Link enhancement for Multiple Antenna - Orthogonal Frequency Division Multiplexing systems

This work presents new algorithms and techniques for link enhancement of Multiple antenna OFDM systems. Specifically improvements are suggested in the areas of (a) Beam-tilting antennas (b) Pilot Powerloading and (c) Multiple Input Multiple Output (MIMO) Detection and Decoding.;It was shown that the higher directivity of beam-tilting directional antennas can result in significant increases in capacity and considerable reduction in bit error rates. The increased number of directions also translates to a higher training overhead. Techniques to reduce this increased overhead (Exhaustive Training at Reduced Frequency, Statistical and MUSIC techniques etc.) were described and characterized. The use of beam-tilting in interference-limited multiuser networks was characterized. The optimum beam directions were decided by using two metrics - the optimum Signal to Interference Noise Ratio (SINR) and Signal to Leakage Noise Ratio (SLNR) and their relative performances were compared. Finally, it was shown that beam-tilting antennas can be used to "spread" users data in space (akin to spreading the code in frequency in CDMA). This can be used to increase the number of concurrent users communicating to a base station and occupying the same spectrum range.;In the area of pilot design, conventionally when channel state information is available at the transmitter, only the data is precoded and equal power is distributed among all pilots. Also, the proportion of power between data and pilots is arbitrarily set. In this work, we develop a framework where for cases where the channel rank is unity, i.e. for Single Input Multiple Output (SIMO) and Multiple Input Single Output (MISO) systems, it is possible to determine exactly how much power should be invested in each subcarrier. For the MIMO case, it is possible to determine the optimum pilot powerloading once a precoder is fixed. Reduction in bit error rates by using the proposed techniques will be illustrated. Also the effect of feedback delay on the proposed technique will be quantified.;Finally, a technique to perform joint MIMO detection and decoding in systems which use convolutional codes is described. Conventionally in MIMO systems these two blocks (detection and decoding) are performed in two separate modules. It is also possible to iteratively use these two blocks to further reduce bit error rates. In the proposed algorithm it is shown that by appropriately designing the interleaver it is possible to track the state of the convolutional encoder as the tree used for MIMO detection is traversed. By utilizing the knowledge of the state, only the valid modulation points are enumerated to subsequent levels of the tree, resulting in increased sparsity of the search sphere. This results in significant improvement in bit error rates as demonstrated by the simulation results. Furthermore, the architecture and VLSI implementation results for the proposed algorithm are presented.