Biblio

Since the Information Networks are added to the current electricity networks, the security and privacy of individuals is challenged. This combination of technologies creates vulnerabilities in the context of smart grid power which disrupt the consumer energy supply. Methods based on encryption are against the countermeasures attacks that have targeted the integrity and confidentiality factors. Although the cryptography strategies are used in Smart Grid, key management which is different in size from tens to millions of keys (for meters), is considered as the critical processes. The Key mismanagement causes to reveal the secret keys for attacker, a symmetric key distribution method is recently suggested by [7] which is based on a symmetric key distribution, this strategy is very suitable for smart electric meters. The problem with this method is its vulnerability to impersonating respondents attack. The proposed approach to solve this problem is to send the both side identifiers in encrypted form based on hash functions and a random value, the proposed solution is appropriate for devices such as meters that have very little computing power.

By applying power usage statistics from smart meters, users are able to save energy in their homes or control smart appliances via home automation systems. However, owing to security and privacy concerns, it is recommended that smart meters (SM) should not have direct communication with smart appliances. In this paper, we propose a design for a smart meter gateway (SMGW) associated with a two-phase authentication mechanism and key management scheme to link a smart grid with smart appliances. With placement of the SMGW, we can reduce the design complexity of SMs as well as enhance the strength of security.

Integration of information technologies with the current power infrastructure promises something further than a smart grid: implementation of smart cities. Power efficient cities will be a significant step toward greener cities and a cleaner environment. However, the extensive use of information technologies in smart cities comes at a cost of reduced privacy. In particular, consumers' power profiles will be accessible by third parties seeking information over consumers' personal habits. In this paper, a methodology for enhancing privacy of electricity consumption patterns is proposed and tested. The proposed method exploits digital connectivity and predictive tools offered via smart grids to morph consumption patterns by grouping consumers via an optimization scheme. To that end, load anticipation, correlation and Theil coefficients are utilized synergistically with genetic algorithms to find an optimal assembly of consumers whose aggregated pattern hides individual consumption features. Results highlight the efficiency of the proposed method in enhancing privacy in the environment of smart cities.

The collection of high frequency metering data in the emerging smart grid gives rise to the concern of consumer privacy. Anonymization of metering data is one of the proposed approaches in the literature, which enables transmission of unmasked data while preserving the privacy of the sender. Distributed anonymization methods can reduce the dependency on service providers, thus promising more privacy for the consumers. However, the distributed communication among the end-users introduces overhead and requires methods to prevent external attacks. In this paper, we propose four variants of a distributed anonymization method for smart metering data privacy, referred to as the Collaborative Anonymity Set Formation (CASF) method. The performance overhead analysis and security analysis of the variants are done using NS-3 simulator and the Scyther tool, respectively. It is shown that the proposed scheme enhances the privacy preservation functionality of an existing anonymization scheme, while being robust against external attacks.

With the rapid development of smart grid, smart meters are deployed at energy consumers' premises to collect real-time usage data. Although such a communication model can help the control center of the energy producer to improve the efficiency and reliability of electricity delivery, it also leads to some security issues. For example, this real-time data involves the customers' privacy. Attackers may violate the privacy for house breaking, or they may tamper with the transmitted data for their own benefits. For this purpose, many data aggregation schemes are proposed for privacy preservation. However, rare of them cares about both the data aggregation and fine-grained access control to improve the data utility. In this paper, we proposes a data aggregation scheme based on attribute decision tree. Security analysis illustrates that our scheme can achieve the data integrity, data privacy preservation and fine- grained data access control. Experiment results show that our scheme are more efficient than existing schemes.

Legal metrology embraces the regulation and control of measuring instruments (MI) used in a diversity of applications including industry, transportation, commerce, medical care and environment protection [3]. Only in Europe, MI are responsible for an annual turnover of more than 500 billion Euros [1]. In developing countries, MI demand has increased substantially due to the adoption of technologies and methods well established in developed countries [3]. MI also can be seen as elementary build blocks for new technologies such as smart grids, Internet of Things and cyber physical systems [1, 2]. Thus legal metrology is crucial to assure the correctness of measurements, protecting the economic system while regulating consumer relations and enhances MI reliability [2].

