Search

Gas classification is a machine learning problem that is important for various applications including monitoring systems, health care, public security, etc. Since measuring the characteristic of gas molecules is greatly affected by external factors such as wind speed and the internal setting of detecting sensors, classification should be done by taking into account the combination of these individual factors, which we call a condition in this paper. In particular, when classifying gas data measured under multiple conditions, the data from each condition need to be integrated, which we call multi-conditioned gas classification. While there have been some studies on gas classification for a single condition, no previous approach deals with the multi-conditioned gas classification problem to the best of our knowledge. In this paper, we propose a novel multi-conditioned gas classification method for the first time. We present a new deep learning network structure that can efficiently extract features from the data of multiple conditions and effectively integrate them, which is referred to as a multi-conditioned gas classification network (MCGCN). We also propose a new training loss function to guarantee good performance reliably for the varying number of given conditions. Experimental results demonstrate the superiority of the proposed method, which achieves accuracies of 99.15% ± 0.41 regardless of the number of conditions with 15 times fewer model parameters in comparison to the existing method.

We propose and demonstrate a method to provide precise time synchronization in distributed control systems using a boundary clock scheme. The major drawback of the boundary clock scheme is the exponential accumulation of time synchronization error as the number of hops increases. To make the error accumulation linearly increase with the number of hops, we first separate the frequency compensation interval (FCI) and the offset and frequency compensation interval (OFCI) and then separately optimize each interval. To demonstrate the performance of this method, we implemented test benches using Ethernet-linked distributed control systems. We measured the peak-to-peak jitter performance along with the maximum time interval error to assess the short- and long-term stability after several hops in distributed control systems. Our method enables the peak-to-peak jitter to be maintained under 107 ns after seven hops. The experimental results show that the performance of time synchronization is dominated by fast jitter rather than frequency error and wander, and the proposed scheme can be used to improve the time synchronization performance in IEEE 1588-compliant control systems using boundary clock. [ABSTRACT FROM AUTHOR]

We report on a fiber-optic transport system for next-generation wireless-communication systems utilizing 4 times 4 multiple-input multiple-output (MIMO) orthogonal-frequency-division-multiple-access (OFDMA) technology. Our system supports time-division-duplex (TDD)-based wireless signals operating at 3.775 GHz. To accommodate the TDD-based MIMO signals over a single strand of optical fiber, we utilize nine-channel coarse wavelength-division-multiplexed optical channels: one for link delay measurement and TDD control signal transmission, four for downlink, and the others for uplink. The system first measures the propagation delay between the central base station (CBS) and the remote antenna (RA) and sends the result to delay modules to compensate for the delay added by the transmission link. This procedure makes the CBS and RA emit the downstream signals simultaneously into the air and, consequently, helps avoid the performance degradation caused by the propagation delay of the radio-over-fiber system. The system then sends the MIMO signals together with TDD control signals to the RA. Our experimental demonstration is carried out with 1-Gb/s OFDMA signals having pilot, control, and data channels. For downstream, the error vector magnitudes (EVMs) are measured to be < -30 dB after a 3.9-km transmission over conventional single-mode fiber when the antenna output power is set to be 24 dBm. For upstream, the best EVMs are measured to be < -35 dB. Both the downstream and upstream performance is limited by the shot and thermal noise of the receiver when the signal power is low, whereas nonlinear distortions of electrical amplifiers start to degrade the system performance as the signal power increases. We also measure the crosstalk between channels. It is measured to be less than -42 dB for all channels, which is found to be caused by board-to-board interference at the RA. [ABSTRACT FROM PUBLISHER]