A carregar...
Projeto de investigação
Sem título
Financiador
Autores
Publicações
Using TURN Servers for Censorship Evasion
Publication . Vilalonga, Afonso; Gallagher, Kevin; Yaǧan, Osman; Resende, João S.; Domingos, Henrique; NOVALincs
As global online censorship intensifies, developing new evasion systems is essential for improving access to uncensored content in highly censored regions. In addition to creating effective censorship evasion systems, the Bridge Distribution Problem also requires the development of systems that allow users to access the necessary information to connect to censorship evasion proxies, such as IP addresses, without censors blocking the communication channels. This extended abstract introduces a novel system that addresses both challenges by using TURN servers as relays to transmit user traffic to censorship evasion proxies or by allowing users to acquire the necessary information to connect to a censorship evasion proxy from brokers where proxy information is available. It also includes a hypothesis on how using TURN servers might help address these challenges.
ReDunT
Publication . Borrego, Dina; Vilalonga, Afonso; Domingos, Henrique; Preguica, Nuno; Boix, Elisa Gonzalez; Ferreira, Carla; Faculdade de Ciências e Tecnologia (FCT)
Conflict-free Replicated Data Types (CRDTs) are a family of data structures that incorporate conflict resolution mechanisms to ensure state convergence and ease the development of highly available distributed systems. Pure operation-based CRDTs provide a unified framework for defining CRDTs based on a partially ordered log of operations. The framework eases the implementation of CRDTs and reduces the risk of introducing bugs. However, developers are still required to manually define the CRDT semantics and compaction functions, which can be complex and error-prone. This paper proposes ReDunT, a framework for the automatic synthesis of compaction functions for pure op-based CRDTs using a rewrite-based semantic approach. We evaluate ReDunT by applying it to a portfolio of pure operation-based CRDTs, including sets and flags, and comparing the obtained results with the manually optimised specifications. We also demonstrate the feasibility of the framework by implementing it as an automatic tool in K semantic framework.
Unidades organizacionais
Descrição
Palavras-chave
Contribuidores
Financiadores
Entidade financiadora
Fundação para a Ciência e a Tecnologia
Programa de financiamento
Número da atribuição
PRT/BD/154787/2023