## Efficient algorithms for constructing (1 + ɛ, β)-spanners in the distributed and streaming models (2004)

Venue: | Distributed Computing |

Citations: | 20 - 6 self |

### BibTeX

@INPROCEEDINGS{Elkin04efficientalgorithms,

author = {Michael Elkin},

title = {Efficient algorithms for constructing (1 + ɛ, β)-spanners in the distributed and streaming models},

booktitle = {Distributed Computing},

year = {2004},

pages = {160--168}

}

### Years of Citing Articles

### OpenURL

### Abstract

For an unweighted undirected graph G = (V, E), and a pair of positive integers α ≥ 1, β ≥ 0, a subgraph G ′ = (V, H), H ⊆ E, is called an (α, β)-spanner of G if for every pair of vertices u, v ∈ V, distG ′(u, v) ≤ α · distG(u, v) + β. It was shown in [20] that for any ɛ> 0, κ = 1, 2,..., there exists an integer β = β(ɛ, κ) such that for every n-vertex graph G there exists a (1 + ɛ, β)-spanner G ′ with O(n 1+1/κ) edges. An efficient distributed protocol for constructing (1+ ɛ, β)-spanners was devised in [18]. The running time and the communication complexity of that protocol are O(n 1+ρ) and O(|E|n ρ), respectively, where ρ is an additional control parameter of the protocol that affects only the additive term β. In this paper we devise a protocol with a drastically improved running time (O(n ρ) as opposed to O(n 1+ρ)) for constructing (1 + ɛ, β)-spanners. Our protocol has the same communication complexity as the protocol of [18], and it constructs spanners with essentially the same properties as the spanners that are constructed by the protocol of [18]. We also show that our protocol for constructing (1+ɛ, β)spanners can be adapted to the streaming model, and devise a streaming algorithm that uses a constant number of passes and O(n 1+1/κ · log n) bits of space for computing allpairs-almost-shortest-paths of length at most by a multiplicative factor (1 + ɛ) and an additive term of β greater than the shortest paths. Our algorithm processes each edge in time O(n ρ), for an arbitrarily small ρ> 0. The only

### Citations

700 | The Space Complexity of Approximating the Frequency Moments
- Alon, Matias, et al.
- 1996
(Show Context)
Citation Context ... size of the input). Furthermore, the algorithm is only allowed to read the input sequentially using a small number of passes through the input. This computational model was developed in the works of =-=[3, 25, 22]-=-, and is important for processing massive data sets (see, e.g., [28], and the references therein). While originally the research in this area focused on numerical problems such as estimating the frequ... |

294 |
Distributed computing: a locality-sensitive approach
- PELEG
(Show Context)
Citation Context ...of a protocol is the (worst-case) number of messages that are sent by the processors during the execution of the protocol. The design of distributed algorithms is a vivid area of research (see, e.g., =-=[29]-=-, and the references therein). Spanners serve as an important tool in this area, and particularly, they were used for routing [32, 7], for constructing synchronizers [31, 6], and for computing almost ... |

260 | Stable distributions, pseudorandom generators, embeddings and data stream computation
- Indyk
- 2000
(Show Context)
Citation Context ... (see, e.g., [28], and the references therein). While originally the research in this area focused on numerical problems such as estimating the frequency moments [3, 16], and L 1 - and L 2 -distances =-=[22, 26]-=-, more recently the problems involving massive graphs have become a subject of study [28, 21]. In this paper we show that our distributed algorithm can be converted into a streaming algorithm for cons... |

251 | Clustering data streams
- Guha, Mishra, et al.
- 2000
(Show Context)
Citation Context ...aming algorithms for clustering in graphs. Despite the fact that the clustering problem is extremely well-studied in different disciplines, we are only aware of very few previous streaming algorithms =-=[24, 12]-=- for clustering metric spaces. However, to the best of our knowledge, these algorithms are not applicable to the problem of clustering in graphs, and hence our result is incomparable with those of [24... |

