Quantized-Feedback Optimal Adaptive Beamforming for FDD Systems

A new approach to adaptive beamforming is developed for frequency division duplex (FDD) systems. Since the method adapts the transmit beamforming vector by computing the optimal direction of adaptation and feeds its quantized version back to the transmitter, we call it the quantized-feedback optimal adaptive (QFOA) beamforming algorithm. This algorithm offers a very fast-convergence and highly adaptive diversity technique. Fast convergence improves the system performance during the start-up period and helps remove the need to continuously tracking the antenna weights during intermittent periods where the transmitter does not transmit to a particular mobile user, or while the connection is temporarily suspended or idle. High convergence speed also provides a fast tracking capability demanded by high mobility applications. We also propose several quantization methods and prove analytically that they guarantee global convergence, irrespective of quantization error, since the gradient is always positive at any nonstationary point in the solution space.