Dynamic frequency allocation in impulsive noise environments

The main focus of present discussions about dynamic frequency allocation is centralised around the ability to detect and avoid other users. In this Letter the discussion is broadened to also include the ability to detect and avoid different types of interfering noise, even from unintentional sources. Two types of noise, impulsive and band-limited additive white Gaussian noise are considered and three types of spectrum sensing methods are evaluated. The performance of the spectrum sensing methods, to detect and avoid interference, is evaluated in terms of symbol error rate for an OFDM communication system. The performed simulations show that the amplitude probability distribution method detects harmful interferences best and the method can significantly decrease the symbol error rate compared to pure energy detection.