The Sparse Approximations of the LMMSE Channel Estimation in OFDM with Transmitter Diversity

The main goals in developing the next generation of wireless communication systems are for delivering multimedia services such as voice, data and image in local coverage networks. These will be a complement to the existing wide area coverage systems, for example to the third generation of mobile communications. In order to provide these services, a high data rate and high quality digital communication system is required in a restricted bandwidth. A major limiting factor is, however, the multipath propagation phenomenon. It causes frequency-selective fading due to different echoes of transmitted symbols overlapping at the receiving end, which can lead to the bit-error-rate (BER) degradation. One way to effectively combat the multipath channel impairments and still provide high-data rates in a limited bandwidth is use of an orthogonal frequency-division multiplexing (OFDM) modulation method and multiple antennas at the transmitting end. OFDM is an attractive multi-carrier modulation (MCM) technique because of its high spectral efficiency and simple single-tap equalizer structure, as it splits the entire bandwidth into a number of overlapping narrow band sub channels requiring lower symbol rates. Furthermore, the inter-symbol interference (ISI) and inter carrier interference (ICI) can be easily eliminated by inserting a cyclic prefix (CP) in front of each transmitted OFDM block. OFDM can be implemented using a coherent or non-coherent detection technique. A coherent detection method generally provides the SNR gain over the non-coherent method as the former modulation technique uses channel state information. This implies, however, a more complex receiver as the channel state information is generally obtained using channel estimation.