Exploring Different Receiver Structures for Radio over FSO Systems with Signal Dependent Noise

This paper investigates the implications of employing multiple transmissions as a diversity mechanism to counter the turbulence induced fading in satellite systems with optical feeder link and RF user links. Such a technique necessitates optimal estimation of the transmitted signal using the different streams on-board the satellite. Unlike traditional combining techniques like Maximal Ratio Combining (MRC) which assume signal independent noise, the aforementioned estimation is made interesting by the signal dependent noise characteristic of the optical processing in such systems. This paper investigates various estimators in the presence of signal dependent noise and assesses them based on performance and complexity. Towards this, the paper devises a signal model to illustrate the effect of various signal dependent and independent noise sources and derives estimators based on (a) averaging, (b) Minimum Mean Square Error (MMSE), and (c) Maximum Likelihood (ML). The bias and the Mean Squared Error (MSE) of these estimators are analyzed and the impact of signal dependent noise characterized. Based on these results and complexity considerations, the paper presents recommendations for the receiver to be used in future generation of satellite systems with optical feeder links.