This paper presents a high performance and hardware efficient packet detection structure, which employs a cross correlator for the M-sample time delayed correlation operation and a signal power calculator using the received input samples less than or equal to a zero-padded suffix of length M. We investigate the detailed characteristics of the proposed packet detector. In this paper, the performance of a class of packet detection algorithms in the ultra-wideband (UWB) channel environments is also studied. The best packet detection algorithm for the multi-band orthogonal frequency division multiplexing (MB-OFDM) UWB transmission is determined through analysis and extensive simulations. The results of analysis show that the proposed packet detection structure has advantages in the hardware complexity as well as performance when compared with the existing packet detection structures. In order to effectively conduct the packet detection before the automatic gain control (AGC) mode, we investigate the effects of both a frequency offset and the initial gain level of a variable gain amplifier (VGA) on the performance of the packet detection. We also suggest a VGA gain control technique to enhance the performance of packet detection.

A packet detection method for zero-padded orthogonal frequency division multiplexing (OFDM) transmission is presented. The proposed algorithm effectively conducts packet detection by employing both an M-sample time delayed cross correlation value, and a received signal power calculated by using the received input samples corresponding to the zero padding (ZP) intervals or less.

In this letter, we propose a novel packet detection and symbol timing synchronization algorithm for MB-OFDM UWB system over multi-path channel. To find more reliable and exact symbol timing, we utilize the merits of the auto- and cross-correlation at the same time. The auto-correlator can gather the energy of the received signal which is distorted by multipath fading. And, the cross-correlator can estimate symbol timing to sample level precision. We define the detection and symbol timing metric as the moving sum of autocorrelation of the pre-computed cross-correlated signals over zero pad duration. Simulation results show that the proposed algorithm can detect the presence of a signal with less than -4 dB SNR and find DFT window with at most 0.1 dB SNR reduction upon the receipt of two symbols.