Application of DFT-based channel estimation for accurate signal cancellation in Cloud-Txn multi-layer broadcasting system

Cloud transmission (Cloud-Txn) with layer-division-multiplexing (LDM) was proposed as a candidate Physical Layer (PHY) technology for the next generation digital TV broadcasting system. In this paper, we investigate the performance of the upper-layer signal cancellation with DFT-based channel estimation techniques, due to their capability of providing near-optimal performance for challenging channels with long delay spread at low Signal to Noise Ratio (SNR). Both Pilot-Aided (PA) and Decision-Directed (DD) techniques are investigated and their performances are compared. The major issue with DFT-based techniques is the “edge effect” for OFDM systems with Null spectrum on either side of the useful spectrum, which serves as guard band. A Cloud-Txn/LDM system usually requires very accurate channel estimation to achieve very low residual upper-layer signal after the cancellation. In this case, the “edge effect” becomes a significant impairment. In this paper, we evaluate different data-padding methods to reduce the “edge effect”, which fill the Null spectrum with Virtual Pilots (VP). It is shown that, with proper data-padding technique, 30 to 40 dB signal cancellation can be achieved.