No description available.

In this thesis, transmitted reference pulse cluster (TRPC) structure is proposed for low data rate ultra-wideband (UWB) communications to meet the short delay lines and low complexity requirements. TRPC contains closely and uniformly spaced pulses, and makes the analog delay lines short and easy to implement. The performance of TRPC over II B channels is analyzed, simulated and compared to that of three existing techniques: conventional TR. non-coherent pulse position modulation (NC-PPM), and frequency shifted reference (FSR) systems. We demonstrate the performance superiority of TRPC over all those systems. especially when implementation issues such as low complexity are taken into consideration. Moreover, novel schemes are proposed to address the physical layer design issues of the TRPC structure. Based on the theoretical bit error rate (BER) derivation for TRPC, three low complexity integration interval determination schemes requiring no channel state information are developed and simulated. In addition, an adaptive threshold method is developed for the TRPC system. which dynamically calculates the optimal detec¬tion threshold for the auto-correlation receiver, and achieves substantial performance gains over all UWB channel models.

UWB communication is an attractive technology that has the potential to provide low-power, low-complexity, and high-speed communications in short range links. One of the main challenges of the UWB communications is the highly frequency-selective channel, which induces hundreds of overlapped copies of the transmitted pulse with different delays and amplitudes. To collect the energy of these multipath components, coherent Rake receivers are proposed, but suffer from high implementation and computational costs on channel estimation. To avoid the stringent channel estimation, several noncoherent receivers, including energy detector (ED) and transmitted reference (TR), are proposed at the cost of degraded performance. In addition, when taking into account practical issues of UWB communications, e.g., non-Gaussian impulsive noise, non-ideal antennas, and limited, significant performance degradation may be introduced by noncoherent receivers. In this dissertation, we will present low-complexity, high-performance, noncoherent receiver designs for UWB communications that i) avoid the stringent channel estimation; ii) lower the computational complexity of the existing receivers with the aid of advanced digital signal processing techniques; and iii) improve the error performance of the noncoherent receivers by accommodating practical imperfections. First, we propose three multi-symbol detectors (MSDs) for multi-symbol different detection (MSDD), which has recently caught attention in UWB communications because of its high performance without requiring explicit channel estimation. To alleviate the non-deterministic polynomial hardness (NP-hard) of MSDD, we analyze the statistical model of MSDD and propose an iterative MSD and two MSDs based on relaxation technique with near-optimal performance and low complexity. Moreover, the error performance of MSDs is further enhanced by exploiting joint soft-input soft-output MSDD and forward error correction codes. Next, we consider the non-Gaussian noise in the presence of multi-access interference, which is impulsive when the number of active users is small. To mitigate the impulsive noise effect, in this dissertation, we propose new differential UWB receivers based on the generalized Gaussian distribution and Laplace distribution and achieve better error performance. Another main issue of UWB communications is the limited radio coverage. To extend the coverage and improve the performance of UWB systems, we focus on a novel differentially encoded decode-and-forward (DF) non-cooperative relaying scheme. Putting emphasis on the general case of multi-hop relaying, we illustrate a novel algorithm for the joint power allocation and path selection (JPAPS), minimizing an approximate of the overall bit error rate (BER). A simplified scheme is also presented, which reduces the complexity to O(N²) and achieves a negligible performance loss. Finally, we concentrate on code-multiplexing (CM) systems, which have recently drawn attention mainly because they enable noncoherent detection without requiring either a delay component, as in TR, or an analog carrier, as in frequency-shifted reference. In this dissertation, we propose a generalized code-multiplexing (GCM) system based on the formulation of a constrained mixed-integer optimization problem. The GCM extends the concept of existing CM while retaining their simple receiver structure, even offering better BER performance and a higher data rate in the sense that more data symbols can be embedded in each transmitted block. Moreover, the impacts of non-ideal antennas on the GCM systems are investigated given some practical antenna measurement data and IEEE 802.15.4a channel environments.