Name
Abstract | ||
|---|---|---|
AbstractEstimation-over-graphs (EoG) is a class of estimation problems that admit a natural graphical representation. Several key problems in robotics and sensor networks, including sensor network localization, synchronization over a group, and simultaneous localization and mapping (SLAM) fall into this category. We pursue two main goals in this work. First, we aim to characterize the impact of the graphical structure of SLAM and related problems on estimation reliability. We draw connections between several notions of graph connectivity and various properties of the underlying estimation problem. In particular, we establish results on the impact of the weighted number of spanning trees on the D-optimality criterion in 2D SLAM. These results enable agents to evaluate estimation reliability based only on the graphical representation of the EoG problem. We then use our findings and study the problem of designing sparse SLAM problems that lead to reliable maximum likelihood estimates through the synthesis of sparse graphs with the maximum weighted tree connectivity. Characterizing graphs with the maximum number of spanning trees is an open problem in general. To tackle this problem, we establish several new theoretical results, including the monotone log-submodularity of the weighted number of spanning trees. We exploit these structures and design a complementary greedy–convex pair of efficient approximation algorithms with provable guarantees. The proposed synthesis framework is applied to various forms of the measurement selection problem in resource-constrained SLAM. Our algorithms and theoretical findings are validated using random graphs, existing and new synthetic SLAM benchmarks, and publicly available real pose-graph SLAM datasets. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1177/0278364918823086 | Periodicals |
Keywords | Field | DocType |
SLAM, measurement selection, resource-constrained SLAM, estimation over graphs, pose-graph pruning, number of spanning trees, graph complexity | Graph,Control engineering,Theoretical computer science,Artificial intelligence,Wireless sensor network,Mathematics,Robotics | Journal |
Volume | Issue | ISSN |
38 | 2-3 | 0278-3649 |
Citations | PageRank | References |
7 | 0.49 | 33 |
Authors | ||
6 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Kasra Khosoussi | 1 | 32 | 6.15 |
| Matthew Giamou | 2 | 9 | 3.92 |
| Gaurav S. Sukhatme | 3 | 5469 | 548.13 |
| Shoudong Huang | 4 | 755 | 62.77 |
| Gamini Dissanayake | 5 | 2226 | 256.36 |
| Jonathan How | 6 | 1759 | 185.09 |