• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권7호
• /
• pp.3543-3557
• /
• 2017

Compression is a very important technique for remotely sensed hyperspectral images. The lossless compression based on the recursive least square (RLS), which eliminates hyperspectral images' redundancy using both spatial and spectral correlations, is an extremely powerful tool for this purpose, but the relatively high computational complexity limits its application to time-critical scenarios. In order to improve the computational efficiency of the algorithm, we optimize its serial version and develop a new parallel implementation on graphics processing units (GPUs). Namely, an optimized recursive least square based on optimal number of prediction bands is introduced firstly. Then we use this approach as a case study to illustrate the advantages and potential challenges of applying GPU parallel optimization principles to the considered problem. The proposed parallel method properly exploits the low-level architecture of GPUs and has been carried out using the compute unified device architecture (CUDA). The GPU parallel implementation is compared with the serial implementation on CPU. Experimental results indicate remarkable acceleration factors and real-time performance, while retaining exactly the same bit rate with regard to the serial version of the compressor.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1991년도 한국자동제어학술회의논문집(국제학술편); KOEX, Seoul; 22-24 Oct. 1991
• /
• pp.1785-1789
• /
• 1991

Recently, a number of papers on adaptive control scheme of manipulators are proposed. Slotine and Li[1] showed an adaptive control scheme for robot manipulator. The controller was described in closed form. And later Niemeyer and Slotine discussed about a computational implementation of the controller in recursive form[2]. Walker proposed another adaptive control scheme which can be implemented by a recursive-form controller[4]. Closed-form description is used for the analysis or design of adaptive control systems while recursive-form realization is used for implementation of the controller. The relation between the closed-form realization and the recursive-form one seems to be inadequately referred. Hence, it makes sense to consider the relation between the closed-form description and the recursive-form one. In this paper, first, we make a simple derivation of an closed-form dynamics description of a robot arm from its recursive-form description. And then we derive the closed-form realization of Walker's scheme applied to manipulators having no kinematic loop. We clarify the difference between the Walker's scheme and Slotine's and evaluate the convergence under the controllers.

• 스마트미디어저널
• /
• 제11권9호
• /
• pp.39-46
• /
• 2022

The WK-recursive network proposed by Vecchia and Sanges[1] is widely used in the design and implementation of local area networks and parallel processing architectures. It provides a high degree of regularity and scalability, which conform well to a design and realization of distributed systems involving a large number of computing elements. In this paper, the routing of a message is investigated on the WK-recursive network, which is key to the performance of this network. We present an efficient shortest path algorithm on the WK-recursive network, which is simpler than Chen and Duh[2] in terms of design complexity.

• 전자공학회논문지C
• /
• 제36C권6호
• /
• pp.37-43
• /
• 1999

고성능 VLSI 설계를 위한 pipeline 형태의 Reed-Solomon을 구현하였다. Shortened RS code의 경우에 있어서 기존의 parallel recursive cell 방식이나[1] 다중 클락 설계와 같은 접근과는 달리 작은 면적에서 단일 클락으로 동작할 수 있는 이중 수소(two-way addressing) 방식의 Euclid 계산을 제안하였다. 이러한 방식은 recursive cell을 병렬 처리하는 Euclid 계산 방식에 비해 면적이나 소비 전력에 있어 장점을 갖고 있음을 synthesis와 전력 모의실험을 통해 검증하였다. 본 설계는 면적상으로 parallerl recursive cell을 이용한 단일 클락euclid 회로가 약 5,000 gate임에 비하여 40% 정도 감소한 3,000 gate 정도에 구현할 수 있었다. 또한 전력 소비면으로는 기존의 recursive cell을 이용한 다중 클락 euclid 회로가 6mW 이상의 전력을 소비하는 반면에 본 설계는 3mW대의 전력 소비를 보여 현격한 차이를 보였다.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 1996년도 춘계학술대회 논문집
• /
• pp.419-423
• /
• 1996

These days, industrial robots are required to have high speed and high precision in doing various tasks. Recently, the adaptive control algorithms for those nonlinear robots have been developed. With spatial vector space, these adaptive algorithms including recursive implementation are simply described. Without sensing joint acceleration and computing the inversion of inertia matrix, these algorithms which include P.D. terms and feedforward terms have global tracking convergence. In this paper, the feasibility of the proposed control method is illustrated by applying to 2 DOF SCARA robot in DSP(Digital Signal Processing).

