The Lightning Network is a payment channel network built on top of the cryptocurrency Bitcoin. It allows Bitcoin to scale by performing transactions off-chain to reduce load on the blockchain. Malicious payment channel participants can try to commit fraud by closing channels with outdated balances. The Lightning Network allows resolving this dispute on the blockchain. However, this mechanism forces the channels' participants to watch the blockchain in regular intervals. It has been proposed to offload this monitoring duty to a third party, called a watchtower. However, existing approaches for watchtowers do not scale as they have storage requirements linear in the number of updates in a channel. In this work, we propose TEE Guard, a new architecture for watchtowers that leverages the features of Trusted Execution Environments to build watchtowers that require only constant memory and are thus able to scale. We show that TEE Guard is deployable because it can run with the existing Bitcoin and Lightning Network protocols. We also show that it is economically viable for a third party to provide watchtower services. As a watchtower needs to be trusted to be watching the blockchain, we also introduce a mechanism that allows customers to verify that a watchtower has been running continuously.

We explore challenges of and present a concept for a decentralized payment service which is based on trusted execution environments. The system guarantees that users can always cash out their funds without depending on the cooperation of other network members, hence minimizing the trust required in other network members. We present an overview of the system, motivate key components for a secure architecture and provide a communication protocol. We prove that the payment service users can cash out their funds at any time without any dependence on other network members.

Ethereum is a platform for deploying smart contracts, which due to their public nature and the financial value of the assets they manage are attractive targets for attacks. With asset management as a main task of smart contracts, access control aspects are naturally part of the application itself, but also of the functions implemented in a smart contract. Therefore, it is desirable to establish the correctness of smart contracts and their access control on application and single-function level through formal methods. However, there is no established methodology of formalising and verifying correctness properties of smart contracts. In this work, we make an attempt in this direction on the basis of a case study. We choose an existing smart contract application which aims to ascertain the integrity of binary files distributed over the Internet by means of decentralised identity management and access control. We formally specify and verify correctness at the level of single functions as well as temporal properties of the overall application. We demonstrate how to use verified low-level correctness properties for showing correctness at the higher level. In addition, we report on our experience with existing verification tools.

Whether a centralized, distributed, or decentralized system approach is selected for Internet-based services affects sovereignty and responsibilities of users and providers alike. Therefore, computer science education can contribute to informed decision-making and citizenship education by teaching power structures of and responsibilities in digital infrastructures. In this practical report, we focus on the example of instant messaging. We analyze three different algorithms for instant messaging that vary in their degree of (de-) centralization. Based on the analysis, we propose a teaching activity called Klemmchat using the concept of computer science unplugged to educate students on the discovered key aspects and trade-offs. We report on results obtained by teaching Klemmchat in two classes in grades 11 and 12. The evaluation shows that the activity is suitable for conveying trade-offs and helping students to engage with the topic. The results, however, leave open whether the acquired understanding affects usage decisions. CCS CONCEPTS• Security and privacy → Social aspects of security and privacy; • Applied computing → Education; • Computer systems organization → Distributed architectures.