http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Toward Deep Learning-based Low Latency Communication in Industrial IoT
Ade Pitra Hermawan,Rizki Rivai Ginanjar,Dong-Seong Kim,Jae-Min Lee 한국통신학회 2019 한국통신학회 학술대회논문집 Vol.2019 No.6
This paper proposes a new direction in order to achieve high throughput and low latency communication in Industrial Internet of Things (IIoT) by utilizing Deep Learning(DL) technique. In order to achieve the goals, congestion in the network shall be avoided. We compare the performance of some DL algorithms in solving network congestion issue. In addition, future research trends regarding to maximize the system performance and to achieve high throughput and low latency communication in IIoT are suggested.
Real-time Data Recovery using Multi-directional LSTM in Wireless Sensor Networks
Ade Pitra Hermawan,Mareska Pratiwi Maharani,Dong-Seong Kim(김동성),Jae-Min Lee(이재민) 한국통신학회 2021 한국통신학회 학술대회논문집 Vol.2021 No.2
In this paper, an algorithm to recover missing data from the sensor in real-time is proposed. Since missing sensor data is a crucial issue in Industrial Internet of Things (IIoT), we employ two different long short-term memory (LSTM) algorithms to handle this issue. The unidirectional LSTM constantly estimates the upcoming data by learning from the previous information, while bidirectional LSTM utilize both past and future information to estimate the missing data. When the system does not receive the data from the sensor devices, the algorithms fill in the missing data based on the predicted value automatically. According to the simulation results, the proposed scheme significantly surpasses the previous works in terms of loss value.
Heterogeneous IoT Sensor Data Classification for Emergency Detection using Machine Learning
Cosmas Ifeanyi Nwakanma,Ade Pitra Hermawan,Jae-Min Lee,Dong Seong Kim 한국통신학회 2021 한국통신학회 학술대회논문집 Vol.2021 No.2
Inertial Measurement Unit (IMU) and Ultra Wide Band (UWB) sensors were integrated to collect heterogeneous data in a smart factory scenario. In this paper, various machine learning algorithms were used to classify the data with a view to detect normal and anomaly situations based on threshold values of the sensor data. System was simulated using keras with GPU 1xTesla K80, 2496 CUDA cores and 12GB GDDR5 VRAM on top of Google colaboratory. Training and testing data were split into 75% and 25% respectively. Classification of the vibration data from the IMU gave Logistic regression (75.9% ), KNN1 (73.37%) and KNN2 (78.4%). In the case of the UWB sensor, KNN1 (100% for movement and respiration), KNN2 (98.60% for movement and 100% for respiration) and Logistic regression gave 100% accuracy both for movement and respiration data. Therefore, it is recommended that based on trade-off, KNN-2 outperformed other machine learning algorithms.
Fast Sensor Data Recovery using Multi-Directional LSTM for Industrial Wireless Sensor
Adinda Riztia Putri,Mareska Pratiwi Maharani,Ade Pitra Hermawan,Jae Min Lee,Dong-Seong Kim 한국통신학회 2021 한국통신학회 학술대회논문집 Vol.2021 No.11
Missing sensor values are common in Wireless Sensor Network due to its unpredictable and fragile nature. While missing sensor values may lead to several serious problems in the industrial environment where the sensors are deployed, it is inevitable at some points. Fast sensor data recovery using multi-directional LSTM is introduced to address this problem where it can handle missing sensor values and replace them with the predicted values from the previous sensor data. Our proposed method shows a promising result of low prediction loss while also achieve fast computation time of RMSE by 0.11, MAE by 0.07, MAPE by 0.01, and computing time by 0.341 ms.