Bluetooth Interference Mitigation in 802.11g

We study Bluetooth (BT) interference mitigation in IEEE 802.11g systems. A robust modified minimum mean squared error (MMSE) algorithm is derived for channel estimation. For interference estimation and cancelation, we consider two cases: the interference non-coincident and coincident with the 802.11g subcarriers. For the former case, non-linear least square (NLS) interference estimation and time-domain cancelation is proposed. The unique 11g long preamble structure is exploited to further improve the NLS estimation efficiency. For the latter case when the interference is coincident with 802.11g subcarriers, we propose to use NLS and polynomial smoothing for interference frequency estimation, and interpolation for channel estimation at the subjected subcarriers. The frequency domain interference signal is then obtained by subtracting the re-constructed 802.11g preamble from the received signal. Thereafter, mitigation at the data transmission stage is also performed in frequency domain. The performance of the proposed algorithms is evaluated in terms of channel estimation mean squared error (MSE) and the bit error rate (BER) by simulations for a range of interference powers.