Provably secure dual-mode publicly verifiable computation protocol in marine wireless sensor networks

Abstract

In the marine wireless sensor networks, marine sensors collect multidimensional data such as temperature, salinity, dissolved oxygen and chlorophyll concentration in the ocean using a single hardware unit for further statistical analysis. Once these data are collected, they will be sent to the satellites or vessels for scientific information processing purposes, e.g. computing the mean, the variance and making regression analysis. Verifiable computation (VC) always allows the computationally weak parties to execute computing function operations over outsourced data sets or perform data sets towards outsourced functions to the cloud and still provides an efficient way to verify the returned result, which is an important issue in marine wireless sensor networks. However, the oceangoing voyage vessels always have low computational abilities, in such a way that they may outsource some computations (that need expensive computation costs by themselves) to the data center on the land (e.g. cloud). The computational results cannot be used directly since the cloud may return an incorrect outcome for some profits. Hence, we design a secure publicly VC protocol called DM−PVC, which supports both public delegation and public verifiability properties and tackles outsourced functions and outsourced data sets in a combined way. We additionally prove the proposed DM−PVC secure in the random oracle model and evaluate its performance in the end.