Gait monitoring is one of the most demanding areas in the rapidly growing mobile health field. We developed a smartphone-based architecture (called “NeuroSENS”) to improve patient-clinician interaction and to promote the prolonged monitoring of neurological gait by the patients themselves. A particular attention was paid to the security and privacy issues in patient’s data transfer, that are assured at three levels in an in-depth defense strategy (data storage, mobile and web apps and data transmission). Although of very wide application, our architecture offers a first application to detect intermittent claudication and gait asymmetry by estimating duty cycle and ratio between odd and even peaks of autocorrelation from vertical accelerometer signal and rotation of the trunk by the fusion of accelerometer, gyroscope and magnetometer signals in 3D. During exercices on volunteers, sensor data were recorded through the presented architecture with different speeds, durations and constrains. Estimated duty cycles, autocorrelation peaks ratios and trunk rotations showed statistically significant difference (p< 0.05) with knee brace compared to free walk. In conclusion, the NeuroSENS architecture can be used to detect walking irregularities using a readily available mobile platform that addresses security and privacy issues.