Download presentation
Presentation is loading. Please wait.
Published bySusanne Guttormsen Modified over 6 years ago
1
Closed-form Algorithms in Hybrid Positioning: Myths and Misconceptions
Niilo Sirola Department of Mathematics Tampere University of Technology, Finland (currently with Taipale Telematics, Finland) Workshop for Positioning, Navigation and Communication 2010
2
Workshop for Positioning, Navigation and Communication 2010
Mobile positioning Given a measurement model y = h(x) + v and the measurements y Find the position x that fits the measurements ”best” Workshop for Positioning, Navigation and Communication 2010
3
Iterative Least Squares
The Gauss-Newton or Taylor series or Iterative/Ordinary/Nonlinear least squares Usual objections to Gauss-Newton initial guess: ”Selection of such a starting point is not simple in practice” convergence is not assured computational load: ”as LS computation is required in each iteration” Workshop for Positioning, Navigation and Communication 2010
4
Examples of closed-form methods
geometrically inspired methods – easy to explain and visualise ”replace each intersection with a line then solve the linear LS problem” Workshop for Positioning, Navigation and Communication 2010
5
Examples of closed-form methods
others are algebraic and more rigorous sometimes come with a proof sometimes can be implemeted by the reader Workshop for Positioning, Navigation and Communication 2010
6
Least-squares vs least-quartic solution
Least-squares solution: Find x such that ‖y – h(x)‖2 is as small as possible Least-quartic solution: ‖y2 – h(x) 2‖2 easier to solve analytically, but the solution is not least squares solution -> is non-optimal in variance sense Workshop for Positioning, Navigation and Communication 2010
7
Workshop for Positioning, Navigation and Communication 2010
Closed-form methods? Some are not even in closed form…. ”…first assume there is no relationship between x,y, and r1 … The final solution is obtained by imposing the relationship.. via another LS computation” ”we can first use (14) to obtain an initial solution…” Workshop for Positioning, Navigation and Communication 2010
8
Workshop for Positioning, Navigation and Communication 2010
Testing method testing range-only methods 12 by 12 kilometer simulated test field, six ranging beacons independent and identically distributed Gaussian measurement noise noise sigma sweeps from 1 m to 10 km 1000 position fixes with random true position for each noise level Workshop for Positioning, Navigation and Communication 2010
9
Workshop for Positioning, Navigation and Communication 2010
Tested algorithms Candidate algorithms: ignore measurements – use the center of stations simple intersection range-Bancroft Gauss-Newton (with and without regularisation) Cheung (2006) Beck (2008) All implemented in Matlab with similar level of optimization Workshop for Positioning, Navigation and Communication 2010
10
Results: RMS position error
Workshop for Positioning, Navigation and Communication 2010
11
Results: normalized error
Workshop for Positioning, Navigation and Communication 2010
12
Back to the objections against GN…
1) requires an initial guess so do several ”closed-form” methods in practical applications rough position usually known from the context: physical constraints, station positions, etc. possible to use a closed-form solution as a starting point Workshop for Positioning, Navigation and Communication 2010
13
2) Convergence not guaranteed
Depends on the quality of the initial guess Regularisation helps Sanity checks recommended - probably should use some with closed-form methods as well! Workshop for Positioning, Navigation and Communication 2010
14
3) Computational complexity
Matlab on a 1.4GHz Celeron laptop station mean: instant simple intersection: 0.3 ms/fix Bancroft: 0.5 ms/fix Cheung: 0.8 ms/fix Beck: 1.2 ms/fix Gauss-Newton: 1.2 – 1.5 ms/fix Workshop for Positioning, Navigation and Communication 2010
15
Workshop for Positioning, Navigation and Communication 2010
Bonus: flexibility Closed-form methods: only ranges: ok only range differences: ok mixed ranges and range differences: some choices range differences + a plane: at least one method mixed ranges, range differences, planes, etc: … huh? Gauss-Newton: combination of any (differentiable) measurements: OK Workshop for Positioning, Navigation and Communication 2010
16
Workshop for Positioning, Navigation and Communication 2010
Conclusions Gauss-Newton was found to be competitive against several closed-form solutions Additional bonus points: Handles also correlated noise Robust numerics Gives an error estimates Extends to time series -> Extended Kalman Filter Questions? Workshop for Positioning, Navigation and Communication 2010
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.