Power distribution cables are the core of a power distribution network in the power system. Phase-to-ground faults account for around 70% of all faults in a power distribution network, and thus their identification and location are essential for the reliability and efficiency of a power system. In this paper, a current-measurement technique with magnetic sensing and computational intelligence for a power distribution cable is proposed, which can reconstruct the fault current while retaining its DC offset. From the DC component, the phase-to-ground fault and faulty phase can be identified accordingly. The technique also enables the double-ended current method to locate the fault since current measurement can then be implemented with magnetic sensors at both ends of the cable at a much lower cost than using current transformers. The implementing procedure of this method is presented for guiding its application in practical scenarios. The technique is envisioned to enhance the situational awareness of the smart grid.