A power system suffers from unexpected various faults, which needs smart devices for accurate estimates of fault detection, classification and direction discrimination to provide inspection, maintenance, and repair of the actual faults. Signal processing-based protection technique can be used to detect the disturbances of the measured signals. This paper presents a digital protection scheme, based on three indices derived from the three-phase input and output current signals of the protected equipment, to detect current disturbances, classify faults, discriminate fault directions, select faulty phase(s), define current unbalances, and identify CT saturation extent. Moreover, it develops a new closed-tripping characteristic with a quadrilateral form for differentiating the fault location (internal/external) with respect to the element protection zone. A power system simulation is carried out using ATP platform, and the technique performance is validated via MATLAB software. Extensive simulation case studies show that the proposed algorithm provides a fast relay response and high accuracy under various operating and fault conditions. It is able to perform successfully the function of fault diagnosis within one periodic cycle. Besides, it responds very well with regard to dependability, security, stability and sensitivity. In addition to that it has the ability to recognize the saturated and unsaturated secondary current signals, measured using the current transformers, in the cases of internal and external faults.