The holy grail of biomedical optical imaging is to perform microscopy deep inside living tissue. Biological tissue scatters light, which prevents the formation of a sharp focus. Wavefront shaping (WFS) makes it possible to focus light through or inside turbid media by spatially shaping the incident wavefront. In this thesis, we tackled some of the most important limitations of the existing WFS methods. We improved the resilience of WFS to measurement noise by introducing a new WFS algorithm, which we called the dual reference algorithm. Our algorithm achieves the maximal interferometric visibility during the measurements, resulting in an optimal signal-to-noise ratio. We showed experimentally that this technique gives us a higher intensity in the focus than the conventional techniques for simultaneous multi-target optimization. Existing WFS algorithms are not optimal if it is known in advance that required corrections should be smooth because they do not use this a priori information. We developed a new WFS algorithm that is optimized for forward scattering materials. In this method, we combined the best of adaptive optics and WFS techniques by introducing a novel scattering and aberration compensation method. Our method, which we call smooth-basis WFS, is noise-resistant and numerically stable, and it can simultaneously find corrections for several targets. We experimentally verified that the smooth-basis WFS outperforms existing WFS techniques, especially in forward scattering samples. Conventionally the focus formed by WFS techniques is scanned using the optical memory effect. We developed a radically new approach, called sparse field focusing (SFF), to scan the generated focus. We experimentally demonstrated that our method achieves a scan range that far exceeds the memory effect range. We presented an analytical model describing the enhancement of the focus, and showed that our experimental results match well with our analytical model. We studied most of the possible reasons affecting the performance of WFS experiments. We reviewed the source of these factors and the possible solutions to mitigate the practical imperfection of the WFS experiments. Utilizing the presented information allows for more effective WFS experiments, optimizing focus enhancement with the lowest number of measurements.