Power grids are undergoing major changes due to rapid growth in renewable energy resources and improvements in battery technology. While these changes enhance sustainability and efficiency, they also create significant management challenges as the complexity of power systems increases. To tackle these challenges, decentralized Internet-of-Things (IoT) solutions are emerging, which arrange local communities into transactive microgrids. Within a transactive microgrid, "prosumers" (i.e., consumers with energy generation and storage capabilities) can trade energy with each other, thereby smoothing the load on the main grid using local supply. It is hard, however, to provide security, safety, and privacy in a decentralized and transactive energy system. On the one hand, prosumers' personal information must be protected from their trade partners and the system operator. On the other hand, the system must be protected from careless or malicious trading, which could destabilize the entire grid. This paper describes Privacy-preserving Energy Transactions (PETra), which is a secure and safe solution for transactive microgrids that enables consumers to trade energy without sacrificing their privacy. PETra builds on distributed ledgers, such as blockchains, and provides anonymity for communication, bidding, and trading.

In the past decade, the information security and threat landscape has grown significantly making it difficult for a single defender to defend against all attacks at the same time. This called for introducing information sharing, a paradigm in which threat indicators are shared in a community of trust to facilitate defenses. Standards for representation, exchange, and consumption of indicators are proposed in the literature, although various issues are undermined. In this paper, we take the position of rethinking information sharing for actionable intelligence, by highlighting various issues that deserve further exploration. We argue that information sharing can benefit from well-defined use models, threat models, well-understood risk by measurement and robust scoring, well-understood and preserved privacy and quality of indicators and robust mechanism to avoid free riding behavior of selfish agents. We call for using the differential nature of data and community structures for optimizing sharing designs and structures.

The ubiquity of the Internet and email, have provided a mostly insecure communication medium for the consumer. During the last few decades, we have seen the development of several ways to secure email messages. However, these solutions are inflexible and difficult to use for encrypting email messages to protect security and privacy while communicating or collaborating via email. Under the current paradigm, the arduous process of setting up email encryption is non-intuitive for the average user. The complexity of the current practices has also yielded to incorrect developers' interpretation of architecture which has resulted in interoperability issues. As a result, the lack of simple and easy-to-use infrastructure in current practices means that the consumers still use plain text emails over insecure networks. In this paper, we introduce and describe a novel, holistic model with new techniques for protecting email messages. The architecture of our innovative model is simpler and easier to use than those currently employed. We use the simplified trust model, which can relieve users from having to perform many complex steps to achieve email security. Utilizing the new techniques presented in this paper can safeguard users' email from unauthorized access and protect their privacy. In addition, a simplified infrastructure enables developers to understand the architecture more readily eliminating interoperability.

Universally Composable (UC) framework provides the strongest security notion for designing fully trusted cryptographic protocols, and it is very challenging on applying UC security in the design of RFID mutual authentication protocols. In this paper, we formulate the necessary conditions for achieving UC secure RFID mutual authentication protocols which can be fully trusted in arbitrary environment, and indicate the inadequacy of some existing schemes under the UC framework. We define the ideal functionality for RFID mutual authentication and propose the first UC secure RFID mutual authentication protocol based on public key encryption and certain trusted third parties which can be modeled as functionalities. We prove the security of our protocol under the strongest adversary model assuming both the tags' and readers' corruptions. We also present two (public) key update protocols for the cases of multiple readers: one uses Message Authentication Code (MAC) and the other uses trusted certificates in Public Key Infrastructure (PKI). Furthermore, we address the relations between our UC framework and the zero-knowledge privacy model proposed by Deng et al. [1].

Recently perceived vulnerabilities in public key infrastructures (PKI) demand that a semantic or cognitive definition of trust is essential for augmenting the security through trust formulations. In this paper, we examine the meaning of trust in PKIs. Properly categorized trust can help in developing intelligent algorithms that can adapt to the security and privacy requirements of the clients. We delineate the different types of trust in a generic PKI model.

