#### DMCA

## Dynamic Pricing for Bandwidth Provisioning (2002)

Venue: | in Proceedings of the 36th Annual Conference on Information Sciences and Systems |

Citations: | 5 - 2 self |

### Citations

1900 |
Markov Decision Processes: Discrete Stochastic Dynamic Programming
- Puterman
- 1994
(Show Context)
Citation Context ... In Section II-C, we converted the pricing POMDP to a belief-state MDP. But this beliefstate MDP has an uncountable state space, and thus techniques such as value iteration or linear programming (see =-=[10]-=-) cannot be applied in this case. There are a number of heuristic techniques for solving such MDPs [11][12][13][14]. Here we explore the policyrollout technique developed by Bertsekas and Castanon [13... |

1179 |
Nonlinear programming
- Bertsekas
- 1999
(Show Context)
Citation Context ... uncountable state space, and thus techniques such as value iteration or linear programming (see [10]) cannot be applied in this case. There are a number of heuristic techniques for solving such MDPs =-=[11][12][-=-13][14]. Here we explore the policyrollout technique developed by Bertsekas and Castanon [13] because of the simplicity of implementation and availability of an obvious \base policy" for rollout.... |

350 | Resource pricing and the evolution of congestion control. Automatica 35
- Gibbens, Kelly
- 1998
(Show Context)
Citation Context ...r unused bandwidth. A key problem in completing such trades is to have a good online pricing scheme. Various pricing schemes have been discussed in the literature for scenarios withsxed sets of users =-=[1]-=-[2][3]. Pricing schemes appropriate for various problems with dynamic user arrivals have also been developed under various network settings [4][5][6]. But these schemes deal with restrictive trac mode... |

156 | Tsitsiklis. Congestion-Dependent Pricing of Network Services.
- Paschalidis, N
- 2000
(Show Context)
Citation Context ...d in the literature for scenarios withsxed sets of users [1][2][3]. Pricing schemes appropriate for various problems with dynamic user arrivals have also been developed under various network settings =-=[4]-=-[5][6]. But these schemes deal with restrictive trac models (e.g. Poisson arrivals) and were designed forsxed-bandwidth dialup connections. Our model more naturally allows the users' bandwidth demands... |

147 |
Perturbation Analysis of Discrete Event Dynamic Systems
- Ho, Cao
- 1991
(Show Context)
Citation Context ...ny times, and every time we estimate the gradient, we take a step in the direction of the gradient. A step size of k such that P k = 1 and P 2 ks1 is known to be appropriate for such algorithms [9]. The derivative of Vsat H (p) along an observed sample path can be computed in an incremental fashion as follows. Dene vL (), for 0 Ls, recursively as follows: v 0 (p) = X fi2A[0;0]g D(p; i) ... |

103 | Tsitsiklis. Simulation-based optimization of Markov reward processes.
- Marbach, N
- 1999
(Show Context)
Citation Context ...ountable state space, and thus techniques such as value iteration or linear programming (see [10]) cannot be applied in this case. There are a number of heuristic techniques for solving such MDPs [11]=-=[12][13][-=-14]. Here we explore the policyrollout technique developed by Bertsekas and Castanon [13] because of the simplicity of implementation and availability of an obvious \base policy" for rollout. Her... |

98 | Rollout algorithms for stochastic scheduling problems
- Bertsekas, Castanon
- 1999
(Show Context)
Citation Context ...able state space, and thus techniques such as value iteration or linear programming (see [10]) cannot be applied in this case. There are a number of heuristic techniques for solving such MDPs [11][12]=-=[13][14].-=- Here we explore the policyrollout technique developed by Bertsekas and Castanon [13] because of the simplicity of implementation and availability of an obvious \base policy" for rollout. Here we... |

35 | A framework for simulation-based network control via hindsight optimization
- Chong, Givan, et al.
- 2000
(Show Context)
Citation Context ... state space, and thus techniques such as value iteration or linear programming (see [10]) cannot be applied in this case. There are a number of heuristic techniques for solving such MDPs [11][12][13]=-=[14]. Her-=-e we explore the policyrollout technique developed by Bertsekas and Castanon [13] because of the simplicity of implementation and availability of an obvious \base policy" for rollout. Here we des... |

11 |
Optimization control { I: basic algorithm and convergence
- Low, Lapsley
- 1999
(Show Context)
Citation Context ...nused bandwidth. A key problem in completing such trades is to have a good online pricing scheme. Various pricing schemes have been discussed in the literature for scenarios withsxed sets of users [1]=-=[2]-=-[3]. Pricing schemes appropriate for various problems with dynamic user arrivals have also been developed under various network settings [4][5][6]. But these schemes deal with restrictive trac models ... |

8 | Pricing of dialup services: an example of congestion-dependent pricing
- Patek, Campos-Nanez
- 2000
(Show Context)
Citation Context ...n the literature for scenarios withsxed sets of users [1][2][3]. Pricing schemes appropriate for various problems with dynamic user arrivals have also been developed under various network settings [4]=-=[5]-=-[6]. But these schemes deal with restrictive trac models (e.g. Poisson arrivals) and were designed forsxed-bandwidth dialup connections. Our model more naturally allows the users' bandwidth demands to... |

4 |
State-dependent pricing and its economic implications
- Wang, Peha
- 1999
(Show Context)
Citation Context ...he literature for scenarios withsxed sets of users [1][2][3]. Pricing schemes appropriate for various problems with dynamic user arrivals have also been developed under various network settings [4][5]=-=[6]-=-. But these schemes deal with restrictive trac models (e.g. Poisson arrivals) and were designed forsxed-bandwidth dialup connections. Our model more naturally allows the users' bandwidth demands to ch... |

3 |
Pricing Priority Classes in a Di#erentiated Services Network, submitted for publication, available as a
- Marbach
- 1999
(Show Context)
Citation Context ...ed bandwidth. A key problem in completing such trades is to have a good online pricing scheme. Various pricing schemes have been discussed in the literature for scenarios withsxed sets of users [1][2]=-=[3]-=-. Pricing schemes appropriate for various problems with dynamic user arrivals have also been developed under various network settings [4][5][6]. But these schemes deal with restrictive trac models (e.... |

3 | Error bounds for rolling-horizon policies in general markov control processes - Hernandez-Lerma, Lasserre - 1990 |

2 |
Optimal pricing for brokermediated bandwidth market
- Savagaonkar, Chong, et al.
- 2001
(Show Context)
Citation Context ...g notation. If ~ is a vector, i denotessi th component of ~ . Contrarily, if is a scalar, i denotes i th power of . Also, for brevity, we present the proofs of the various results elsewhere [7]. II Problem Formulation A Problem description We consider a dynamic market in which the resource being traded is the bandwidth over a single link. We assume that the maximum bandwidth available on th... |