Recently, the wireless local area networks (WLANs) in complex environments, such as underground, underwater, and indoor environments, have attracted considerable attention for both commercial and military applications. However, existing communication techniques based on electro-magnetic (EM) and acoustic signals in these environments suffer from varies of challenges, such as high absorption, long delay, limited frequency bands, and signal blockage. In this thesis, we propose to use reconfigurable magnetic-induction (MI) techniques and reconfigurable smart reflect-arrays to address aforementioned problems. The theoretical model, system design, and control algorithms are presented. The proposed systems are validated by a series of in-lab experiments.