Consecutive mean excision algorithms in narrowband or short time interference mitigation

Narrowband interference and impulse cancellation are parallel problems in the frequency domain and in the time domain, respectively. This paper presents iterative consecutive mean excision (CME) algorithms that can be used for these purposes. One CME algorithm uses all the data in the initial step whereas a more powerful forward CME (FCME) algorithm uses only a small part of the data. Iteration is shown to be an advantage if compared to noniterative algorithms, but the benefits depend on signals to be cancelled. The CME algorithms also provide a direct way to set the interference detection threshold. The paper discusses the implementation and interference mitigation capability of the CME algorithms. In the simulations their performance is compared to noniterative similar kind of notch filter algorithms in a delay estimation problem. It is shown that these methods offer remarkable interference tolerance improvements especially against very narrowband interference and about 40 dB improvements in tolerable interference-to-signal ratio for interference whose bandwidth is up to 40 % of the bandwidth of the desired signal. It is also shown that averaging is an important element in interference cancellation in delay estimation if the bandwidth of the interference exceeds 10 % of the bandwidth of the desired signal.