الفهرس 
ABSTRACTOnly 14 pages are availabe for public view
 |  | from | 153 |  |  |
| | | from | 153 | | |
Abstract
Signal reconstruction plays a crucial role in rebuilding signals from incomplete information. This thesis proposes a novel technique for transmitting signals by utilizing bandpass spectral information and introduces a new recovery algorithm that reconstructs the baseband signal’s energy from a window of its spectrum. The algorithm’s performance is extensively evaluated for various signal types and data rates, including pulse signals, bit streams, analog signals, and human speech signals, under both noiseless and noisy channel conditions. The results showcase the technique’s ability to efficiently transmit signals over bandlimited channels, requiring less spectral power compared to conventional schemes and enabling bandwidth optimization.
The proposed technique involves transmitting a bandpass window of the signal’s spectrum and employing an adaptive bandpass filter (ABPF) to reconstruct most of the signal’s energy, resulting in significant bandwidth optimization. Promising reconstruction performance is demonstrated for pulse and stream of pulses signals, even with varying data rates. The thesis addresses the sagging problem encountered during the transmission of low-power or small-band segments and proposes two solutions to overcome this challenge. Furthermore, the technique achieves an impressive reconstruction efficiency for speech signals and high bandwidth utilization. Also, the proposed scheme achieves notable improvements in reconstruction performance for pulse signals and streams of pulses across different data rates.
To address the increasing demand for channel resources within limited spectrums driven by wireless technology, the thesis proposes a noise reduction scheme based on a noise reject filter (NRF). This scheme enhances the performance of the baseband signal transmission and reconstruction algorithm by mitigating the impact of additive white Gaussian noise (AWGN), optimizing bandwidth utilization. Additionally, the proposed scheme demonstrates its capability to reconstruct regular data in radio frequency identification (RFID) systems, both in the presence and absence of malware.