Based on the overall vision of the Sixth Generation Mobile Communications (6G) and the different dimensions of the sub-vision, the article outlines key technology trends and challenges that are likely to achieve these wonderful visions. This article is based on new basic technologies such as new composites, meta-materials, non-orthogonal waveforms, high-frequency millimeter and terahertz waves generation, to network-based technologies such as wireless new Internet infrastructure and deep-sea communication networks. Emerging application technologies such as perceptual interconnection and human body communication are introduced in different categories to select key technologies that are full of opportunities and challenges in the 6G era. This article cites the relevant content and research results of some of the top international academic journals, and strives to be as rigorous and reasonable as possible, with a view to attracting industry attention to the development trend of 6G mobile communications.

The timely and efficient processing of alarm information in the power backbone communication network can provide an opportunity for fault handling. By applying artificial intelligence technology based on historical data, the occurrence of alarms can be predicted. A complete data set is the basic condition for successful prediction. The data collected in the real network often has problems such as uneven distribution and missing data. How to construct a complete data set is a key issue based on artificial intelligence technology. This paper proposes a data preprocessing method for alarm prediction in power backbone communication network. Data enhancement technology is used to complete the class balance of data, so that the final data set conforms to the data size input by the model. The simulation results show that the data pre-processing method based on data enhancement can improve the accuracy of alarm prediction to a large extent.

This paper proposes and demonstrates an Optical Signal Noise Ratio (OSNR) monitoring scheme using the signal power nonlinear transformation and Deep Neural Networks (DNN). The features of signal after 2th, 4th, and 8th transformation and corresponding Fast Fourier Transformation (FFT) depend on the OSNR of signal. By utilizing the DNN to extract those OSNR depended specific features, the OSNR value can be estimated. Simulation results for 28 Gbaud Polarization Division Multiplexing (PDM)-Quadrature Phase Shift Keying (QPSK), PDM-8 Phase Shift Keying (PSK), PDM-8 Quadrature Amplitude Modulation (QAM) and PDM-16QAM signals show that the OSNR monitoring with mean estimation standard errors of 0.10, 0.09, 0.33 and 0.46 dB in back-to-back case and 0.43, 0.34, 0.66 and 0.79 dB in 2 000, 1 040, 1 040 and 800 km single mode fiber transmission case with input optical power of 4, 4, 3 and 3 dBm, respectively.

In Complementary Metal Oxide Semiconductor (CMOS) camera-based visible light communication system using the rolling shutter effect, the number of pixel columns representing one bit will decrease and the fluctuation of grayscale values became serious with the increasing of system bit rate, which leads to the degradation of demodulation performance. A novel demodulation algorithm based on global and local dynamic thresholds and segmented downsampling is proposed. The global dynamic threshold is fitted by grayscale values of all columns in an image, and frame headers contained in the image are located. Then the local dynamic thresholds are fitted respectively by the grayscale values in each data frame. Finally, the result of grayscale decision is segmented downsampling to recover the transmitted bit sequence. The experimental results show that when the column scanning frequency of the rolling shutter is 103.68 kHz, the algorithm is feasible with 2.962 pixels/bit, where the maximum bit rate and net data rate can reach 35 and 10.8 kbit/s, respectively.

Aiming at the problem that the current quantum random number generation technology can not meet the safety and high rate requirement in the communication field, we propose an experimental scheme based on measuring the vacuum fluctuations of the light field to produce quantum true random numbers in this paper. Different from the experimental schemes reported in the past, we theoretically analyze the influence of the gain of the local oscillator on the relative quantum entropy content of the original random numbers in the quantum orthogonal component measurement of quantum random number generation system. The quantum condition minimum entropy is used to quantify the randomness of the original random numbers under the worst case assumption that the classical noise is completely controlled by the eavesdropper. Based on the theoretical analysis, the vacuum noise fluctuations are experimentally amplified by relatively increasing the local oscillator intensity while controlling the electronic gain and the classical electronic noise. The minimum entropy content introduced by the quantum noise in the system is improved. At the same time, the vacuum quantum noise is broadband Gaussian white noise, which effectively increases the bandwidth of extractor frequency-band, as well as the utilization rate of quantum entropy source and the security of the quantum random number generation system while increasing the rate of quantum random number generation. The results show that the Toeplitz-hash extractor based on the security information theory is provable. The quantum random number generation of 6.7Gbit/s is realized. The three standard tests for random numbers of Nist, Diehard and TestU01-SmallCrush are used to verify the true randomness of the quantum random numbers generated in this scheme, which provides a new way to increase the rate of generation of vacuum quantum random number generators.

Quantum Key Distribution (QKD) is considered as a secure solution to provide secret keys for the security of optical networks. In order to meet the security demands of fields like finance and military, the integration of QKD into optical networks has attracted much attention. At present, some studies have focused on the integration of QKD in optical networks, while researches on the survivability of optical networks based on QKD is still in a blank. Aiming at the survivability of optical networks based on QKD, this paper proposes protection schemes for key provisioning service, and evaluates the network performances through simulation. The results verify that the proposed method has better network blocking probability, resource utilization and key consumption rate. The proposed scheme provides an important advice for the development of the survivability of optical networks based on QKD.

