Department of Information Systems, Cork Institute of Technology, Cork T12 P928, Ireland.
School of Computer Science and Statistics, Trinity College Dublin, Dublin 2, Ireland.
Sensors (Basel). 2019 Sep 19;19(18):4049. doi: 10.3390/s19184049.
Participatory sensing is a process whereby mobile device users (or participants) collect environmental data on behalf of a service provider who can then build a service based upon these data. To attract submissions of such data, the service provider will often need to incentivize potential participants by offering a reward. However, for the privacy conscious, the attractiveness of such rewards may be offset by the fact that the receipt of a reward requires users to either divulge their real identity or provide a traceable pseudonym. An incentivization mechanism must therefore facilitate data submission and rewarding in a way that does not violate participant privacy. This paper presents Privacy-Aware Incentivization (PAI), a decentralized peer-to-peer exchange platform that enables the following: (i) Anonymous, unlinkable and protected data submission; (ii) Adaptive, tunable and incentive-compatible reward computation; (iii) Anonymous and untraceable reward allocation and spending. PAI makes rewards allocated to a participant untraceable and unlinkable and incorporates an adaptive and tunable incentivization mechanism which ensures that real-time rewards reflect current environmental conditions and the importance of the data being sought. The allocation of rewards to data submissions only if they are truthful (i.e., incentive compatibility) is also facilitated in a privacy-preserving manner. The approach is evaluated using proofs and experiments.
参与式感知是一种移动设备用户（或参与者）代表服务提供商收集环境数据的过程，然后服务提供商可以基于这些数据构建服务。为了吸引此类数据的提交，服务提供商通常需要通过提供奖励来激励潜在参与者。然而，对于注重隐私的人来说，奖励的吸引力可能会因为接受奖励需要用户透露真实身份或提供可追踪的化名而被抵消。因此，激励机制必须以不侵犯参与者隐私的方式促进数据提交和奖励。本文提出了隐私感知激励（Privacy-Aware Incentivization，PAI），这是一种去中心化的点对点交换平台，能够实现以下功能：（i）匿名、不可链接和受保护的数据提交；（ii）自适应、可调谐和激励相容的奖励计算；（iii）匿名和不可追踪的奖励分配和支出。PAI 使分配给参与者的奖励无法追踪和链接，并结合了自适应和可调谐的激励机制，确保实时奖励反映当前环境条件和所寻求数据的重要性。只有在数据提交是真实的情况下（即激励相容性）才会分配奖励，这也以隐私保护的方式得到促进。该方法使用证明和实验进行了评估。
