On the evaluation of cloud web services for crowdsourcing mobile applications

Despite the growing importance of mobile crowdsourcing applications and cloud computing, little is known about the actual performances of web services deployed within public cloud computing platforms. In order to provide an assessment of the achievable performances in such scenario, we design and implement a back-end general architecture for mobile applications requiring crowdsourcing. We deploy our back-end in the Microsoft Azure cloud computing platform using the PaaS approach, and we evaluate its performance in terms of autoscaling, response time and request rate; while varying the number of instances, the instance type, and the number of concurrent users. Our results shed light on the achievable performances of web services aimed at ingesting crowdsourced data.