Comparing AWS Streaming Services: A Use Case on ECG Data Streams

The goal of this research paper is to compare Amazon Kinesis Data Streams and SQS via a practical use case for processing electrocardiogram medical data. Both services are tested on a large number of concurrent data streams up to 10.000 concurrent data streams, measuring and evaluating their performance. In both cases, the producer is a programmed data generator that supplies the data to the services. The motivation of the paper is the streaming of textual files with 125-500 samples per second written as integers, representing electrocardiograms from sensors. We set a research hypothesis that Amazon Kinesis will outperform SQS service for streaming medical data collected from sensors. The pros and cons are listed for each of the Amazon services taking into account the price, use cases when each of the services will be more applicable, and the scalability. The concluding results from this experiment are that both of the services provide great performance but they will mainly provide better results if each of them is applied to a specific use case.