Operational practices of both power transmission and distribution grids are impacted by the increasing connection of renewable energy resources, electric vehicles and energy storage systems. A timely and accurate knowledge of the system state, coupled with a level of automation able to protect or control the system, enables a better operation of the electrical grids. In this respect, synchrophasor networks, might provide system operators with wide-area, synchronized, low latency, high refresh-rate measurements from phasor measurement units (PMUs). This thesis focuses on the design and validation of resilient synchrophasor networks in order to address some of the challenges that are preventing the adoption in real-fields of what proposed by the literature. First, the architectural layers of synchrophasor networks are described with particular focus on the time dissemination techniques suitable for phasor measurement units. Then, we propose a new data-pushing logic able to minimize the latency introduced at the concentration point without increasing the data incompleteness. The data-pushing logic is validated in three real synchrophasor networks adopting different telecom infrastructures (i.e., 4G-LTE, optical fiber links and twisted pairs). In the context of reliable operation of synchrophasor networks, we propose an algorithm able to deal with intentional or unintentional bad data measurements. We also impersonate an attacker and we show that it is possible to forge delay attacks undetectable by state-of-the-art bad-data detection algorithms that can lead to physical grid damage. We use the proposed bad data algorithm as an effective way of neutralizing these attacks. The IEEE C37.118 Class-P Std compliant synchrophasor extraction algorithm adopted in the real-fields is also implemented and validated in a GPS-synchronized real-time simulator. The simulated PMUs are used to validate two applications for the real-time protection and control of electrical networks by means of synchrophasor technology. The first one is a protection scheme that relies on PMU-based real-time state estimation processes to detect a fault and identify the faulted line. Then, we validate in a real-time hardware-in-the-loop (HIL) setup, the Grid Explicit Congestion Notification (GECN) voltage control mechanism for distribution networks, already presented in the literature. We show that, by leveraging on the accurate knowledge of the system state, GECN is able to control the state of the system in time-scales of seconds, in order to maintain its voltage level within predefined limits.