In amplify forward cooperative relaying, which strategy performs better : CSI assisted amplification or Fixed gain amplification? Why? Is it a thumb rule or behavior may vary for communication systems?
More specifically, assuming a dual-hop AF relay system without a direct link, Fig. 2 of [1] compares the outage probability performance of a dual-hop nonregenerative system employing a fixed gain relay with that of an equivalent (in terms of average
introduced gain) system with a varible gain relay. For the medium to large average
SNR region, systems with variable gain relays outperform those with fixed gain relays. However, the surprising and interesting result is that the gap in performance is not as much as one would expect in comparison with the difference in
implementation and complexity of both relays. Note also that systems with fixed gain relays can even slightly outperform systems with variable gain relays at low average SNR. This is due to the fact that the variable gain relay has a gain floor
when alpha_1 is too small, which is a relatively frequent event in the low average SNR region. Finally, note that as is increased, the range of average SNR in which fixed gain relays outperform variable gain relays extends to the right. One should bear in mind, however, that the performance of fixed gain relays shown in the figure can be thought of as an upper bound on the performance of practical relays. This due to the
fact that fixed gain relays suffer from saturation that degrades the performance of these relays at low average SNR ranges.
[1] M. O. Hasna and M.-S. Alouni, "A performance study of dual-hop transmissions with fixed gain relays", IEEE Trans. Wireless Commun., vol. 3, no. 6, pp. 1963-1968, Nov. 2004
I agree with Daniel in his answer. Another important point is the complexity of the two types. More channel estimations are requred in the CSI-assisted relaying scheme where the relay gain is depending on the instentaneous first hop channel. In contrast, the fixed gain relay could be of two types: blind and semi-blind. In the blind fixed gain relaying, the relay gain is completely independent of the first hop channel, whereas in the semi-blind fixed gain relay, the gain is a function of the average value of first hop channel. Thanks