Abstract
Considerable works on adaptive schemes for transmit beamforming in distributed networks have emerged in the past years, where a noise-free received signal strength (RSS) measurement and a static environment were often considered. In practical environments, however, system uncertainties may rise and the aforementioned ideal assumptions may fail in these settings. Therefore, we focus on robust designs in this thesis and proposed a systematic analytical framework on the convergence of a general set of adaptive schemes under the condition that measurement of RSS at the receiver side is corrupted by noise. In addition, for time-varying channel and time-varying network topology, we defined a set of robustness criteria that can be used as comparison metrics for existing adaptive schemes. By utilizing the proposed analytical frameworks and metrics, we develop an bio-inspired scheme, BioRARSA2, that possess significantly superior robustness with respect to environmental variations and system uncertainties, where the improved robustness is further validated through extensive numerical simulations.