A large portion of Internet traffic passes through middleboxes that read or modify messages. However, as more traffic is protected with TLS, middleboxes are becoming unable to provide their functions. To leverage middlebox functionality while preserving communication security, secure middlebox channel protocols have been designed as extensions of TLS. A key idea is that the endpoints explicitly incorporate middleboxes into the TLS handshake and grant each middlebox either the read or the write permission for their messages. Because each middlebox has the least data access privilege, these protocols are resilient against the compromise of a single middlebox. However, the existing studies have not comprehensively analyzed the communication security under the scenarios where multiple middleboxes are compromised. In this paper, we present novel attacks that break the security of the existing protocols under such scenarios and then modify maTLS, the state-of-the-art protocol, so that all the attacks are prevented with marginal overhead.

Consumer privacy leakage from data names poses a serious threat to Content-Centric Networking (CCN) networks. Obfuscating names is a promising countermeasure, and anonymizers with deterministic encryption schemes have been proposed to provide data privacy while enabling CCN features, such as in-network caching. Existing studies assume a weak threat model in which anonymizers are honest, and their obfuscation schemes are not resilient against privacy attacks such as name guessing attacks. This paper designs a name obfuscation framework based on the realistic assumption that anonymizers are semi-honest. The framework strengthens data privacy using multiple keys and separates obfuscation for prefixes and suffixes, and is implemented on a P4 switch to provide Tbps forwarding speed.

Several platforms to efficiently retrieve content from Internet of Things (IoT) devices installed in various locations have been proposed for information-centric networking (ICN). However, location privacy is at stake in such platforms because consumers retrieve content by specifying the plaintext names of the locations of their interest (LOIs). Previous studies on IP have leveraged k-anonymity of location offered by a trusted proxy called an anonymizer to hide LOIs. Specifically, an anonymizer sends content requests to k locations in a location anonymity set, which comprises an LOI and the other dummy locations. This technique can be applied to ICN; however, two problems need to be solved: the adversary models are unrealistic and the requirements for a location anonymity set have been defined in ad-hoc manners. In this study, we assume a semi-honest anonymizer and define the requirements rigorously using the notions of entropy and t-closeness. Next, we design an architecture for location privacy protection and an algorithm for location anonymity set generation. Finally, we evaluate the overhead incurred by our architecture and the quality of generated location anonymity sets through experiments under a realistic scenario. Our results indicate that our architecture and algorithm offer strong location privacy with marginal overhead.

Federated learning is a privacy-preserving learning system where participants locally update a shared model with their own training data. Despite the advantage that training data are not sent to a server, there is still a risk that a state-of-the-art model inversion attack, which may be conducted by the server, infers training data from the models updated by the participants, referred to as individual models. A solution to prevent such attacks is differential privacy, where each participant adds noise to the individual model before sending it to the server. Differential privacy, however, sacrifices the quality of the shared model in compensation for the fact that participants' training data are not leaked. This paper proposes a federated learning system that is resistant to model inversion attacks without sacrificing the quality of the shared model. The core idea is that each participant divides the individual model into model fragments, shuffles, and aggregates them to prevent adversaries from inferring training data. The other benefit of the proposed system is that the resulting shared model is identical to the shared model generated with the naive federated learning.

Named Data Networking (NDN) is one of promising next generation Internet architectures that aim to realize efficient content distribution. However, in terms of producer anonymity, NDN has a serious problem that adversaries can easily learn who publishes what content due to its feature that content is inherently tied to the producer by the content name and the signature. In this article, we first define producer anonymity rigorously in terms of content-producer unlinkability, and then design a system to achieve it. Our design is based on hidden service, which is an onion routing-based system in IP, however, we improve it to take full advantage of NDN. We demonstrate that our system provides a level of anonymity comparable to hidden service with lower overhead through analysis and experiment.

Verifying remote computing environments, such as computing nodes in fog and edge computing, has gained considerable attention. This poster extends an existing remote attestation method so that it can verify that obtained results are generated by trusted computing nodes as well as remote computing nodes are trusted.