• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.9
• /
• pp.49-61
• /
• 1996

Back-propagation(BP) algorithm needs a lot of time to train the artificial neural network (ANN) to get high accuracy level in classification tasks. So there have been extensive researches to process back-propagation algorithm on parallel processors. This paper prsents a linear systolic array which calculates forward-backward propagation of BP algorithm at the same time using effective space-time transformation and PE structure. First, we analyze data flow of forwared and backward propagations and then, represent the BP algorithm into data dapendency graph (DG) which shows parallelism inherent in the BP algorithm. Next, apply space-time transformation on the DG of ANN is turn with orthogonal direction projection. By doing so, we can get a snakelike systolic array. Also we calculate the interval of input for parallel processing, calculate the indices to make the right datas be used at the right PE when forward and bvackward propagations are processed in the same PE. And then verify the correctness of output when forward and backward propagations are executed at the same time. By doing so, the proposed system maximizes parallelism of BP algorithm, minimizes th enumber of PEs. And it reduces the execution time by 2 times through making idle PEs participate in forward-backward propagation at the same time.

• The KIPS Transactions:PartC
• /
• v.12C no.3 s.99
• /
• pp.437-442
• /
• 2005

The measurement of RTT (Round Trip Time) can be used for the analysis of Internet congestion. However, simple measuring of RTT which measures only hun around time of a packet can not infer a packet forward/backward delay variation. In this thesis, we present a new algorithm which can be used for the estimation of forward/backward delay variation of packets. These delay variations are implication of network congestion state. In this algorithm, the reference forward/backward delay can be determined based on the minimum RTT value. The delay variation of each packet can be calculated by comparing reference delay with the packet delay. We verified our proposed algorithm by NS-2 simulation and delay measuring in a real network.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2014.04a
• /
• pp.357-359
• /
• 2014

그룹키란 그룹 내의 사용자들이 암호화와 복호화를 위하여 공유하는 비밀키를 의미하며, 그룹키는 Backward secrecy와 Forward secrecy를 모두 제공하여야 한다. Backward secrecy는 새로운 그룹 사용자가 과거의 데이터를 읽지 못하도록 하는 것이고, Forward secrecy는 탈퇴한 사용자가 이후의 데이터를 읽지 못하도록 하는 것이다. 본 연구에서는 해쉬체인을 사용하여 네트워크 환경에서 Backward secrecy와 Forward secrecy를 효율적으로 제공하는 기법을 제안한다.

• Journal of information and communication convergence engineering
• /
• v.7 no.4
• /
• pp.521-524
• /
• 2009

Park proposed a forward & backward Secure key management scheme in wireless sensor networks for Process Control Systems (PCSs) or Supervisory Control and Data Acquisition (SCADA) systems [7]. The scheme, however, is still vulnerable to an attack called "sandwich attack": two nodes captured at times $t_1$ and $t_2$, respectively, surrenders all the group keys used between times $t_1$ and $t_2$. In this paper, we propose a fix to the scheme, which can limit the vulnerable time duration to an arbitrarily chosen time span while keeping the forward and backward secrecy of the scheme untouched.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2004.05a
• /
• pp.837-842
• /
• 2004

The positions of rotating sound sources have been localized by experiments with the Doppler effects removed. In order to do-Dopplerize the sound signals emitted from moving sources, two kinds of signal reconstruction methods were applied. One is the forward propagation method and the other is the backward propagation method. Forward propagation method analyze the source emission time based on the instantaneous distance between sensors and the assumed source position, then the signals are reconstructed with respect to the emission time. On the other hand, the backward method uses time delay to do-Dopplerize the acquired data for the received time of reference. In both techniques, the reconstructed signal data were processed using beamforming algorithm to produce power distributions at the frequency of interest. Experiments have been carried out for varying frequencies, rotating speeds and the object distances. Forward propagation method has shown better performance in locating source position than the backward propagation method.

• 유체기계공업학회:학술대회논문집
• /
• 2005.12a
• /
• pp.650-654
• /
• 2005

Computational and experimental studies on the forward-sweep inducer for the rocket-engine turbopump are presented in comparison with the conventional backward-sweep inducer. It is shown that back flows at the inlet decreases for forward-sweep inducers. And the low pressure region at the back flow are also decreased, which is assumed to promote the suction performance of the inducers. The backward-sweep inducer shows almost the same suction performance as that of the backward-sweep inducers although it has small inlet tip diameter and shorter length. And the efficiency of the forward-type inducer shows better results than the backward-sweep inducer due to the small size of backflows.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.6
• /
• pp.514-519
• /
• 2009

The projection approximation subspace tracking (PAST) is one of the attractive subspace tracking algorithms, because it estimates the signal subspace adaptively and continuously. Furthermore, the computational complexity is relatively low. However, the algorithm still has room for improvement in the subspace estimation accuracy. FE-PAST (Forward-Backward PAST) is one of the results from the improvement studies. In this paper, we propose a new algorithm to improve the orthogonality of the FB-PAST (Forward-Backward PAST).

• The Journal of Churna Manual Medicine for Spine and Nerves
• /
• v.7 no.1
• /
• pp.35-47
• /
• 2012

Objectives : The purpose of this study is to investigate the clinical application of Rocking forward, Rocking backward Exercise and MET(Muscle Energy Techniques) to three patients who had sway-back posture. Methods : Three patients diagnosed as sway-back posture, not have discogetic pain, were hospitalized Bu-Chun Jaseng hospital of Oriental Medicine about 20 days and they were treated by Rocking forward, Rocking backward Exercise and MET. To measure the outcome of the patients' improvement, we observed the X-ray of Whole spine and Numerical Rating Scale(NRS). Results : After being treated by Rocking forward, Rocking backward Exercise and MET, the patients' posture was significantly improved and the pain was reduced. The X-ray of Whole spine became almost normal and the improvement of patients' in the NRS score was detected. Conclusions : These results suggest that Rocking forward, Rocking backward Exercise and MET were effective treatment on patients who had sway-back posture.

• Korean Journal of Mathematics
• /
• v.6 no.2
• /
• pp.259-279
• /
• 1998

For the forward-backward semimartingale, we can define the backward semimartingale flow which is generated by the backward canonical stochastic differential equation. Therefore, we define the backward self-similar stochastic processes, and we study the backward self-similar stochastic flows through the canonical stochastic differential equations.

• The Journal of Korean Physical Therapy
• /
• v.31 no.5
• /
• pp.292-297
• /
• 2019

Purpose: This study examined the effects of backward walking training with task orientation on the functional walking ability of children with cerebral palsy. Methods: This study was a single-blinded, randomized controlled trial with a crossover design conducted at a single rehabilitation facility with cross-over to the other intervention arm following a two-week break. For a total of 12 children with spastic hemiplegia cerebral palsy, the forward walking training group (n=6) underwent training three times a week for three weeks, 40 minutes a day, and the backward walking training group (n=6) was also trained under the same conditions. To identify the functional walking ability, variables, such as the walking speed, stride length, and step length, were measured using a walk analyzer (OptoGait, Microgate S.r.l, Italy). Results: Both groups showed significant increases in walking speed, stride length, and step length (p<0.01). The backward walking group showed more significant improvement in the walking speed from pre- to post-test (p<0.05). The gait characteristics were similar in the two groups (stride length and step length) but the walking speed in the backward walking group showed a mean difference between the positive effects higher than the forward walking group. Conclusion: Task-oriented backward walking training, which was conducted on the ground, may be a more effective treatment approach for improving the walking functions of spastic hemiplegia children than forward walk training.