Signal processing for ultra-wideband wireless communications

Abstract

Unlicensed ultra-wideband (UWB) technology has received signi cant research interest for communication, localization and mapping due to its large bandwidth, low power and low complexity Therefore, UWB technology can be considered for short range based applications for next generation wireless networks especially in internet of things and wireless sensor networks. One of the main challenges it faces is the interference from impulsive noise and narrowband communications, since it has low power and huge bandwidth. Further, the Nyquist sampling rate of the UWB signals is very high, which results in costly and complex system design for practical applications. In this thesis, the impact of impulsive noise in UWB wireless communication channels and a novel robust UWB receiver design is investigated that utilizes the received UWB signal cluster sparsity characteristics to mitigate impulsive noise. Further, multiple UWB signal clusters (due to hundreds of multipath) work as a diversity scheme in the proposed receiver design to reduce UWB signal blanking in the presence of impulsive noise. Then the e ect of narrowband interferences (NBIs) on the UWB system is considered, since wideband width UWB signals overlap with high power narrowband wireless communication devices. The sparsity-based NBI mitigation method is proposed that exploits distinct characteristics of UWB signals and NBI. The proposed NBI mitigation method does not require a non-linear operator such as a limiter or a blanker. Improved performance of the proposed UWB receiver has been validated UWB signal transmission in multipath fading channels. A sub-Nyquist rate UWB receiver is designed by exploiting the sparsity of UWB signals to reduce the sampling rate and power consumption of a UWB system. A deterministic (partial) UWB waveform-matched measurement matrix is proposed. The proposed measurement matrix has circulant structure and is sparse in nature. The proposed matrix is easy to implement in hardware and is operationally time e cient as needed in a practical system. Further, bit error rate performance of the corresponding UWB system and the operational time complexity with the proposed measurement matrix are analyzed and compared to the existing measurement matrices for a sub-Nyquist rate receiver design. Next, UWB system for wireless sensor network (WSN) using massive antenna arrays (MAAa) at fusion center (FC) for distributed detection is proposed and analyzed. The coherent and energy based fusion rules are analyzed for the proposed WSN over multiple access channels. The trade-o between performance and implementation complexity of the coherent and energy based fusion is studied. Further, it is shown that MAAs at FC and various level of channel knowledge can enhance the performance of energy-based detector in UWB sensor network with simple system implementation and signal processing requirement. Performance of the proposed UWB sensor network with reference to probability of detection, false alarm, and error are analyzed over standard IEEE 802.15.4a multipath channels and results are validated using simulations. The impact of various design parameters such as the number of sensors, receiver antennas, sensor quality, and integration interval on thesystem performance is also analyzed. Lastly, the impact of various system parameters on the performance metrics such as bit error rate, time of arrival is highlighted and demonstrated. Semi-analytical results are compared with Monte' Carlo simulations to verify the correctness of derived expressions for UWB systems. The robust transmitter and receiver designs proposed in the thesis have been compared with the existing designs are proved to be better than the existing designs, techniques, and algorithms, both analytically and empirically.