155 | Computing on data streams
- Henzinger, Raghavan, et al.
- 1998
(Show Context)
Citation Context ... size of the input). Furthermore, the algorithm is only allowed to read the input sequentially using a small number of passes through the input. This computational model was developed in the works of =-=[3, 25, 22]-=-, and is important for processing massive data sets (see, e.g., [28], and the references therein). While originally the research in this area focused on numerical problems such as estimating the frequ... |

154 |
Graph spanners
- Peleg, Schäffer
- 1989
(Show Context)
Citation Context ... close approximation of the metric space U of the graph G. Graph spanners have multiple applications in the areas of Graph Algorithms and Distributed Computing, and were subject of extensive research =-=[30, 4, 13, 11, 17, 20, 18, 33, 9, 10]-=- in the course of the last fifteen years. In the area of Distributed Computing spanners were found particularly useful in the following (henceforth called distributed) model of computation. In this mo... |

152 | The probabilistic communication complexity of set intersection - Kalyanasundaram, Schnitger - 1992 |

133 |
A trade-off between space and efficiency for routing tables
- Peleg, Upfal
- 1989
(Show Context)
Citation Context ...ign of distributed algorithms is a vivid area of research (see, e.g., [29], and the references therein). Spanners serve as an important tool in this area, and particularly, they were used for routing =-=[32, 7]-=-, for constructing synchronizers [31, 6], and for computing almost shortest paths [4, 18]. In the 90’s most of the study of spanners and their applications focused on spanners whose metric space disto... |

127 |
An optimal synchronizer for the hypercube
- Peleg, Ullman
- 1989
(Show Context)
Citation Context ... area of research (see, e.g., [29], and the references therein). Spanners serve as an important tool in this area, and particularly, they were used for routing [32, 7], for constructing synchronizers =-=[31, 6]-=-, and for computing almost shortest paths [4, 18]. In the 90’s most of the study of spanners and their applications focused on spanners whose metric space distorts the original metric space by at most... |

79 | New sparseness results on graph spanners
- Chandra, Das, et al.
- 1995
(Show Context)
Citation Context ... close approximation of the metric space U of the graph G. Graph spanners have multiple applications in the areas of Graph Algorithms and Distributed Computing, and were subject of extensive research =-=[30, 4, 13, 11, 17, 20, 18, 33, 9, 10]-=- in the course of the last fifteen years. In the area of Distributed Computing spanners were found particularly useful in the following (henceforth called distributed) model of computation. In this mo... |

79 | All pairs almost shortest paths
- Dor, Halperin, et al.
(Show Context)
Citation Context ... close approximation of the metric space U of the graph G. Graph spanners have multiple applications in the areas of Graph Algorithms and Distributed Computing, and were subject of extensive research =-=[30, 4, 13, 11, 17, 20, 18, 33, 9, 10]-=- in the course of the last fifteen years. In the area of Distributed Computing spanners were found particularly useful in the following (henceforth called distributed) model of computation. In this mo... |

75 | Routing with polynomial communication-space tradeo
- Awerbuch, Peleg
- 1992
(Show Context)
Citation Context ...ign of distributed algorithms is a vivid area of research (see, e.g., [29], and the references therein). Spanners serve as an important tool in this area, and particularly, they were used for routing =-=[32, 7]-=-, for constructing synchronizers [31, 6], and for computing almost shortest paths [4, 18]. In the 90’s most of the study of spanners and their applications focused on spanners whose metric space disto... |

71 |
On the exponent of the all pairs shortest path problem
- Alon, Galil, et al.
- 1997
(Show Context)
Citation Context ...urvey [35] and the references therein). We remark that many of the algorithms for all-pairs-(almost)shortest-paths in the traditional model of computation rely heavily either on matrix-multiplication =-=[2, 23, 34]-=-, or on multiple BFS traversals of the entire graph [1, 17, 15]. Both these approaches appear to be completely unsuitable for the streaming model of computation.sNote also that the space complexity of... |

71 | Better streaming algorithms for clustering problems
- Charikar, O’Callaghan, et al.
- 2003
(Show Context)
Citation Context ...aming algorithms for clustering in graphs. Despite the fact that the clustering problem is extremely well-studied in different disciplines, we are only aware of very few previous streaming algorithms =-=[24, 12]-=- for clustering metric spaces. However, to the best of our knowledge, these algorithms are not applicable to the problem of clustering in graphs, and hence our result is incomparable with those of [24... |

