Automated Assessment of Aggregate Query Imprecision in Dynamic Environments

Queries are widely used for acquiring data distributed in opportunistically formed mobile networks. However, when queries are executed in such dynamic settings, the returned result may not be consistent, i.e., it may not accurately reflect the state of the environment. It can thus be difficult to reason about the meaning of a query’s result. Reasoning about imperfections in the result becomes even more complex when in-network aggregation is employed, since only a single aggregate value is returned. We define the semantics of aggregate queries in terms of a qualitative description of consistency and a quantitative measure of imprecision. We provide a protocol that performs in-network aggregation while simultaneously generating quality assessments for the query result. The protocol enables intuitive interpretations of the semantics associated with an aggregate query’s execution in a dynamic environment.