Edge computing offers computing resources at network edges to provide low-latency cloud services or computing offloading for mobile computing services and IoT applications. One essential problem is how to effectively allocate crowdsourced edge resources to users in a highly dynamic and unpredicted environment. This cannot be realized without a truthful open market, while incorporating privacy-preserving requirement of individual users. The existing differentially private auction models, designed for single-sided single-item auction, is not applicable in the two-sided combinatorial edge markets. In this thesis, we first discuss edge computing and how crowdsourcing and an open market can help to expand its coverage. We propose an online privacy-preserving truthful double auction mechanism...[ Read more ]

Edge computing offers computing resources at network edges to provide low-latency cloud services or computing offloading for mobile computing services and IoT applications. One essential problem is how to effectively allocate crowdsourced edge resources to users in a highly dynamic and unpredicted environment. This cannot be realized without a truthful open market, while incorporating privacy-preserving requirement of individual users. The existing differentially private auction models, designed for single-sided single-item auction, is not applicable in the two-sided combinatorial edge markets. In this thesis, we first discuss edge computing and how crowdsourcing and an open market can help to expand its coverage. We propose an online privacy-preserving truthful double auction mechanism for dynamic resource cooperation at the edge. To cope with uncertainties in market behaviours, we decompose the online optimization into a series of single-round auctions such that their objectives can be iteratively obtained to capture the temporally-coupled nature of the problem. Based on that, we design and implement a near-optimal allocation policy with efficiency and privacy guarantee. verified through theoretical analysis and experiment.