Friday, March 27, 2015

Case study: Smart Grid

Before 2020, it is expected that the European countries will improve their energy distribution networks by replacing electricity meters with smart meters. One of the corner stones of this improvement is the introduction of the Smart Grid concept: a large network of energy providers and energy consumers. Motivated by the lower-carbon generation, combined with greatly improved efficiency on the demand side, it is expected that consumers will interaction more often with the energy supply system by delivering the excess energy from their solar panels, wind turbines and heat pumps to the network.

The Smart Grid will be a large and complex system and information will be collected, processed, stored, communicated at all levels: an electricity smart meter in a house collects measurements on a regular basis (e.g. every 10 minutes) and forwards those measures to the nodes up in the Smart Grid. These measures are used remotely and  automatically for many purposes: for instance, the real time management of the energy balance in the network and the computation of the bill with fully dynamic tariffs. The Smart Meter can also offer additional services like the remote control of the switch. This Smart Grid revolution introduces many challenges, as part of the HEAT project we are especially interested in improving the overall security and preserving the privacy of the user.

Four years ago, NXP Semiconductors Belgium N.V. actively participated in European Commission expert groups for privacy and security of the grid. NXP also made the link between the Smart Grid and the more generic problem of the security of the industrial  automation control systems. These cooperation's have led to the definition of new NXP products that are now being deployed together with several partners, first during a prototyping phase in Belgium  and later in a production phase in Germany and France and in a near future in production in the UK.

However, to realize such a smart grid system requires that privacy and data protection law must be taken into account as an important requirement for the design of smart metering systems. For example, this was a concern for the Dutch senate who blocked two smart metering bills in 2009. One very promising solution to alleviate these concerns, but which still enables utility companies to detect fraud and apply advanced statistics (e.g. required for billing), is enabling the computation on the encrypted meter readings. Techniques which make the computation on encrypted data possible is known as homomorphic encryption and are the topic of study within this project. More specifically, we are interested in designing schemes which are efficient and memory-friendly and allow arbitrary computation on ciphertexts: a technique which is known as fully homomorphic encryption (FHE).

No comments:

Post a Comment