Trustworthy People-Centric Sensing: Privacy, security and user incentives road-map

Mobile crowd sensing architectural frameworks: A comprehensive survey

“…However, trustworthiness evaluation may contradict users' security and privacy considerations. Thus, mechanisms are required for providing some form of 'accountable anonymity', linking users to their actions, without revealing their identity [8]. In general, as noted in [11], MCS systems should be adaptive and robust in order to identify on one hand inadequate user participation and on the other hand obsolete, redundant, inconsistent and/or inaccurate data collection.…”

“…Privacy concerns are raised, considering, understanding, choosing and controlling the information users share, with whom and for how long [8]. For example, users would like to have access to a smart parking service, but they are reluctant in disclosing sensitive information (such as users' current location, point of interests or personal activities).…”

“…For example, users would like to have access to a smart parking service, but they are reluctant in disclosing sensitive information (such as users' current location, point of interests or personal activities). As noted in [8], sensed data, tagged with spatio-temporal information, could be exploited in order to infer even users' daily routines and habits. Thus, it is imperative to preserve the privacy of the individuals.…”

“…At this point it should be noted that privacy is user specific and to this respect privacy schemes and techniques should take into account and efficiently address variations in users' willingness to share data in the risk of other people acquiring/inferring private information. A potential solution would be to adopt a fully decentralized privacy preserving architecture, enabling users to fully control their own personal data (e.g., decide with whom they are going to share them and when [8], [7]), with the risk however of acquiring potentially low quality data. [8] advances the adaptive combination of centralized solutions that allow for identity management and accountability and decentralized architectures that enable the dissemination of selected information in a peer to peer manner.…”

“…Our view is that a coherent and unified architectural framework, including all key elements related to MCS processes and taking into account their inter-dependencies could constitute a step forward to enabling the unimpeded advancement of MCS. Even though there is a number of related research works that endeavor to define an architectural framework (e.g., [1], [7], [8], [9]), shedding more light to the underlying issues, balancing their inter-play and accounting for and systematically addressing all challenges in a generic, concrete, reliable and robust manner could enable MCS systems to achieve their full potentials. To this respect, in the current work, having as a starting point the MCS unique characteristics and open challenges posed, we comprehensively and critically survey recent MCS systems, models and architectures proposed in related research literature, while analyzing their distinct features and discussing on their relative merits and weaknesses.…”