A Comparison of Streaming Sensory Data Transfer Methods in Remote Processing Environment

The advancements in Internet of Things (IOT) has resulted in ubiquitous availability and use of sensors in remote single board computers like raspberry pi or IOT device specific hardware. Though every device has a local storage and processing component but it is not fast and scalable enough to process the huge amount of generated by the sensors involved. The data has to be offloaded to different large computing and processing systems using network. This data can further be used for monitoring, prediction or analysis. Even though there are different communication protocols and architectural styles, namely REST, SOAP, socket and RPCs to efficiently transfer this data but no concrete quantitative data is present to show the efficiency of these methods in different environments. This paper describes these numerous telemetry approaches available to transfer data to a remote server and presents different configurations where each method can prove to be the most effective with respect to different type of sensor data. It also provides a concrete analysis on the type of communication mechanism optimum for a particular length of data transmission.