Mobile crowdsensing (MCS) has been emerging as a new sensing paradigm where vast numbers of mobile devices are used for sensing and collecting data in various applica- tions. Unlike traditional sensor networks (or the static sensing paradigm), which use pre-deployed sensors to collect specific information at fixed locations, MCS leverages a large number of participants (smart mobile device users) to jointly perform sensing and other crowd sourcing tasks. The MCS solution brings several advantages, includ- ing low infrastructure cost, real-time and wide coverage, and potential integration with human intelligence. However, it also faces several research challenges. These in- clude participant selection, incentive mechanisms, and privacy protection and so on. In this work, we mainly focus on designing scalable privacy-preserving participant selection with appropriate incentive for mobile crowdsensing system.Auction based participant selection has been widely used for current MCS systems to achieve user incentive and task assignment optimization. However, participant se- lection problems solved with auction-based approaches usually involve participants’ privacy concerns because a participant’s bids may contain her private information, and disclosure of participants’ bids may disclose their private information as well. Following the classical VCG auction, we carefully design a scalable grouping based privacy-preserving participant selection scheme, which uses Lagrange polynomial in- terpolation (LPI) to perturb participants? bids within groups. The proposed solution, which built on the current MCS platform, can protect such bid privacy in a tempo- rally and spatially dynamic MCS system. Later, we analyze the bidding game of our proposed solution with three implications to prove the security.To address the participant grouping problem with the constraint of communication cost during participant bidding process, we propose two algorithms: sorting and dynamic programming (DP). We prove that sorting algorithm could efficiently achieve a feasible solution with a certain approximation ratio for different problems, while dynamic programming algorithm is proved to provide the optimal solution. However, the selection scheme with cloud-based MCS platform suffers from high overheads, poor scalability and more important. To address this issue and to enhance the protection of user privacy, we further propose a set of novel privacy-preserving grouping methods, which place participants into small groups over hierarchical edge clouds. Our design goal is to group participants in a way that minimizes the com- munication cost during secure sharing/bidding, while satisfying each participant’s requirement for privacy preservation. For different scenarios and optimization functions, we propose a set of grouping schemes to fulfill this goal.For all of above work, extensive simulations over both synthetic and real-life datasets are conducted to verify the efficiency and security, and confirm the effectiveness of proposed mechanisms.