68 |
On the all-pairs-shortest-path problem in unweighted undirected graphs
- Seidel
- 1995
(Show Context)
Citation Context ...urvey [35] and the references therein). We remark that many of the algorithms for all-pairs-(almost)shortest-paths in the traditional model of computation rely heavily either on matrix-multiplication =-=[2, 23, 34]-=-, or on multiple BFS traversals of the entire graph [1, 17, 15]. Both these approaches appear to be completely unsuitable for the streaming model of computation.sNote also that the space complexity of... |

66 | Fast estimation of diameter and shortest paths (without matrix multiplication
- Aingworth, Chekuri, et al.
- 1999
(Show Context)
Citation Context ...the algorithms for all-pairs-(almost)shortest-paths in the traditional model of computation rely heavily either on matrix-multiplication [2, 23, 34], or on multiple BFS traversals of the entire graph =-=[1, 17, 15]-=-. Both these approaches appear to be completely unsuitable for the streaming model of computation.sNote also that the space complexity of our streaming algorithm O(n 1+1/κ · log n) is not far from opt... |

64 | Computing almost shortest paths
- Elkin
- 2001
(Show Context)
Citation Context ...ger β = β(ɛ, κ) such that for every n-vertex graph G there exists a (1 + ɛ, β)-spanner G ′ with O(n 1+1/κ ) edges. An efficient distributed protocol for constructing (1+ ɛ, β)-spanners was devised in =-=[18]-=-. The running time and the communication complexity of that protocol are O(n 1+ρ ) and O(|E|n ρ ), respectively, where ρ is an additional control parameter of the protocol that affects only the additi... |

59 | On graph problems in a semi-streaming model
- Feigenbaum, Kannan, et al.
- 2005
(Show Context)
Citation Context ... ...$5.00. [Extended Abstract] Jian Zhang ‡ Computer Science Department Yale University, P.O. Box 208285 New Haven, CT 06520-8285 USA zhang-jian@cs.yale.edu previously known algorithm for the problem =-=[21]-=- constructs paths of length κ times greater than the shortest paths, has the same space requirements as our algorithm, but requires O(n 1+1/κ ) time for processing each edge of the input graph. Howeve... |

55 | Exact and Approximate Distances in Graphs - A Survey
- Zwick
- 2001
(Show Context)
Citation Context ...rk that in the traditional (not streaming) model of computation, the problem of computing almost shortest paths was the focus of extensive recent research ([17, 15, 18]; see also the excellent survey =-=[35]-=- and the references therein). We remark that many of the algorithms for all-pairs-(almost)shortest-paths in the traditional model of computation rely heavily either on matrix-multiplication [2, 23, 34... |

54 | Network synchronization with polylogarithmic overhead
- Awerbuch, Peleg
- 1990
(Show Context)
Citation Context ... area of research (see, e.g., [29], and the references therein). Spanners serve as an important tool in this area, and particularly, they were used for routing [32, 7], for constructing synchronizers =-=[31, 6]-=-, and for computing almost shortest paths [4, 18]. In the 90’s most of the study of spanners and their applications focused on spanners whose metric space distorts the original metric space by at most... |

54 | An approximate L 1 -difference algorithm for massive data streams - Feigenbaum, Kannan, et al. - 1999 |

41 | Near linear time construction of sparse neighborhood covers
- Awerbuch, Berger, et al.
- 1998
(Show Context)
Citation Context |

40 |
Approximate distance oracles for unweighted graphs in expected O(n 2 ) time
- Baswana, Sen
(Show Context)
Citation Context |

36 | A faster distributed protocol for constructing a minimum spanning tree
- Elkin
- 2004
(Show Context)
Citation Context ...treaming model of computation. It is also conceivable that the number of passes that is used by our algorithm will be reduced in the future. Finally, we remark that our adaptation of the algorithm of =-=[19]-=- for constructing neighborhood covers for the streaming model is one of the first existing streaming algorithms for clustering in graphs. Despite the fact that the clustering problem is extremely well... |

