Exploring new sources of diversity in wireless systems

This paper introduces a new signal processing technique for designing transmit data to improve reliability of communications over wireless fading channels when there is a single transmit and multiple receive antennas using space-time convolutional codes. The original space-time convolutional code design was based on combining transmit diversity and coding to obtain reliability over fading channels when there are multiple transmit antennas. In this paper, we demonstrate that by appropriate transmit data design, we can implement space-time convolutional codes when there is only a single transmit and multiple receive antennas. We achieve this objective by transmitting multiple symbols during a symbol period from one transmit antenna. Multiple symbols are transmitted by multiplying each symbol with a random variable and superimposing them on top of each other. Inducing randomness into the channel is equivalent to creating additional paths, called virtual paths. Although these virtual paths are statistically dependent, by optimizing these induced random variables we can achieve block error rate performance improvement close to that of space-time convolutional codes on systems with multiple physical transmit antennas. Simulation results compare the performance of this technique and that of multiple transmit antenna systems with equal number of receive antennas for QPSK signal constellation when 4 and 16 state space-time convolutional code designs are applied.