5th generation (5G) millimeter wave (mmWave) positioning systems are of growing interest for application in operating environments where global navigation satellite system (GNSS) signals are unavailable or unreliable, promising enhancements to positioning accuracy. Application environments range from warehouses and indoor areas to dense urban spaces. However, in real dynamic operating conditions, brief signal outages are expected due to both environmental features and moving objects such as cars or pedestrians. During 5G signal outages, a positioning system must rely on alternative positioning systems and sensors. Inertial navigation systems (INS) provide a self-contained positioning solution unaffected by environmental factors. However, when operating alone INS suffers from unbounded drift in position error. Automotive radar, or electronic scanning radar (ESR) are low-cost sensors integrated in most modern vehicles, and are of increasing interest to positioning applications. This paper presents an extended Kalman filter (EKF) fusion architecture integrating a 5G positioning system with pose corrections from an ESR scan to map registration algorithm. 5G measurements are simulated in a quasi-real environment, and all radar and INS data are collected from real road tests within a GNSS-denied indoor parking garage. An array of 5G signal outages of varying lengths and characteristics are inflicted on the positioning system. The radar-aided positioning system maintains an average root mean squared error of 0.6m during 5G signal outages, improving the 5G/INS performance by 70% across the tested scenarios.