Causal and Δ-causal broadcast in opportunistic networks

Abstract

Causal broadcast is a fundamental communication abstraction for many distributed applications. Several implementations of this abstraction have been proposed over the last decades for traditional networks, that is, networks that assume the existence of a continuous bi-directional end-to-end path between any pair of nodes. Opportunistic networks constitute a kind of networks in which this assumption cannot be made, though, so the implementation of causal broadcast in such networks must be addressed differently. This paper presents two algorithms based on causal barriers that can ensure the causally-ordered delivery of broadcast messages in an opportunistic network, considering both cases where the messages propagate in the network without or with a bounded lifetime. The latter case is especially interesting in networks that must run for a long time, or with a population of nodes that changes continuously.

A Novel Memory Concurrent Editing Model for Large-Scale Video Streams in Edge Computing

“…The copies do not require complex operational transformations and can be updated independently and in parallel without the occurrence of conflicts as assured. Unlike OT algorithms that focus on operation transformation, ORDT is a data structure in its definition, and its core idea is not how to resolve conflicts but to directly design a data structure that can avoid conflict generation [14,[18][19][20][21]. Currently, CRDT algorithms are divided into two main categories: (1) State-based CRDTs (State-based CRDTs or Convergent Replicated Data Types, CvRDTs), which include the current state of the sending node copy in each message, and (2) Operation-based CRDTs (Operation-based CRDT or Commutative Replicated Data Types, CmRDTs), which will carry the updated operations performed by the local node from the last broadcast on each message.…”