Power profiling of multimedia sensor node with name-based segment streaming

Multimedia streaming from miniaturized sensors is attractive for a wide range of web-based applications, including surveillance and Internet of Things (IoT) applications. This paper profiles the power consumption in a wireless video sensor node. We compare the power consumption of video streaming frameworks based on a manifest file, such as the Hypertext Transfer Protocol (HTTP) Live Streaming (HLS), with a Wireless Video Sensor Network Platform compatible Dynamic Adaptive Streaming over HTTP (WVSNP-DASH) framework. The WVSNP-DASH framework is based on independently playable video segments that convey the metadata required for playback in their names (and do not require a manifest file). The power consumption components of the video capture and storage pipeline are evaluated. The presented extensive power profiling measurements provide real-world empirical data on architectural design decisions for multimedia sensor nodes suitable for IoT applications. Our measurement results indicate that the name-based WVSNP-DASH framework is well suited for flexible low-power web-based video streaming from miniaturized sensors.