In view of the delay problem of optical transport network, this paper analyzes the factors caused by the delay of optical transport network, and elaborates the proportion of the factors that introduce delays according to the two scenarios of long-haul backbone network and local network. For these two scenarios, this paper proposes strategies to realize low-latency optical transport networks, which provides a valuable reference for optimization of low-latency optical transmission networks.

In order to overcome the Erbium Doped Fiber Amplifier (EDFA) gain fluctuations when monitoring long-distance multi-span fiber links with EDFA for existing Optical Time Domain Reflectometry (OTDR), an additional fill light is needed to maintain power stability. In this paper, based on linear frequency modulated optical time domain reflectometry, a long-distance multi-span fiber monitoring technology that does not require supplemental light to pass through EDFA is proposed. This technique utilizes chirped quasi-continuous light as the probe light to avoid fluctuations in the EDFA gain. At the same time, the fractional-domain Fourier filtering is used to remove the spontaneous emission white noise accumulated in the cascaded EDFA, thereby improving the signal-to-noise ratio of the received signal. The experimental results show that the technology can accurately detect the reflection events at the EDFA and the connection points in the long-distance multi-span fiber link.

In term of the cross-sensitivity problem of ambient temperature in the process, a double-parameter sensor based on Twin-Core Fibers (TCF) and Long Period Gratings (LPG) is proposed. The theoretical mode is established, and the stimulated analysis is performed. The simulation results show that the bending sensitivity of 0.205 nm/cm with the micro-bending range of 20~60 cm and the temperature sensitivity of 1.208 nm/℃ with the temperature range of 0~100 ℃ are achieved. Based on a spectrometer having a wavelength distinguishability of 0.1 nm, the sensor can reach a distinguishability of 0.2 cm for bending and 0.079 ℃ for temperature. This sensor has advantages of simple structure and high sensitivity, and can effectively eliminate interference of ambient temperature which is available for shape monitoring of buildings such as tracks and bridges.

With the rapid development of optical communication technology, the requirement of the size and performance of the modulator are higher. Therefore, an electro-optic modulator based on Aubry-André-Harper (AAH) resonator is proposed in this paper. The device is based on silicon and etched circular air holes to form triangular photonic crystal structure. AAH resonator is introduced into the line defect of photonic crystal. The p-n junction is formed by doping the line defect. According to the carrier dispersion effect of silicon, the refractive index of the substrate plate is changed by changing the modulation voltage, which results in the displacement of the resonant wavelength and achieves the modulation purpose. The Lumerical simulation software is used to analyze the device performance. The simulation results show that the 3 dB bandwidth is 1.01 nm on the device with 9.03 μm length. When the refractive index difference reaches 0.01, Transverse Electic (TE) mode narrow bandwidth modulation with 1 550 nm wavelength can be realized. The insertion loss of the device is 1.33 dB and the extinction ratio is 29.13 dB. The process error has little effect on the device performance. The electro-optic modulator is easy to integrate, small in size and has high extinction ratio, which is of great significance for the development of integrated optical devices.

Since each user's demodulation reference signals are superimposed and interfere with each other in the case where multiple users simultaneously transmit signals at the same frequency, it is difficult to accurately estimate the channel impulse response for each user. In this paper, a special signal matrix is designed for the transmission format 4 of the 5G system Physical Uplink Control Channel (PUCCH) by using the characteristics of demodulation reference signals of different users. Then we combine this signal matrix with traditional single-user channel estimation algorithms to propose three channel estimation algorithms including Least Squares (LS)-based spline interpolation, LS-based Discrete Cosine Transform (DCT) interpolation algorithm, and Liner Minimum Mean Squared Error (LMMSE)-based DCT interpolation algorithm. Finally, the scenes of two users and four users are simulated. The simulation results show that three proposed channel estimation algorithms are applicable to the format 4 of 5G system PUCCH, which has great reference value for engineering applications.

In order to achieve fingerprint location based on the strength of low energy bluetooth received signal, a new fingerprint location method is proposed, which does not need to collect too many data points to establish fingerprint database and can correct the positioning error in time. In the off-line phase, we use the low rank matrix completion theory to recover the fingerprint database with only a small number of acquisition points to reduce the workload. In the online stage, after clustering and matching the fingerprint database, the localization problem is transformed into compressed sensing model to improve positioning accuracy. Experimental results show that the proposed positioning algorithm can effectively reduce about 40% workload of early data acquisition and reduce the positioning error from 0.93 m to 0.71 m.

Pilot sequences are inevitably multiplexed in different cells, resulting in pilot pollutions. The interference caused by pilot pollution does not decrease as the number of base station antennas increases. Pilot decontamination is one of the main reasons, which limits the performance of massive Multiple-Input Multiple-Output (MIMO) systems. To reduce the effect of pilot pollution, pilot assignment scheme with time-shift joint space partitioning is proposed, which is based on time and space domains. The pilot decontamination can be mitigated by reducing the number of reused times of the pilot sequence in two directions. We also derive the achievable uplink users access rate and Signal-Interference-Noise Ratio (SINR)in the system. The simulation results show that the proposed scheme can effectively reduce pilot pollution and improve the system performance.