Sensing OFDM Signals using periodically inserted pilots

A common feature in OFDM is that some pilot subcarriers repeat in certain OFDM blocks. In this paper we propose sensing methods using the repetition structure of the pilots. The derivation of the methods is based on the optimal likelihood ratio test (LRT) after taking the auto-correlation. Although the derived LRT is optimal, it cannot be used in practice as it needs some unknown information of the channel and noise. To make the methods practical, we simplify the LRT to obtain the approximated LRT (ALRT). Carrier frequency offset (CFO) is another major obstacle to the proposed methods and also to many other methods. To handle the problem, we propose a method to estimate the composite CFO and compensate the composite CFO in time domain. The frequency offset estimation method uses multiple taps of auto-correlation of the signal, which can be used for other applications as well. The proposed methods are robust to frequency offset, noise power uncertainty, frequency selective channel, time delay uncertainty, and interferences.