With the growth of Internet in many different aspects of life, users are required to share private information more than ever. Hence, users need a privacy management tool that can enforce complex and customized privacy policies. In this paper, we propose a privacy management system that not only allows users to define complex privacy policies for data sharing actions, but also monitors users' behavior and relationships to generate realistic policies. In addition, the proposed system utilizes formal modeling and model-checking approach to prove that information disclosures are valid and privacy policies are consistent with one another.

We present CARiSMA, a tool that is originally designed to support model-based security analysis of IT systems. In our recent work, we added several new functionalities to CARiSMA to support the privacy of personal data. Moreover, we introduced a mechanism to assist the system designers to perform a CARiSMA analysis by automatically initializing an appropriate CARiSMA analysis concerning security and privacy requirements. The motivation for our work is Article 25 of Regulation (EU) 2016/679, which requires appropriate technical and organizational controls must be implemented for ensuring that, by default, the processing of personal data complies with the principles on processing of personal data. This implies that initially IT systems must be analyzed to verify if such principles are respected. System models allow the system developers to handle the complexity of systems and to focus on key aspects such as privacy and security. CARiSMA is available at http://carisma.umlsec.de and our screen cast at https://youtu.be/b5zeHig3ARw.

Many enterprises are transitioning towards data-driven business processes. There are numerous situations where multiple parties would like to share data towards a common goal if it were possible to simultaneously protect the privacy and security of the individuals and organizations described in the data. Existing solutions for multi-party analytics that follow the so called Data Lake paradigm have parties transfer their raw data to a trusted third-party (i.e., mediator), which then performs the desired analysis on the global data, and shares the results with the parties. However, such a solution does not fit many applications such as Healthcare, Finance, and the Internet-of-Things, where privacy is a strong concern. Motivated by the increasing demands for data privacy, we study the problem of privacy-preserving multi-party data analytics, where the goal is to enable analytics on multi-party data without compromising the data privacy of each individual party. In this paper, we first propose a secure sum protocol with strong security guarantees. The proposed secure sum protocol is resistant to collusion attacks even with N-2 parties colluding, where N denotes the total number of collaborating parties. We then use this protocol to propose two secure gradient descent algorithms, one for horizontally partitioned data, and the other for vertically partitioned data. The proposed framework is generic and applies to a wide class of machine learning problems. We demonstrate our solution for two popular use-cases, regression and classification, and evaluate the performance of the proposed solution in terms of the obtained model accuracy, latency and communication cost. In addition, we perform a scalability analysis to evaluate the performance of the proposed solution as the data size and the number of parties increase.

Deep Learning has recently become hugely popular in machine learning for its ability to solve end-to-end learning systems, in which the features and the classifiers are learned simultaneously, providing significant improvements in classification accuracy in the presence of highly-structured and large databases. Its success is due to a combination of recent algorithmic breakthroughs, increasingly powerful computers, and access to significant amounts of data. Researchers have also considered privacy implications of deep learning. Models are typically trained in a centralized manner with all the data being processed by the same training algorithm. If the data is a collection of users' private data, including habits, personal pictures, geographical positions, interests, and more, the centralized server will have access to sensitive information that could potentially be mishandled. To tackle this problem, collaborative deep learning models have recently been proposed where parties locally train their deep learning structures and only share a subset of the parameters in the attempt to keep their respective training sets private. Parameters can also be obfuscated via differential privacy (DP) to make information extraction even more challenging, as proposed by Shokri and Shmatikov at CCS'15. Unfortunately, we show that any privacy-preserving collaborative deep learning is susceptible to a powerful attack that we devise in this paper. In particular, we show that a distributed, federated, or decentralized deep learning approach is fundamentally broken and does not protect the training sets of honest participants. The attack we developed exploits the real-time nature of the learning process that allows the adversary to train a Generative Adversarial Network (GAN) that generates prototypical samples of the targeted training set that was meant to be private (the samples generated by the GAN are intended to come from the same distribution as the training data). Interestingly, we show that record-level differential privacy applied to the shared parameters of the model, as suggested in previous work, is ineffective (i.e., record-level DP is not designed to address our attack).

