Large antenna arrays at the transmitter (TX) has recently been shown to achieve remarkable intra-cell interference suppression at low complexity. However, building large arrays in practice, would require the use of power-efficient RF amplifiers, which generally have poor linearity characteristics and hence would require the use of input signals with a very small peak-to-average power ratio (PAPR). In this talk, we consider the single-user Multiple-Input Single-Output (MISO) channel for the case where the TX antennas are constrained to transmit signals having constant envelope (CE). We show that, with per-antenna CE transmission the effective channel seen by the receiver is a SISO AWGN channel with its input constrained to lie in a doughnut-shaped region. For a broad class of fading channels, analysis of the effective doughnut channel shows that under a per-antenna CE input constraint, i) compared to an average-only total TX power constrained MISO channel, the extra total TX power required to achieve a desired information rate is small and bounded, ii) with N TX antennas an O(N) array power gain is achievable, and iii) for a desired information rate, using power-efficient amplifiers with CE inputs would require significantly less total TX power when compared to using highly linear (power-inefficient) amplifiers with high PAPR inputs.
Dr.Saif Khan Mohammed did his B.Tech in Computer Science and Engg from IIT Delhi in 1998. He then joined Philips Semiconductors, Bangalore as Design Engineer in the ASIC group. Later he joined Ishoni Networks, California and worked as Senior Chip Architecture Engineer from 1999 to 2003. In 2003 he came back to India and joined Texas Instruments, Bangalore and worked till 2007 as Design Engineer (Systems and Algorithms Group), in the Wireless Solutions Group. From 2006-2010 he joined IISc Bangalore for his PhD. Later he carried out his post doctoral research for one year at Linkoping University Sweden. Currently he is working as Assistant Professor in the Communication Systems Division, Department of Electrical Engg. of Linkoping University, Sweden. His research interests are: Coding and signal processing for Large MIMO Systems, wireless communications, statistical signal processing, information and coding theory.