Scalable Mismatch Correction for Time-interleaved Analog-to-Digital Converters in OFDM Reception1

Realization of all-digital baseband receiver processing for multi-Gigabit communication requires analog-to-digital converters (ADCs) of sufficient rate and output resolution. A promising architecture for this purpose is the time-interleaved ADC (TI-ADC), in which L “sub-ADCs” are employed in parallel. However, the gain, timing and voltage-offset mismatches between the sub-ADCs, if left uncompensated, lead to error floors in receiver performance. A standard technique for gain and timing mismatch correction is to use L FIR filters, with tap lengths increasing with the mismatch levels. In this paper, we investigate the use of TI-ADCs in OFDM receivers, and provide a scalable technique for mismatch compensation whose complexity is independent of L and the mismatch levels. We achieve this by decomposing the FFT operator that is at the core of the OFDM receiver into eigenmodes, and showing that, even for large values of L and mismatch levels as high as 25%, two eigenmodes suffice to provide an accurate description of the mismatch-perturbed FFT operator. We provide simulation results that show that the associated mismatch compensation algorithm is successful is eliminating the mismatch-induced error floor.