We present an expert system based on a Bayesian network to advise a Smart Grid developer on how to introduce privacy measures and the associated equipment (i.e., communication protocol, smart meter type) based on his needs. The implementation of the application is completely described in  and accesible through this link.
This system should work as follows: the user (i.e., a SG developer) provides the application with information related to the desired communication model (i.e., its topology) and the optional compliance with some control requirements (real-time performance, sustainability, dependability and survivability). Then, the system computes a recommendation of a privacy technique that suits these conditions giving its required smart meter type, according to their computational capabilities The system also finds the most suitable communication standard to transfer data from the user domain to the utility (specifically, one to connect the appliances with the smart meter in the HAN and another to connect the meter with the gateways in the AMI). To achieve this, the user has to set the requirements of each protocol in terms of reliability, security and performance.The privacy scenario (i.e., home, industrial, vehicular) also has to be established, as each one concedes different level of importance to security requirements.
- "Recommender System for Privacy-Preserving Solutions in Smart Metering",
Pervasive and Mobile Computing, vol. 41, Pervasive and Mobile Computing, pp. 205-218, 10/2017. (I.F.: 2.974)
Nowadays, Smart Grid is envisaged to provide several benefits to both customers and grid operators. However, Smart Meters introduce many privacy issues if consumption data is analysed. In this paper we analyse the main techniques that address privacy when collecting electricity readings. In addition to privacy, it is equally important to preserve efficiency to carry on with monitoring operations, so further control requirements and communication protocols are also studied. Our aim is to provide guidance to installers who intend to integrate such mechanisms on the grid, presenting an expert system to recommend an appropriate deployment strategy.Impact Factor: 2.974Journal Citation Reports® Science Edition (Thomson Reuters, 2017)