This paper presents an approach to formalizing and enforcing a class of use privacy properties in data-driven systems. In contrast to prior work, we focus on use restrictions on proxies (i.e. strong predictors) of protected information types. Our definition relates proxy use to intermediate computations that occur in a program, and identify two essential properties that characterize this behavior: 1) its result is strongly associated with the protected information type in question, and 2) it is likely to causally affect the final output of the program. For a specific instantiation of this definition, we present a program analysis technique that detects instances of proxy use in a model, and provides a witness that identifies which parts of the corresponding program exhibit the behavior. Recognizing that not all instances of proxy use of a protected information type are inappropriate, we make use of a normative judgment oracle that makes this inappropriateness determination for a given witness. Our repair algorithm uses the witness of an inappropriate proxy use to transform the model into one that provably does not exhibit proxy use, while avoiding changes that unduly affect classification accuracy. Using a corpus of social datasets, our evaluation shows that these algorithms are able to detect proxy use instances that would be difficult to find using existing techniques, and subsequently remove them while maintaining acceptable classification performance.

User privacy is an important issue in a blockchain based transaction system. Bitcoin, being one of the most widely used blockchain based transaction system, fails to provide enough protection on users' privacy. Many subsequent studies focus on establishing a system that hides the linkage between the identities (pseudonyms) of users and the transactions they carry out in order to provide a high level of anonymity. Examples include Zerocoin, Zerocash and so on. It thus becomes an interesting question whether such new transaction systems do provide enough protection on users' privacy. In this paper, we propose a novel and effective approach for de-anonymizing these transaction systems by leveraging information in the system that is not directly related, including the number of transactions made by each identity and time stamp of sending and receiving. Combining probability studies with optimization tools, we establish a model which allows us to determine, among all possible ways of linking between transactions and identities, the one that is most likely to be true. Subsequent transaction graph analysis could then be carried out, leading to the de-anonymization of the system. To solve the model, we provide exact algorithms based on mixed integer linear programming. Our research also establishes interesting relationships between the de-anonymization problem and other problems studied in the literature of theoretical computer science, e.g., the graph matching problem and scheduling problem.

Personalization, recommendations, and user modeling can be pow- erful tools to improve people?s experiences with technology and to help them nd information. However, we also know that people underestimate how much of their personal information is used by our technology and they generally do not understand how much algorithms can discover about them. Both privacy and ethical tech- nology have issues of consent at their heart. This talk will look at how to consider issues of privacy and consent when users cannot explicitly state their preferences, The Creepy Factor, and how to balance users? concerns with the bene ts personalized technology can o er.

The security of Wireless Sensor Networks (WSNs) is vital in several applications such as the tracking and monitoring of endangered species such as pandas in a national park or soldiers in a battlefield. This kind of applications requires the anonymity of the source, known as Source Location Privacy (SLP). The main aim is to prevent an adversary from tracing back a real event to the originator by analyzing the network traffic. Previous techniques have achieved high anonymity such as Dummy Uniform Distribution (DUD), Dummy Adaptive Distribution (DAD) and Controlled Dummy Adaptive Distribution (CAD). However, these techniques increase the overall overhead of the network. To overcome this shortcoming, a new technique is presented: Exponential Dummy Adaptive Distribution (EDAD). In this technique, an exponential distribution is used instead of the uniform distribution to reduce the overhead without sacrificing the anonymity of the source. The exponential distribution improves the lifetime of the network since it decreases the number of transmitted packets within the network. It is straightforward and easy to implement because it has only one parameter $łambda$ that controls the transmitting rate of the network nodes. The conducted adversary model is global, which has a full view of the network and is able to perform sophisticated attacks such as rate monitoring and time correlation. The simulation results show that the proposed technique provides less overhead and high anonymity with reasonable delay and delivery ratio. Three different analysis models are developed to confirm the validation of our technique. These models are visualization model, a neural network model, and a steganography model.