• 대한전자공학회논문지SP
• /
• 제37권4호
• /
• pp.65-71
• /
• 2000

본 논문에서는 SDFT(sliding discrete-Fourier transform)을 순환식(recursive)으로 구현할 때 유한 비트 고정소수점 계산하여 발생하는 오차의 영향을 해석적으로 구하는 방법을 제시하고 이의 유도 과정을 기술하였다. 유한 비트 오차는 계수의 양자화 때문에 발생하는 계수오차와, 곱셈연산 후 반올림되는 유효자리 때문에 발생하는 반올림오차로 구성된다. 각각의 오차는 주파수 스펙트럼 추정오차를 야기 시키며, 이 스펙트럼 오차의 전력과 실제 스펙트럼의 전력 비(noise-to-signal power ratio NSR)를 진동계수를 표현하는 비트 수, DFT 값을 표현하는 비트 수, 그리고 DFT 구간길이에 대한 식으로 유도하였다. 유도과정은 SDFT 순환식(recursive equation)을 통해 유도한 오차방정식(error-dynamic equation)과 계수오차 및 반올림오차의 확률분포특성에 근거하였다. 해석적으로 유도한 NSR 결과를 시뮬레이션 실험을 통해 얻은 결과와 비교하여 타당성을 확인하였다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 추계종합학술대회 논문집
• /
• pp.651-654
• /
• 1999

This paper investigates the varying characteristics of the performance degradation resulting from the different combination of wordlength in fixed-point implementation of recursive sinusoidal transform. The performance degradation is analytically derived in the form of noise-to-signal power ratio. The best wordlength combination is shown to be the equal length distribution of the given number of bits between the transform coefficients and the data. The analysis results are also verified through the computer simulations.

• 한국통신학회:학술대회논문집
• /
• 한국통신학회 2006년도 하계종합학술발표회 논문 초록집
• /
• pp.607-607
• /
• 2006

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제5권12호
• /
• pp.2326-2339
• /
• 2011

To reduce the implementation complexity of the Vertical Bell Labs layered space-time (V-BLAST) systems with respect to the zero-forcing (ZF) criterion, a computationally efficient recursive algorithm is proposed. A fast implementation of the proposed algorithm is developed and its complexity is analyzed in detail. The proposed algorithm matches the ZF-OSIC detection well, and its three significant advantages can be demonstrated by analyses and simulations. Firstly, its speedups over the conventional ZF-OSIC with norm-based ordering, the original fast recursive algorithm (FRA) and the fastest known algorithm (FKA) in the number of flops are 1.58, 2.33 and 1.22, respectively. Secondly, a much simpler implementation than FRA and FKA can be expected. Finally, the storage requirements are lower than those of FRA and FKA. These advantages make the proposed algorithm more efficient and practical.

• 한국음향학회지
• /
• 제18권2호
• /
• pp.61-65
• /
• 1999

본 논문에서는 필터의 계수들을 이용하여 재귀적 3차 Volterra 필터의 안정성에 대한 충분 조건을 유도하고자 한다. Volterra 필터는 기억성이 있는 비선형 시스템을 모델링하는데 있어서 매우 효과적이다. 그러나 비재귀적 Volterra 필터는 비선형 시스템을 모델링하는데 있어서 많은 수의 계수를 필요로 한다는 것이 잘 알려져 있다. 이러한 이유로 비재귀적 구현보다 적은 수의 계수를 요구하는 재귀적 Volterra 필터를 많이 고려한다. 그러나, 재귀적 Volterra 필터의 주된 문제점은 필터에 내재하는 비안정성이다. 본 논문에서는 필터의 입력신호의 크기가 유한한 경우에 있어서 재귀적 이산시간 3차 Volterra 필터의 출력이 유한하게 되는 간단한 조건에 대하여 보고하고자 한다.