An authorization model for multi provider queries

Abstract

We present a novel approach for the specification and enforcement of authorizations that enables controlled data sharing for collaborative queries in the cloud. Data authorities can establish authorizations regulating access to their data distinguishing three visibility levels (no visibility, encrypted visibility, and plaintext visibility). Authorizations are enforced in the query execution by possibly restricting operation assignments to other parties and by adjusting visibility of data on-the-fly. Our approach enables users and data authorities to fully enjoy the benefits and economic savings of the competitive open cloud market, while maintaining control over data.

Toward Owners’ Control in Digital Data Markets

“…Although several approaches have been designed to protect the privacy of the users (e.g., [18]), this problem is particularly difficult in the data market scenario because the analysis may involve data of different owners with different (possibly contrasting) requirements that should be enforced. Also, since the result of an analysis over a (collection of) dataset(s) can leak sensitive information on the data that have been analyzed (and hence on the individuals to whom data refer), the data market should support data trading while protecting the privacy of all involved parties, thus avoiding both direct and indirect leakage of sensitive information (e.g., [28]).…”

Multi-Provider Secure Processing of Sensors Data

“…Given a query formulated by a final user and its query tree plan (e.g., see Figure 2(d)), it is necessary to assign each operation in the tree to a subject respecting the authorizations, while maximizing performance and minimizing economic costs. Note that, as illustrated in [4], encryption can be profitably used to enable the assignment of operations to less expensive, but also less trusted, cloud providers. For instance, considering the example in Figure 2, encryption of attribute ECG would enable Google to evaluate the selection condition.…”

“…2) Estimate economic costs: the algorithm simulates the operation execution by the candidate and generates the profile of the resulting relation, extended with an estimate of the economic cost. 3) Check uniform visibility: since some operators compare or combine attributes, the algorithm verifies the authorization policy also a posteriori, to check whether uniform visibility is satisfied [4], and possibly restricts the set of valid candidates. 4) Assign candidate: the algorithm, according to the greedy approach, assigns the operation to the valid candidate with the lowest economic cost.…”

“…In this paper, we address the problem of supporting a secure and efficient query processing over data collected and controlled by different parties (data authorities). In particular, our reference scenario (Figure 1) is characterized by multiple data authorities collecting data from sensors that aim to collaborate for the evaluation of queries over their data, possibly involving cloud providers that offer computing power at limited prices but that are not necessarily fully trusted to access the data content [4]. Our approach allows the data authorities to independently define authorizations over their data regulating their release to users, other data authorities, and cloud providers.…”

“…Running example authority, cloud provider, final user) can access their data, in plaintext or in encrypted form, we rely on the authorization model in[4]. According to this model, an authorization[P, E] → S over relation R allows subject S to access the set P of attributes in plaintext and the set E of attributes in encrypted form.…”