Hardware Optimized Cell Averaging Estimation (HO-CAE) for Threshold Based Signal Detection

As the number of active wireless devices increases, more of the electromagnetic spectrum is allocated to the communication networks that support them, leading to more spec-trum congestion and interference. Congestion and interference can vary with time and location providing an opportunity for dynamic spectrum access and reactive spectrum avoidance. To utilize dynamic spectrum sensing, a cognitive radar system needs to be able to sense the spectrum, identify available bands and select the next waveform within a single pulse repetition interval (PRI). Given the short duration of a PRI and the dynamic way communications networks access the spectrum, sensing must utilize computationally efficient techniques that lend themselves to real-time operation. This work proposes a method for calculating an energy threshold that can determine if any given frequency in a spectrum snapshot is occupied. Hardware optimized cell averaging estimation (HO-CAE) uses averaging windows to estimate the noise power of the spectrum. U sing a noise power estimate, a threshold can be calculated to determine if a signal is present in each bin of the power spectrum.