TY  - JOUR
A1  - Ambassa, Pacome L.
A1  - Kayem, Anne Voluntas dei Massah
A1  - Wolthusen, Stephen D.
A1  - Meinel, Christoph
T1  - Inferring private user behaviour based on information leakage
JF  - Smart Micro-Grid Systems Security and Privacy
N2  - In rural/remote areas, resource constrained smart micro-grid (RCSMG) architectures can provide a cost-effective power supply alternative in cases when connectivity to the national power grid is impeded by factors such as load shedding. RCSMG architectures can be designed to handle communications over a distributed lossy network in order to minimise operation costs. However, due to the unreliable nature of lossy networks communication data can be distorted by noise additions that alter the veracity of the data. In this chapter, we consider cases in which an adversary who is internal to the RCSMG, deliberately distorts communicated data to gain an unfair advantage over the RCSMG’s users. The adversary’s goal is to mask malicious data manipulations as distortions due to additive noise due to communication channel unreliability. Distinguishing malicious data distortions from benign distortions is important in ensuring trustworthiness of the RCSMG. Perturbation data anonymisation algorithms can be used to alter transmitted data to ensure that adversarial manipulation of the data reveals no information that the adversary can take advantage of. However, because existing data perturbation anonymisation algorithms operate by using additive noise to anonymise data, using these algorithms in the RCSMG context is challenging. This is due to the fact that distinguishing benign noise additions from malicious noise additions is a difficult problem. In this chapter, we present a brief survey of cases of privacy violations due to inferences drawn from observed power consumption patterns in RCSMGs centred on inference, and propose a method of mitigating these risks. The lesson here is that while RCSMGs give users more control over power management and distribution, good anonymisation is essential to protecting personal information on RCSMGs.
KW  - Approximation algorithms
KW  - Electrical products
KW  - Home appliances
KW  - Load modeling
KW  - Monitoring
KW  - Power demand
KW  - Wireless sensor networks
KW  - Distributed snapshot algorithm
KW  - Micro-grid networks
KW  - Power consumption characterization
KW  - Sensor networks
Y1  - 2018
SN  - 978-3-319-91427-5
SN  - 978-3-319-91426-8
U6  - https://doi.org/10.1007/978-3-319-91427-5_7
VL  - 71
SP  - 145
EP  - 159
PB  - Springer
CY  - Dordrecht
ER  -