ETRI Journal最新文献

In this study, a planar printed circuit board (PCB) coil with FR4 substrate was designed and simulated using the finite element method, and the results were analyzed in the frequency domain. This coil can be used in wireless power transfer (WPT) as a transmitter or receiver, eliminating wires. It can also be used as the receiver in radio frequency energy-harvesting (RF-EH) systems by optimizing the planar PCB coil to convert radio-wave energy into electricity, and it can be employed as an excitation (transmitter) or receiver coil in nuclear magnetic resonance (NMR) spectroscopy. This PCB coil can replace the conventional coil, yielding a reduced occupied volume, a fine-tuned design, reduced weight, and increased efficiency. Based on the calculated gain, power, and electromagnetic and electric field results, this planar PCB coil can be implemented in WPT, NMR spectroscopy, and RF-EH devices with minor changes. In applications such as NMR spectroscopy, it can be used as a transceiver planar PCB coil. In this design, at frequencies of 915 MHz and 40 MHz with 5 mm between coils, we received powers of 287.3 $�\mu$W and 480 $�\mu$W, respectively, which are suitable for an NMR coil or RF-EH system.

Recently, polar code has attracted the attention of many scholars and has been developed as a code technology in coded-cooperative communication. We propose a polar code scheme based on Plotkin structure and quasi-uniform punching (PC-QUP). Then we apply the PC-QUP to coded-cooperative scenario and built to a new coded-cooperative scheme, which is called PCC-QUP scheme. The coded-cooperative scheme based on polar code is studied on the aspects of codeword construction and performance optimization. Further, we apply the proposed schemes to space–time block coding (STBC) to explore the performance of the scheme. Monte Carlo simulation results show that the proposed cooperative PCC-QUP-STBC scheme can obtain a lower bit error ratio (BER) than its corresponding noncooperative scheme.

With an explosive increase of data produced annually, researchers have been attempting to develop solutions for systems that can effectively handle large amounts of data. Single-operating-system (OS) non-uniform memory access (NUMA) abstraction technology is an important technology that ensures the compatibility of single-node programming interfaces across multiple nodes owing to its higher cost efficiency compared with scale-up systems. However, existing technologies have not been successful in optimizing user performance. In this paper, we introduce a single-OS NUMA abstraction technology that ensures full compatibility with the existing OS while improving the performance at both hypervisor and guest levels. Benchmark results show that the proposed technique can improve performance by up to 4.74$�\times$ on average in terms of execution time compared with the existing state-of-the-art open-source technology.

To meet increasing traffic requirements in mobile networks, small base stations (SBSs) are densely deployed, overlapping existing network architecture and increasing system capacity. However, densely deployed SBSs increase energy consumption and interference. Although these problems already exist because of densely deployed SBSs, even more SBSs are needed to meet increasing traffic demands. Hence, base station (BS) switching operations have been used to minimize energy consumption while guaranteeing quality-of-service (QoS) for users. In this study, to optimize energy efficiency, we propose the use of deep reinforcement learning (DRL) to create a BS switching operation strategy with a traffic prediction model. First, a federated long short-term memory (LSTM) model is introduced to predict user traffic demands from user trajectory information. Next, the DRL-based BS switching operation scheme determines the switching operations for the SBSs using the predicted traffic demand. Experimental results confirm that the proposed scheme outperforms existing approaches in terms of energy efficiency, signal-to-interference noise ratio, handover metrics, and prediction performance.

In this paper, we describe a neural network-based application that recommends multiple items using dialog context input and simultaneously outputs a response sentence. Further, we describe a multi-item recommendation by specifying it as a set of clothing recommendations. For this, a multimodal fusion approach that can process both cloth-related text and images is required. We also examine achieving the requirements of downstream models using a pretrained language model. Moreover, we propose a gate-based multimodal fusion and multiprompt learning based on a pretrained language model. Specifically, we propose an automatic evaluation technique to solve the one-to-many mapping problem of multi-item recommendations. A fashion-domain multimodal dataset based on Koreans is constructed and tested. Various experimental environment settings are verified using an automatic evaluation method. The results show that our proposed method can be used to obtain confidence scores for multi-item recommendation results, which is different from traditional accuracy evaluation.

This paper presents methods for reducing the peak-to-average power ratio (PAPR) of the orthogonal time frequency space (OTFS) signal. These methods mainly consist of two operations: symbol interferometry (SI) and either $�\mu$-law or $�A$-law companding. SI spreads the data of one OTFS symbol onto all symbols and is implemented using a simple inverse fast Fourier transform operation on each OTFS symbol. During the second operation, the PAPR of the OTFS signal is significantly reduced. For our performance analysis, the complementary cumulative distribution function, probability density function, and bit error rate are illustrated through simulations performed in MATLAB. The performance is also analyzed using a solid-state power amplifier at the transmitter and compared with OTFS, $�\mu$-law-based OTFS, and SI OTFS systems. The results indicate that the proposed OTFS system achieves a low PAPR.

In this paper, we mathematically investigate a downlink non-orthogonal multiple access (NOMA) system for short-packet communications (SPC) in which the near users are used as full-duplex (FD) relays to forward intended signals from the source to a far user. In addition, partial relay selection is employed to enhance the performance of the FD relays under the impact of imperfect interference cancellation. At the far user, selection combining (SC) or maximal ratio combining (MRC) is employed to combine the signals received from the source and the selected FD relay. The analytical expressions for the average block error rate (BLER) of two users over flat Rayleigh fading channels are derived. Furthermore, closed-form asymptotic expressions of the average BLERs at the near and far users in high signal-to-noise ratio (SNR) regimes are obtained. The numerical results show that the analytical BLERs of the near user and far user closely match the simulation results.

This paper presents the performance evaluation of an amplify-and-forward (AF) relay-assisted multiuser multiple input–multiple output (MU–MIMO) downlink transmission system for correlated fading channels. The overall system performance was improved by incorporating a double-quadrature spatial modulation (DQSM) scheme. The bit error rate (BER) performance and detection complexity of the AF–MU–MIMO–DQSM system were analyzed and compared with those of a conventional AF–MU–MIMO system under the same conditions and parameters. The results showed that the correlated fading channel severely affected the performance of systems with higher spectral efficiency (SE). Considering an SE of 12 bpcu/user, the AF–MU–MIMO–DQSM system yielded a gain of up to 3 dB in BER performance compared with that of its conventional counterpart for the analyzed cases. In terms of detection complexity, the AF–MU–MIMO–DQSM system showed a reduction of up to 56 % compared with that of the conventional system for the optimal maximum likelihood detection criterion.