Traditional wireless personal-area sensor network accesses physiological signals from a small number of sensor nodes under a low data communication rate. Some clinical applications, e.g. gait control and posture analysis, may need to access physiological signals from a large number of sensor nodes across body under a faster data communication speed in real-time or at least near real-time. This clinical need introduces a new challenge in the design of high-throughput wireless bodyarea sensor networks with a large capacity of sensor nodes. This paper describes the design of a high-throughput, multi-nodes wireless sensor network platform based on the state-of-the-art single-chip microcontroller unit technology with built-in Wi-Fi connectivity. Experimental test demonstrated that the developed wireless sensor network platform was able to support multi-nodes sensor access (10-20) concurrently under a high data sampling rate (>500Hz) in near real-time with a very low packet lost rate (<1%). The designed platform may serve as an alternative solution complement to those wireless sensor networks enabled by the low-power, low-communication rate technology.