Towards a reliable Architecture for Crowdsourcing

Extension of navigation applications to indoor use is a current research focus. There exist several concepts; however, all share the problem of ascertaining infrastructure data. Further, maintenance of the data is challenging. "Crowdsourcing" is an ansatz to outsource these tasks from the service provider to the user. There exist different derivates of crowdsourcing; howerver, no commonly accepted definition. An attempt of a definition is provided, resulting in a crowdsourcing taxonomy. Utilizing the taxonomy conceived, an architecture extension compliant to German data protection laws is introduced in order to support different classes of crowdsourcing. The extension conceived centres the "Indoor Navigation Server Access Network Entity" (INSANE) which acts as a proxy for crowdsourcing communication, effectively concealing the crowdsourcers' identity from the crowdfunders. Addressing security and privacy issues, a principal user management as well as concealing techniques are applied. Avoiding a single point of failure at the INSANE, utilisation of a distributed hash table is proposed. The concept is exemplarily implemented and proven to be of good performance and scalability.