We propose a novel corridor or hallway gait monitoring system based on radar signal processing, unsupervised learning, and an association method. Unlike some of our previous works where we designed an in-package lens antenna to sharpen the beam of a commercially available radar system, our algorithm proposed in this paper could be paired with any type of MIMO FMCW radar without need for radar antenna alteration to capture human gait in a highly cluttered environment. We validate the improved system functionality through accurately capturing the spatiotemporal gait values (e.g., speed, step points, step count, step time, step length, and cadence) in a long hallway.