A comparison of different routing schemes for the robust network loading problem: polyhedral results and computation
read more
Citations
Staffing and scheduling flexible call centers by two-stage robust optimization
An optimization model for robust FSO network dimensioning
Solving the bifurcated and nonbifurcated robust network loading problem with k-adaptive routing
A Polyhedral Study of the Robust Capacitated Edge Activation Problem
References
The Price of Robustness
The price of the robustness
Adjustable robust solutions of uncertain linear programs
Solving two-stage robust optimization problems using a column-and-constraint generation method
Related Papers (5)
Frequently Asked Questions (12)
Q2. What future works have the authors mentioned in the paper "A comparison of different routing schemes for the robust network loading problem: polyhedral results and computation" ?
The main scope of the paper is to study the Robust Network Loading Problem. The authors discussed two classes of inequalities ( non-negativity constraints and robust cutset inequalities ) that can be proved to be facet defining under the same assumptions in all the considered settings. Namely, their results suggest that volume routing yields cost reductions close to those obtained using dynamic routing but requires computational times similar to those obtained for static routing. While for static routing compact formulations can be as fast as Benders decomposition algorithms, the situation is different for volume routing for which Benders decomposition clearly outperforms the compact formulation.
Q3. What is the way to solve volume routing?
their results suggest that volume routing yields cost reductions close to those obtained using dynamic routing but requires computational times similar to those obtained for static routing.
Q4. What is the author's approach to the problem with dynamic routing and splittable flows?
In [33] the author studies the problem with dynamic routing and splittable flows under the Hose model, proposing a branch-and-cut procedure related to bilevel optimization.
Q5. What is the definition of affine routing?
Affine routing [41] restricts the flows to be affine functions of the demands, as it applies to network optimization what has long been known as affine decision rules in adjustable robust optimization [14].
Q6. How does the paper show that volume routing is able to reduce costs?
The authors also show that volume routing yields cost reductions over static routing in half of the instances, while not requiring more computational time.
Q7. What is the trade-off between flexibility and tractability?
Volume routing seems to offer the best trade-off between flexibility and tractability, while requiring as little information as static routing when it comes to decentralized implementations.
Q8. What is the way to generate robust cutset inequalities?
The authors generate robust cutset and 3-partition inequalities, according to one of the following configurations:0 : No cuts.1 : Heuristic and exact separation of robust cutset inequalities at the root node only.
Q9. How many instances of volume routing can be solved?
Regarding the solution costs, the authors see that volume routing yields a positive cost reduction in 26 instances out of 48, which ranges up to 12.8 %.
Q10. What is the difference between compact and static routing?
compact formulations are less efficient for the problem with volume routing than they are for the one with static routing, which is probably due to the larger number of variables and constraints present in fV RNL.
Q11. How many instances could not be solved within the time limit?
Dynamic routing is, as expected, harder to solve than static and volume and 21 instances could not be solved within the time limit.
Q12. What is the way to separate inequalities?
The authors note that one could directly separate inequalities with integer µ and b being the upper integer of the corresponding Benders cut (also known as rounded metric inequalities).