GlobData is a project that aims to design and implement a middleware tool offering the abstraction of a global object database repository. This tool, called Copla, supports transactional access to geographically distributed persistent objects independent of their location. Additionally, it supports replication of data according to different consistency criteria. For this purpose, Copla implements a number of consistency protocols offering different tradeoffs between performance and fault-tolerance. This paper presents the work on strong consistency protocols for the GlobData system. Two protocols are presented: a voting protocol and a nonvoting protocol. Both these protocols rely on the use of atomic broadcast as a building block to serialize conflicting transactions. The paper also introduces the total order protocol being developed to support large-scale replication.
In ubiquitous networks, the multiple devices carried by an user may unintentionally expose information about her habits or preferences. This information leakage can compromise the users' right to privacy. A common approach to increase privacy is to hide the user real identity under a pseudonym. Unfortunately, pseudonyms may interfere with the reputation systems that are often used to assert the reliability of the information provided by the participants in the network. This paper presents a framework for combining anonymity with reputation and shows that it can be configured to provide a desired degree of balance between these two conflicting goals. The proposed solution leverages on well-known cryptographic techniques, such as public key infrastructure and blind signatures.
Flooding is an expensive but unavoidable operation in some application scenarios devised for Mobile Ad Hoc Networks (MANETs). In this paper, we present a novel algorithm to reduce the overhead imposed by flooding operations. The algorithm improves previous results by using a distributed function to elect the nodes that will provide the highest additional coverage to previous retransmissions. The algorithm does not require any signalling or imposes special requirements on the participating devices.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.