An algorithm for blind symbol rate estimation using second order cyclostationarity

The problem of blind symbol rate estimation of digitally modulated signals is addressed here as a necessary preprocessing stage for modulation classification. We exploit the second order cyclostationarity of linearly modulated signals, with a special focus on a blind scenario with noise and multipath effects. An alternative blind approach using the cyclic domain energy profile is proposed here, to minimize the effects of channel conditions and parameter uncertainty. Compared to the earlier estimators using cyclostationarity, the new scheme shows better performance in the low SNR region with no susceptibility to the pulse shape variation and multipath fading and with a relatively lower computational complexity. The algorithm is extended further to estimate symbol rates of interfering carrier frequencies, which is useful for spectrum sensing applications.