Rate and Power Adaptation Over Slow Fading Channels With Noisy Quantized Feedback

The objective of this work is to maximize the expected transmission rate over slow fading channels when quantized channel feedback is available for rate and power control. This problem has been recently studied in the literature assuming noiseless feedback. We consider a more realistic model, where the feedback indices are subject errors that are hardly avoidable in a wireless environment. Our scheme considers a binary symmetric channel (BSC) model for every feedback bit. A channel optimized scaler quantizer (COSQ) is designed to utilize noisy feedback in our system. The impact of noisy quantized feedback is studied when multi-layer coding is employed to combat the uncertainty about the channel state information (CSI) at the transmitter. We also study temporal power adaptation in this system. We show that a noisy feedback system based on COSQ design never performs worse than a no-feedback system and its performance approaches a noiseless feedback scenario as the feedback link quality improves.