Currently, there are no known explicit algorithms for the great majority of graph problems in the dynamic distributed message-passing model. Instead, most state-of-the-art dynamic distributed algorithms are constructed by composing a static algorithm for the problem at hand with a simulation technique that converts static algorithms to dynamic ones. We argue that this powerful methodology does not provide satisfactory solutions for many important dynamic distributed problems, and this necessitates developing algorithms for these problems from scratch. In this paper we develop a fully dynamic distributed algorithm for maintaining sparse spanners. Our algorithm improves drastically the quiescence time of the state-of-the-art algorithm for the problem. Moreover, we show that the quiescence time of our algorithm is optimal up to a small constant factor. In addition, our algorithm improves significantly upon the state-of-the-art algorithm in all efficiency parameters, specifically, it has smaller quiescence message and space complexities, and smaller local processing time. Finally, our algorithm is self-contained and fairly simple, and is, consequently, amenable to implementation on unsophisticated network devices.

Recently great attention has been given to point-of-failure swap rerouting, an efficient technique for routing in presence of transient failures. According to this technique, a message follows the normal routing table information unless the next hop has failed; in this case, it is redirected towards a precomputed link, called swap; once this link has been crossed, normal routing is resumed. The amount of precomputed information required in addition to the routing table is rather small: a single link per each destination. Several efficient serial algorithms have been presented to compute this information; none of them can unfortunately be efficiently implemented in a distributed environment. In this paper we present protocols, based on a new strategy, that allow the efficient computation of all the optimal swap edges under several optimization criteria.

Abstract — In this paper, we study the dynamic version of the distributed all-pairs shortest paths problem. Most of the solutions given in the literature for this problem, either (i) work under the assumption that before dealing with an edge operation, the algorithm for the previous operation has to be terminated, that is, they are not able to update shortest paths concurrently, or (ii) concurrently update shortest paths, but their convergence can be very slow (possibly infinite). In this paper we propose a partially dynamic algorithm that overcomes most of these limitations. In particular, it is able to concurrently update shortest paths and in many cases its convergence is quite fast. These properties are highlighted by an experimental study whose aim is to show the effectiveness of the proposed algorithms also in the practical case. Index Terms — Distributed networks, dynamic algorithms, shortest paths, routing, experimental evaluation, network

Author manuscript, published in ""

(will be inserted by the editor) Engineering a new algorithm for distributed shortest paths on dynamic networks

Engineering a new loop-free shortest paths

Abstract. The problem of finding and updating shortest paths in distributed networks is considered crucial in today’s practical applications. In the recent past, there has been a renewed interest in devising new efficient distance-vector algorithms as an attractive alternative to link-state solutions for large-scale Ethernet networks, in which scalability and reliability are key issues or the nodes can have limited storage capabilities. In this paper we present Distributed Computation Pruning (DCP), a new technique, which can be combined with every distance-vector routing algorithm based on shortest paths, allowing to reduce the total number of messages sent by that algorithm and its space occupancy per node. To check its effectiveness, we combined DCP with DUAL (Diffuse Update ALgorithm), one of the most popular distance-vector algorithm in the literature, which is part of CISCO’s widely used EIGRP protocol, and with the recently introduced LFR (Loop Free Routing) which has been shown to have good performances on real networks. We give experimental evidence that these combinations lead to a significant gain both in terms of number of messages sent and memory requirements per node. 1