추정에 근거한 적절한 보상 바이어스를 더해줌으로써 일반적인 병렬 곱셈기와 비교하여 50%의 면적을 줄일 수 있다. 본 논문에서는 CSD 곱셈기에 적합한 고정길이 곱셈기의 구조를 제시하고 전파 캐리 선택절차를 이용한 부호확장제거방법과 결합함으로서 새로운 곱셈기구현 방안을 제시한다. 이 곱셈기의 응용으로서 SSB/BPSK-DS/CDMA 전송방식에 사용되는 힐버트 트랜스포머를 43탭 FIR 필터로 구현하고 기존의 compensation 벡터 방법과 비교하여 약 34%의 부호확장 오버헤드를 줄일 수 있음을 보인다.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.46 no.9
• /
• pp.33-38
• /
• 2009

The group CSD (GCSD) multiplier was recently proposed based on the variation of canonic signed digit (CSD) encoding and partial product sharing. This multiplier provides an efficient design when the multiplications are performed only with a few predetermined coefficients (e.g., FFT). In many DSP applications such as FFT, the (2W-1)-bit product obtained from W-bit multiplicand and W-bit multiplier is quantized to W-bits by eliminating the (W-1) least-significant bits. This paper presents an error compensation method for a fixed-width GCSD multiplier that receives a W-bit input and produces a W-bit product. To efficiently compensate for the quantization error, the encoded signals from the GCSD multiplier are used for the generation of error compensation bias. By Synopsys simulations, it is shown that the proposed method leads to up to 84% reduction in power consumption and up to 79% reduction in area compared with the fixed-width modified Booth multiplier.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.42 no.10 s.340
• /
• pp.29-34
• /
• 2005

The maximum quantization error has serious effect on the performance of fixed-width multipliers that receive W-bit inputs and produce W-bit products. In this paper, we analyze the error bound of fixed-width modified Booth multipliers. Then, the estimation method for the number of additional columns for fixed-width multipliers is proposed to limit the maximum quantization error within a desired bound. In addition, it is shown that our methodology can be extended to reduced-width multipliers. By simulations, it is shown that the proposed error analysis method is useful to the practical design of fixed-width modified Booth multipliers.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.2C
• /
• pp.298-305
• /
• 2004

This paper presents an error compensation method for a fixed-width modified Booth multiplier that receives a W-bit input and produces a W-bit product. To efficiently compensate for the quantization error, Booth encoder outputs (not multiplier coefficients) are used for the generation of error compensation bias. The truncated bits are divided into two groups depending upon their effects on the quantization error. Then, different error compensation methods are applied to each group. By simulations, it is shown that quantization error can be reduced up to 50％ by the proposed error compensation method compared with the existing method with approximately the same hardware overhead in the bias generation circuit. It is also shown that the proposed method leads to up to 40％ reduction in area and power consumption of a multiplier compared with the ideal multiplier.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.6 no.2
• /
• pp.199-205
• /
• 2011

This paper presents a QR decomposition processor by exploiting Givens rotation for the GDFE (Generalized Decision Feedback Equalizer). A Givens rotation consists of phase extraction, sine/cosine generation and angle rotation parts. Combining two-stage method (coarse and fine stage) and the fixed-width modified-Booth multiplier, we design an efficient QR decomposition processor. By simulations, it is shown that the proposed QR decomposition processor can be a feasible solution for GDFE.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.6C
• /
• pp.632-639
• /
• 2002

To reduce the area and power consumption in constant coefficient multiplications, the constant coefficient can be encoded using canonic signed digit(CSD) representation. When the partial product terms are added depending on the nonzero bit(1 or -1) positions in the CSD-encoded multiplier, all sign bits are properly extended before the addition takes place. In this paper, to reduce the overhead due to sign extension, a new method is proposed based on the fact that carry propagation in the sign extension part can be controlled such that a desired input bit can be propagated as a carry. Also, a fixed-width multiplier design method suitable for CSD multiplication is proposed. As an application, 43-tap filbert transformer for SSB/BPSK-DS/CDMA is implemented. It is shown that, about 16∼28% adders can be saved by the proposed method compared with the conventional methods.

