For beyond 5G wireless communication systems, inband full-duplex (IBFD) is one of the important candidate technologies for realizing high capacity and low latency transmissions. However, in an IBFD system, the base station (BS) can perform uplink (UL) and downlink (DL) processes simultaneously, and hence there exists severe self-interference (SI) produced by the simultaneous UL and DL communications in the system. Therefore, to realize the IBFD system, it is necessary to deal with the SI problem. To cancel SI, adaptive filter is one of the possible candidates, which can be implemented with low cost and complexity. In our previous study, we proposed a digital SI cancellation method for IBFD system by using adaptive filter. To verify the performance of the method in our previous study, we implemented an IBFD system with 3-stage of SI cancellation: antenna, analog circuit, and digital cancellation. From the experiment results, it can be seen that our implemented system can finally offer significant SI cancellation performance, so that the demodulated output can reach $10^{-2}$ order bit error rate (BER).
This paper studies the trajectory optimization of an automated guided vehicle (AGV) for a smart factory where material handling with high wireless communication quality is required. This design aims to maximize the sum of spectral efficiency under complex continuous-value constraints on movement and combinatorial constraints on cooperative movements. To tackle the NP-hard optimization problem, this paper proposes a hybrid algorithm that combines global and local optimization, where global optimization meets the constraints with relaxation and local optimization considers all constraints on movements. Specifically, the AGVs travel by local path planning based on dynamic window approach (DWA), given the globally optimal trajectory that is solved using dynamic programming (DP). Finally, simulation results provide the high communication quality achieved by the proposed algorithm compared to the benchmark algorithms.
Traditional methods of coffee bean inspection mainly rely on manual labor, which results in issues such as low efficiency and subjectivity. However, with the advancement of computer vision and machine learning technologies, automated coffee bean inspection has become possible. Therefore, this study aims to develop a system that can distinguish between good and bad coffee beans using two main components: YOLOv7 and convolutional neural network (CNN). The image recognition model in this study is divided into three categories: broken, insect-infested, and mold, all of which employ transfer learning. Using YOLOv7, the coffee beans are easily recognized and processed by the image classification model Then, the captured coffee beans are used as the input for the image classification model. If the output results are all negative, it means that the bean is good, and it will be kept. However, if there is at least one output indicating a defect, the bean will be labeled as the corresponding defect type. In the end, using the DenseNet201 model, we achieve an accuracy of 98.97% in classify defective coffee beans.
This paper presents the physical-layer cell identity (PCID) detection probability in the presence of a large carrier frequency offset (CFO) including a large Doppler shift and phase noise from a local oscillator for the 5G New Radio (NR) terrestrial network (TN) and non-terrestrial network (NTN) in the millimeter-wave bands. We employ the same PCID detection method with CFO estimation for both the TN and NTN, which employs cyclic prefix (CP)-based fractional frequency offset (FFO) estimation and joint integer frequency offset (IFO) estimation with the primary synchronization signal sequence (PSS). Link-level simulation results show that the CFO due to the frequency stability of a local oscillator of a set of user equipment of up to 20 ppm and a large Doppler shift resulting from satellite movement are suppressed to a low level employing the same PCID detection method with CFO compensation. We also show that the impact of the time-varying phase noise on the PCID detection probability is slight. Accordingly, the high PCID detection probabilities of approximately 95% and 99% are achieved at the average received signal-to-noise ratio (SNR) of approximately 0dB and -4dB for the Tapped Delay Line (TDL)-C and NTN-TDL-A channel models in a non-line-of-sight (NLOS) environment, and the NTN-TDL-C channel model in a LOS environment, respectively.
This paper proposes estimation methods for carrier frequency offset (CFO) and channel response using synchronization signals for the Physical Broad-cast Channel (PBCH) of the 3GPP New Radio (NR) specifications. In the proposed methods, we estimate and compensate the CFO of the received PBCH by employing the primary synchronization signal (PSS) followed by estimation of the channel response using the demodulation reference signal (DMRS) and the secondary synchronization signal (SSS). Simulation results show that the proposed CFO estimation method achieves almost the identical average block error rate (BLER) for the PBCH as that without CFO. We also show that the channel estimation method using the DMRS and SSS decreases the required average received SNR at the average BLER of $10^{-3}$ by approximately 0.4 dB compared to that with only the DMRS. Finally, we show the effect of successive cancellation (SC) decoding and cyclic redundancy check (CRC)-aided successive cancellation list (CA-SCL) decoding on the BLER of the PBCH with a low polar coding rate.
