Abstract | ||
---|---|---|
This paper proposes a coded distributed graph processing framework to alleviate the communication bottleneck in large-scale distributed graph processing. In particular, we propose a topology-aware coded computing ( TACC) algorithm that has two salient features. First, we propose a topology-aware graph allocation strategy. Second, we propose a coded aggregation scheme that combines the intermediate computations for graph processes while constructing coded messages. The proposed setup builds on a trade-off between computation and communication, in that increasing the computation load at the distributed parties can in turn reduce the communication load. We demonstrate the effectiveness of the TACC algorithm by comparing the communication load with existing setups on a Google web graph for PageRank computations. In particular, we show that the proposed coding strategy can lead up to 82% improvement in reducing the communication load when compared to the state-of-the-art. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1109/icassp.2019.8682227 | 2019 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP) |
Keywords | Field | DocType |
Distributed computing, large-scale graph processing, graph signal filtering | Resource management,PageRank,Bottleneck,Mathematical optimization,Computer science,Coding (social sciences),Theoretical computer science,Sparse matrix,Computation,Encoding (memory),Salient | Conference |
ISSN | Citations | PageRank |
1520-6149 | 0 | 0.34 |
References | Authors | |
0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Bagak Guler | 1 | 0 | 0.34 |
Amir Salman Avestimehr | 2 | 1880 | 157.39 |
Antonio Ortega | 3 | 4720 | 493.26 |