Frame Detection of OFDM System with Periodic Pattern Preamble

This paper deals with frame detection of OFDM system with periodic pattern preamble transmitted over frequency-selective channels. The systems utilizing such periodic pattern preamble include IEEE 802.11a/g/n and IEEE 802.16, etc. By applying Log-likelihood ratio testing, two different approaches to considering statistical characteristics of channels have been used to derive two simpler frame detectors. Both detectors and their approximation at high signal noise ratio are insensitive to frequency offset and the latter is more computationally effective. An approximated method to analyze statistical characteristics of detector is derived. Neyman-Person rule is used to obtain detection decision threshold. Finally, computer simulations are conducted to evaluate the detection performance of presented detectors.