In order to suppress the impulsive noise interference, a joint frequency-domain/time-domain scheme is proposed. In contract to conventional methods, the impact of impulsive noise can be detected by monitoring the instantaneous power of the guard band in the frequency domain; meanwhile the main information of impulsive noise, including burst duration, instantaneous power and arrived time, can be estimated as well. Then a time-domain window function with adaptive parameters, which are decided in terms of the estimated information of the impulsive noise and the current carrier-to-noise ratio (CNR), is employed to suppress the impulsive interference. Simulation results confirm the validity of the proposed scheme, which improves the bit error rate (BER) performance for the ISDB-T receivers in both AWGN channel and Rayleigh fading channel.

For another serious interference for ISDB-T in fast fading channels, the ICI destroys the orthogonality between the subcarriers of OFDM signal, which will cause degradation on system performance. A new technique named compressed sensing is exploited to estimate the channel state information (CSI) with very limited pilots. Thereby, an iterative equalizer of Parallel Interference Cancellation (PIC) is adopted to suppress ICI and compensate the OFDM signal. Computer simulation has been conducted, and it is clearly shown that the proposed method can effectively improve the BER performance of OFDM systems, in comparison with the conventional Least Square (LS) method.

In this talk, we will begin by discussing the how web search works under the hood. In particular, we will see that web search can be framed as a ranking problem and tackled via machine learning methods. Then I will discuss my work on improving the cost-effectiveness of building search engines through the use of semi-supervised learning. Finally, we will briefly discuss other ranking problems, such as those in computational biology and natural language processing, that can be addressed in a similar framework.