Zhijiang Xu

With the rapid development of Internet of Things (IOT) applications, the performance of existing wireless localization methods under non-line-of-sight (NLOS) transmission environments is seriously challenged. Therefore, this paper proposes a new definition of residual, i.e., the angle residual. The key target of angle residual is to detect line-of-sight (LOS) links from measured and calculated angle parameters. The target position is then estimated by two-step weighted least squares (TS-WLS) algorithm, where only positioning parameters of LOS link are included. Simulation shows that the proposed algorithm performs better than traditional algorithms, especially when the mobile node (MN) position and the NLOS link distribution are randomly generated. Therefore, the proposed method effectively improves the positioning performance and enhances the localization stability in real NLOS transmission environments.

Vibration signal analysis is the most common technique used for mechanical vibration monitoring. By using vibration sensors, the fault prognosis of rotating machinery provides a way to detect possible machine damage at an early stage and prevent property losses by taking appropriate measures. We first propose a digital integrator in frequency domain by combining fast Fourier transform with digital filtering. The velocity and displacement signals are, respectively, obtained from an acceleration signal by means of two digital integrators. We then propose a fast method for the calculation of the envelope spectra and instantaneous frequency by using the spectral properties of the signals. Cepstrum is also introduced in order to detect the unidentifiable periodic signal in the power spectrum. Further, a fault prognosis algorithm is presented by exploiting these spectral analyses. Finally, we design and implement a visualized real-time vibration analyzer on a Raspberry Pi embedded system, where our fault prognosis algorithm is the core algorithm. The real-time signals of acceleration, velocity, displacement of vibration, as well as their corresponding spectra and statistics, are visualized. The developed fault prognosis system has been successfully deployed in a water company.

For secure transmission of low cost single antenna communication nodes in wireless crowdsensing networks under static channel, a physical layer communication scheme is proposed, where each digital modulated symbol is encrypted by a random key at the transmitter and decrypted with the same key at the receiver. The legal users exploit the synchronized chaotic sequence and the two-stage block interleaver to generate a complex random variable (random key), whereby its envelope obeys the Rayleigh distribution and its phase obeys the uniformly distribution. The modulated symbol is multiplied by the complex random variable (encryption) to imitate the Rayleigh fading of the channel at the transmitting end. The received symbol is divided by the identical complex random variable (decryption) to recover the transmitted message before the digital demodulation at the receiving end. Simulation results show that the bit error ratio (BER) performance of the legitimate users is consistent with the theoretical value of the Rayleigh fading channel, while the corresponding BER of the eavesdropper is too high (about 0.5) to intercept any information.

Deploying the long-term evolution (LTE) system in unlicensed spectrum, which is also known as LTE-U, is an appealing technology for future cellular networks. However, the performance of legacy WiFi users will be degraded by the coexistence of LTE-U in the same unlicensed spectrum. In this paper, we investigate the WiFi access probability in a LTE-U and WiFi coexistence scenario where LTE-U users employ the listen- before-talk (LBT) protocols to access the channel. We derive the WiFi access probability expression as a function of the numbers of LTE-U and WiFi users. Numerical results show that the larger LTE-U CW size and offload users from LTE-U to WiFi, the better performance of WiFi access probability can be achieved. Our results provide a significant reference for admission control in LTE-U/WiFi coexistence networks.

The non-line-of-sight (NLOS) error significantly reduces the accuracy of trilateration positioning algorithm in wireless sensor networks (WSN). Therefore, by considering the WSN including both the static nodes (SN) and the mobile nodes (MN), this paper proposes a localization algorithm aiming at the NLOS mitigation, where the NLOS corrupted localization is modeled as a constraint least square (CLS) problem. Among this algorithm, the localization residual is employed to construct the cost function, and a grouping strategy is exploited to further improve the positioning accuracy. The simulations demonstrate that the proposed algorithm can effectively reduce the accuracy loss in scenarios with large NLOS errors or SN numbers.

To achieve secure communication between legitimate users, a physical layer encryption scheme based on random rotation of the modulated symbol is proposed. By exploiting the random behavior and the reciprocity property of the wireless channel, the channel state information (CSI) shared between transmitter and legitimate receivers is used as a initial seed to generate chaotic sequence. The transmitter uses the chaotic sequence to rotate the modulated symbol to enhance communication security and to reduce eavesdroppers' ability to demodulated symbols correctly. Due to the fact that the eavesdropper does not posses any information about the legitimate channel because the channel response is unique to the location of the transmitter and receiver as well as the environment, the receiver is able to demodulate the random rotated symbols correctly while the eavesdroppers demodulate them erroneously. Simulation results show that bit-error-rate (BER) of the legal user matches theoretical results perfectly while the eavesdroppers' BER stays around 0.5, which means that the proposed scheme keeps data transmission under security.