• Proceedings of the IEEK Conference
• /
• 2003.07b
• /
• pp.867-870
• /
• 2003

본 논문은 멀티미디어 데이터 처리를 위한 효율적인 RISC 프로세서 유닛의 설계를 목표로 Vector 프로세서의 SIMD(Single Instruction Multiple Data) 개념을 바탕으로 고정된 연산기 데이터 비트 수에 비해 상대적으로 작은 비트수의 데이터 연산의 부분 병렬화를 통하여 멀티미디어 데이터 연산의 기본이 되는 곱셈누적(MAC : Multiply and Accumulate) 연산의 성능을 향상 시킨다. 또한 기존의 MMX나 VIS 등과 같은 범용 프로세서들의 부분 병렬화를 위해 전 처리 과정의 필요충분조건인 데이터의 연속성을 위해 서로 다른 길이의 데이터 흑은 비트 수가 작은 멀티미디어의 데이터를 하나의 데이터로 재처리 하는 재정렬 혹은 Packing/Unpacking 과정이 성능 전체적인 성능 저하에 작용하게 되므로 본 논문에서는 기존의 프로세서의 연산기 구조를 재이용하여 병렬 곱셈을 위한 연산기 구조를 구현하고 이를 위한 데이터 정렬 연산 구조를 제안한다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.1
• /
• pp.123-130
• /
• 2016

In this paper, a high-performance arithmetic unit which can efficiently accelerate a number of algorithms for multimedia application was designed. The 3-stage pipelined arithmetic unit can execute 38 operations for complex and fixed-point data by using efficient configuration for four 16-bit by 16-bit multipliers, new sign extension method for carry-save data, and correction constant scheme to eliminate sign-extension in compression operation of multiple partial multiplication results. The arithmetic unit has about 300-MHz operating frequency and about 37,000 gates on 45nm CMOS technology and its estimated performance is 300 MCOPS(Million Complex Operations Per Second). Because the arithmetic unit has high processing rate and supports a number of operations dedicated to various applications, it can be efficiently applicable to multimedia processors.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2018.10a
• /
• pp.190-192
• /
• 2018

This paper describes a design of an elliptic curve cryptography (ECC) processor that supports five pseudo-random curves and five Koblitz curves over binary field defined by the NIST standard. The ECC processor adopts the Lopez-Dahab projective coordinate system so that scalar multiplication is computed with modular multiplier and XORs. A word-based Montgomery multiplier of $32-b{\times}32-b$ was designed to implement ECCs of various key lengths using fixed-size hardware. The hardware operation of the ECC processor was verified by FPGA implementation. The ECC processor synthesized using a 0.18-um CMOS cell library occupies 10,674 gate equivalents (GEs) and 9 Kbits RAM at 100 MHz, and the estimated maximum clock frequency is 154 MHz.

• Journal of IKEEE
• /
• v.23 no.1
• /
• pp.58-67
• /
• 2019

A design of an elliptic curve cryptography (ECC) processor that supports both pseudo-random curves and Koblitz curves over $GF(2^m)$ defined by the NIST standard is described in this paper. A finite field arithmetic circuit based on a word-based Montgomery multiplier was designed to support five key lengths using a datapath of fixed size, as well as to achieve a lightweight hardware implementation. In addition, Lopez-Dahab's coordinate system was adopted to remove the finite field division operation. The ECC processor was implemented in the FPGA verification platform and the hardware operation was verified by Elliptic Curve Diffie-Hellman (ECDH) key exchange protocol operation. The ECC processor that was synthesized with a 180-nm CMOS cell library occupied 10,674 gate equivalents (GEs) and a dual-port RAM of 9 kbits, and the maximum clock frequency was estimated at 154 MHz. The scalar multiplication operation over the 223-bit pseudo-random elliptic curve takes 1,112,221 clock cycles and has a throughput of 32.3 kbps.