One of the most important issues for magnetic-confinement fusion research is the so-called anomalous transport across magnetic field lines, i.e. transport that is in excess of that caused by collisional processes. The need to reduce anomalous transport in order to increase the efficiency of a prospective fusion reactor must be addressed through an investigation of its fundamental underlying causes. This thesis is divided into two distinct components: one experimental and instrumental, and the other theoretical and based on numerical modeling. The experimental part consists of the design and installation of a new diagnostic for core turbulence fluctuations in the TCV tokamak. An extensive conceptual investigation of a number of possible solutions, including Beam Emission Spectroscopy, Reflectometry, Cross Polarization, Collective Scattering and different Imaging techniques, was carried out at first. A number of criteria, such as difficulties in data interpretation, costs, variety of physics issues that could be addressed and expected performance, were used to compare the different techniques for specific application to the TCV tokamak. The expected signal to noise ratio and the required sampling frequency for TCV were estimated on the basis of a large number of linear, local gyrokinetic simulations of plasma fluctuations. This work led to the choice of a Zernike phase contrast imaging system in a tangential launching configuration. The diagnostic was specifically designed to provide information on turbulence features up to now unknown. In particular, it is characterized by an outstanding spatial resolution and by the capability to measure a very broad range of fluctuations, from ion to electron Larmor radius scales, thus covering the major part of the instabilities expected to be at play in TCV. The spectrum accessible covers the wavenumber region from 0.9 cm-1 to 60 cm-1 at 24 radial positions with 3 MHz bandwidth. The diagnostic is an imaging technique and is therefore also well suited to investigate inhomogeneous spatial regions, where the need for an excellent spatial resolution is greatest. Additionally, it was also designed as translatable to broaden the region of study, which can extend up to the magnetic axis, in selected configurations. The translatable design combined with the flexibility of TCV in terms of plasma positioning in the vacuum vessel allows the phase contrast system to measure fluctuations across virtually the whole plasma minor radius. The diagnostic is sensitive both to radial and poloidal wave numbers, depending on the configuration. A parallel project to the development and installation of the phase contrast imaging system was the installation of a prototype Doppler reflectometer operating in a homodyne configuration, both in X and in O mode polarization. The reflectometer was operated parasitically to assess its performance which proved to be excellent; it is now routinely available on TCV. The theoretical part of the thesis consisted of extensive modeling of the effect of plasma shape, in particular triangularity, on turbulent transport by means of linear and nonlinear gyrokinetic simulations. This was motivated by experiments on TCV that had shown a dramatic improvement in confinement, up to a factor of two, in inverting the sign of the triangularity from positive to negative. Negative triangularity was indeed found to have a stabilizing influence on ion scale instabilities, specifically on the so called trapped electron mode (TEM). Simulations were carried out on actual TCV shots and the variation of the heat flux with triangularity calculated by the nonlinear simulations is in fair agreement with the experimental results. Linear simulations and a simple analytical model explain, in agreement with nonlinear runs, the resulting stabilization as a result of a rather complex modification of the toroidal precessional drift of trapped particles exerted by negative triangularity.