36 |
All pairs shortest paths for graphs with small integer length edges
- Galil, Margalit
- 1997
(Show Context)
Citation Context ...urvey [35] and the references therein). We remark that many of the algorithms for all-pairs-(almost)shortest-paths in the traditional model of computation rely heavily either on matrix-multiplication =-=[2, 23, 34]-=-, or on multiple BFS traversals of the entire graph [1, 17, 15]. Both these approaches appear to be completely unsuitable for the streaming model of computation.sNote also that the space complexity of... |

35 | All-pairs small stretch paths
- Cohen, Zwick
(Show Context)
Citation Context ..., and write a new path on top of it. We remark that in the traditional (not streaming) model of computation, the problem of computing almost shortest paths was the focus of extensive recent research (=-=[17, 15, 18]-=-; see also the excellent survey [35] and the references therein). We remark that many of the algorithms for all-pairs-(almost)shortest-paths in the traditional model of computation rely heavily either... |

34 | A simple linear time algorithm for computing a (2k − 1)-spanner of O(n 1+1/k ) size in weighted graphs
- Baswana, Sen
- 2003
(Show Context)
Citation Context |

32 | Polylog-time and near-linear work approximation scheme for undirected shortest paths
- Cohen
(Show Context)
Citation Context ...m is an improved version of the algorithm of [18]. The distributed algorithm of [18] is based on the following sequential construction (which dates back to the parallel construction of hopsets due to =-=[14]-=-). We now give a high level view of the sequential algorithm. Given a graph, a (κ, W )-cover can be constructed for the graph. Intuitively, there are only a small number of large clusters (cluster con... |

31 | 1 + ɛ, β)-Spanner Constructions for General Graphs - Elkin, Peleg |

29 | Roundtrip spanners and roundtrip routing in directed graphs
- Roditty, Thorup, et al.
(Show Context)
Citation Context |

27 |
An improved data stream algorithm for frequency moments
- Coppersmith, Kumar
- 2004
(Show Context)
Citation Context ...rtant for processing massive data sets (see, e.g., [28], and the references therein). While originally the research in this area focused on numerical problems such as estimating the frequency moments =-=[3, 16]-=-, and L 1 - and L 2 -distances [22, 26], more recently the problems involving massive graphs have become a subject of study [28, 21]. In this paper we show that our distributed algorithm can be conver... |

18 |
Adapting to asynchronous dynamic networks
- Awerbuch, Patt-Shamir, et al.
- 1992
(Show Context)
Citation Context ...enter of C ′ and the center of C ′′ . 3. Add to the spanner all the edges of the set ¡ C∈¢ ℓ+1 EG(C). Note that the above algorithm is a synchronous protocol. By constructing and using a synchronizer =-=[5]-=-, this protocol can be converted into an asynchronous protocol. 3. ANALYSIS OF THE TIME AND MESSAGE COMPLEXITY In this section we analyze the time and message complexity of Algorithm 2.3. In the distr... |

10 |
On finding common neighborhoods in massive graphs
- Buchsbaum, Giancarlo, et al.
- 2003
(Show Context)
Citation Context ...(|E| · n ρ ). 4. ADAPTATION TO THE STREAMING MODEL In this section we adapt Algorithm 2.3 to the streaming model and devise an algorithm for computing all-pairsalmost-shortest-paths in this model. In =-=[8, 21]-=-, it was shown that in the streaming model, finding the common neighborhood (distance 1 neighborhood) of two vertices requires Ω(n) bits of space in constant number of passes. We extend this result to... |

8 |
Fast algorithms for t-spanners and stretch-t paths
- Cohen
- 1999
(Show Context)
Citation Context |

8 |
Data streams: Algorithms and applications. 2003. Available at http://athos.rutgers.edu/∼muthu/stream-1-1.ps
- Muthukrishnan
(Show Context)
Citation Context ... input sequentially using a small number of passes through the input. This computational model was developed in the works of [3, 25, 22], and is important for processing massive data sets (see, e.g., =-=[28]-=-, and the references therein). While originally the research in this area focused on numerical problems such as estimating the frequency moments [3, 16], and L 1 - and L 2 -distances [22, 26], more re... |