Generating synthetic data is a well-known option to limit disclosure risk in sensitive data releases. The usual approach is to build a model for the population and then generate a synthetic data set solely based on the model. We argue that building an accurate population model is difficult and we propose instead to approximate the original data as closely as privacy constraints permit. To enforce an ex ante privacy level when generating synthetic data, we introduce a new privacy model called $ε$ synthetic privacy. Then, we describe a synthetic data generation method that satisfies $ε$-synthetic privacy. Finally, we evaluate the utility of the synthetic data generated with our method.

We address the problem of ciphertext-policy attribute-based encryption with fine access control, a cryptographic primitive which has many concrete application scenarios such as Pay-TV, e-Health, Cloud Storage and so on. In this context we improve on previous LSSS based techniques by building on previous work of Hohenberger and Waters at PKC'13 and proposing a construction that achieves ciphertext size linear in the minimum between the size of the boolean access formula and the number of its clauses. Our construction also supports fast decryption. We also propose two interesting extensions: the first one aims at reducing storage and computation at the user side and is useful in the context of lightweight devices or devices using a cloud operator. The second proposes the use of multiple authorities to mitigate key escrow by the authority.

The cloud computing paradigm enables enterprises to realise significant cost savings whilst boosting their agility and productivity. However, security and privacy concerns generally deter enterprises from migrating their critical data to the cloud. One way to alleviate these concerns, hence bolster the adoption of cloud computing, is to devise adequate security policies that control the manner in which these data are stored and accessed in the cloud. Nevertheless, for enterprises to entrust these policies, a framework capable of providing assurances about their correctness is required. This work proposes such a framework. In particular, it proposes an approach that enables enterprises to define their own view of what constitutes a correct policy through the formulation of an appropriate set of well-formedness constraints. These constraints are expressed ontologically thus enabling–-by virtue of semantic inferencing–- automated reasoning about their satisfaction by the policies.

Some IoT data are time-sensitive and cannot be processed in clouds, which are too far away from IoT devices. Fog computing, located as close as possible to data sources at the edge of IoT systems, deals with this problem. Some IoT data are sensitive and require privacy controls. The proposed Policy Enforcement Fog Module (PEFM), running within a single fog, operates close to data sources connected to their fog, and enforces privacy policies for all sensitive IoT data generated by these data sources. PEFM distinguishes two kinds of fog data processing. First, fog nodes process data for local IoT applications, running within the local fog. All real-time data processing must be local to satisfy real-time constraints. Second, fog nodes disseminate data to nodes beyond the local fog (including remote fogs and clouds) for remote (and non-real-time) IoT applications. PEFM has two components for these two kinds of fog data processing. First, Local Policy Enforcement Module (LPEM), performs direct privacy policy enforcement for sensitive data accessed by local IoT applications. Second, Remote Policy Enforcement Module (RPEM), sets up a mechanism for indirectly enforcing privacy policies for sensitive data sent to remote IoT applications. RPEM is based on creating and disseminating Active Data Bundles-software constructs bundling inseparably sensitive data, their privacy policies, and an execution engine able to enforce privacy policies. To prove effectiveness and efficiency of the solution, we developed a proof-of-concept scenario for a smart home IoT application. We investigate privacy threats for sensitive IoT data and show a framework for using PEFM to overcome these threats.

Many organizations process personal information in the course of normal operations. Improper disclosure of this information can be damaging, so organizations must obey privacy laws and regulations that impose restrictions on its release or risk penalties. Since electronic management of personal information must be held in strict compliance with the law, software systems designed for such purposes must have some guarantee of compliance. To support this, we develop a general methodology for designing and implementing verifiable information systems. This paper develops the design of the History Aware Programming Language into a framework for creating systems that can be mechanically checked against privacy specifications. We apply this framework to create and verify a prototypical Electronic Medical Record System (EMRS) expressed as a set of actor components and first-order linear temporal logic specifications in assume-guarantee form. We then show that the implementation of the EMRS provably enforces a formalized Health Insurance Portability and Accountability Act (HIPAA) policy using a combination of model checking and static analysis techniques.

This paper argues that standard privacy policy principles are unsuitable for wearable devices, and introduces a proposal to test the role of digital literacy on privacy concerns and behaviors, in an effort to devise modified privacy policies that are appropriate for wearable devices.