Digital transform for a selective channel estimation (application to multicarrier data transmission)

Multicarrier modulations have proven they can handle frequency selective transmission channels efficiently. Multicarrier transmissions (also called parallel techniques) are principally developed for digital broadcasting (audio and TV) and digitizing telephone subscriber loops (asynchronous digital subscriber loop-ADSL techniques). This paper generalizes the DMT-OFDM digital modulations (discrete multitone-orthogonal frequency division multiplex), from the usual use of the Fourier transform to other digital transforms manipulating exclusively real numbers or, when possible, bits. We particularly pay attention to Hadamard transforms, due to their simple implementation. The performances of our system (called HDM, for Hadamard division multiplex) are presented for the urban mobile radio channel. This channel is estimated by an interpolating method based on digital transforms (the usual Fourier transform and, less common, Hadamard transform). For equivalent performances (in terms of bit error rate (BER) and spectral efficiency), the main advantage of the Hadamard approach is its very simple implementation, allowing cheap mass production.