#### DMCA

## Structured Admission Control Policy in Heterogeneous Wireless Networks with Mesh Underlay

Citations: | 3 - 3 self |

### Citations

1848 | Markov Decision Processes Discrete Stochastic Dynamic Programming - Puterman - 1994 |

343 |
Applied Probability Models with Optimization Applications
- Ross
- 1970
(Show Context)
Citation Context ...ller nodes to improve the system performance. Through numerical analysis and simulations, we show that the proposed algorithm has a convergence speed much faster than that of the Value Iteration (VI) =-=[13]-=- algorithm and leads to a CAC scheme significantly more cost efficient than the Complete Sharing and Guard-Channel policies. The rest of the paper is organized as follows. Related works and their cont... |

251 | Exploiting structure in policy construction
- Boutilier, Dearden, et al.
- 1995
(Show Context)
Citation Context ...rmulation is used to characterize the structure of possible optimal policies. Then, knowledge of the policy structure can be exploited to design efficient numerical methods to find the optimal policy =-=[10]-=-. As an example, in [4], it is shown that the optimal CAC policy for a single cellular Base-Station (BS) is the well-known guard-channel policy, which is fully determined by a single threshold that ca... |

165 | On optimal call admission control in cellular networks
- Ramjee, Nagarajan, et al.
- 1997
(Show Context)
Citation Context ...ce-sharing systems. CAC schemes for homogeneous cellular networks have been extensively studied. These schemes can be classified into nearoptimal heuristics [3] and decision-theoretic optimal methods =-=[4]-=-. Dynamic Programming (DP) and Markov Decision Processes (MDP) [5] are often used in the design of optimal CAC algorithms. However, for almost all realistic modelings of networking systems, the comput... |

159 | On the complexity of solving Markov decision problems
- Littman, Dean, et al.
- 1995
(Show Context)
Citation Context ...odelings of networking systems, the computational load of finding an optimal policy by MDP algorithms is very high. In fact, there is no known strongly-polynomial time algorithm to solve MDP problems =-=[6]-=-. This can hinder the application of optimal CAC schemes in practical scenarios. Furthermore, in HWNs with their increased complexity, the “curse of dimensionality” problem [7] makes it impractical to... |

109 |
Local search in combinatorial optimization
- AARTS, LENSTRA
- 1997
(Show Context)
Citation Context ...t criterion and combinatorial search. In this section, we propose an efficient algorithm which determines the parameters of an optimal threshold policy using combinatorial localized search algorithms =-=[31]-=-. Given a CAC policy Ψ, the system can be analyzed as a Markov chain with states (i, n) as depicted in Fig. 3. This Markov chain still has a large state space. However, observing the separation betwee... |

78 |
The stochastic knapsack problem
- Ross, Tsang
- 1989
(Show Context)
Citation Context ...ly apply admission control methods that are previously used in homogeneous wireless networks. A more effective use of DP-based methods is to obtain structural results for optimal control problems [7]–=-=[9]-=-. In this approach, a DP formulation is used to characterize the structure of possible optimal policies. Then, knowledge of the policy structure can be exploited to design efficient numerical methods ... |

61 | Integration of 802.11 and third-generation wireless data networks
- Buddhikot, Chandranmenon, et al.
- 2003
(Show Context)
Citation Context ...rogeneous Wireless Networks (HWN) consist of several layers of different overlapping wireless access technologies such as an integration of WiMAX or 3G at the overlay with IEEE 802.11 at the underlay =-=[2]-=-. This multi-layer architecture provides users with the option of choosing between different available wireless access technologies based on users’ traffic profile, mobility pattern, and QoS preferenc... |

55 |
ªChannel Occupancy Times and Handoff Rate for Mobile Computing
- Fang, Chlamtac, et al.
- 1998
(Show Context)
Citation Context ...able I: η i,j hw = φ(i, j)ηi h and ηi hwc = φ(i, 0)ηi h , where index 0 is reserved for cellular overlay. It is clear that the above memoryless assumptions result in exponential channel holding times =-=[25]-=-. In this paper, we focus on call-level QoS, which is common in CAC literature. Further, the fixed channel allocation (FCA) scheme is used. FCA easily applies to various wireless technologies with cha... |

50 |
On time and space decomposition f complex structures
- Courtois
- 1985
(Show Context)
Citation Context ...v) is in state k can now be calculated as ∑ ˆλ k vλ (h\n)o(⃗u)/(λho(⃗u) + λno(⃗u)). (17) λ (k) (h\n)v = ⃗u∈U(k) To find birth-death rates, we use the Markov chain state aggregation method proposed in =-=[30]-=-. The transition rate from subset U(i) to U(j) denoted by qv(i, j) can be found as qv(i, j) = ∑ ∑ ⃗x∈U(i) ⃗y∈U(j) qt(⃗x, ⃗y) πt(⃗x) π U(i) (18) where QT = [qt(⃗x, ⃗y)]. The birth-death rates can be ca... |

43 | A Dynamic Call Admission Policy with Precision QoS Guarantee using Stochastic Control for Mobile Wireless Networks
- Wu, Wong, et al.
- 2002
(Show Context)
Citation Context ...ts and how to reserve resources in a resource-sharing systems. CAC schemes for homogeneous cellular networks have been extensively studied. These schemes can be classified into nearoptimal heuristics =-=[3]-=- and decision-theoretic optimal methods [4]. Dynamic Programming (DP) and Markov Decision Processes (MDP) [5] are often used in the design of optimal CAC algorithms. However, for almost all realistic ... |

43 |
Personal communication systems using multiple hierarchical cellular overlays
- Hu, Rappaport
- 1995
(Show Context)
Citation Context ...12]. Interrupted Poisson Processes (IPP) and Markov-Modulated Poisson Processes (MMPP) have been used to model the behavior of overflow processes in the context of hierarchical cellular networks [16]–=-=[18]-=-, [24]. A common assumption among these studies is that the traffic statistics of all micro-cells are equal, i.e., they have identical arrival and hand-off rates. This reduces analysis complexity by a... |

33 |
The interrupted poisson process as an overflow process
- Kuczura
- 1973
(Show Context)
Citation Context ...ar networks. Different models with various complexities are studied in [16]–[19]. These models are built upon prior research on the the characterization of overflow processes for loss queuing systems =-=[20]-=-–[22]. Initially, the overflow traffic has been modeled as a Poisson process [23], however, it was later shown that this approximation is inaccurate, and that by failing to model the burstiness of the... |

32 |
Tijms. A First Course in Stochastic Models
- C
- 2003
(Show Context)
Citation Context ... problem. This reflects our concern about long-run QoS performance. In addition, since the decision epochs can be at any randomly distributed time, a Semi-Markov Decision Process (SMDP) model is used =-=[27]-=-. In what follows, we present a general framework that will be used throughout the rest of the paper. We start with a finitehorizon optimal cost function, and we show that the solution to the infinite... |

27 |
The Analysis of A Queue Arising in Overflow Models
- Meier-Hellstern
- 1989
(Show Context)
Citation Context ...ean of off-time period. An IPP can be expressed as an MMPP with two states as QIP P = [ ] −β β , ΛIP P = α −α [ 0 0 0 λ ] . (10) To find the equivalent IPP of an MMPP, we employ the technique used in =-=[21]-=-. The method is to match the first two noncentral moments of the arrival rate of the MMPP and an appropriately chosen time constant. Let us denote by πQi the stationary probability vector associated w... |

27 | A survey of Markov decision models for control of networks of queues. Queueing Systems
- Weber
- 1993
(Show Context)
Citation Context ... is triggered by event e. If this cost is less than blocking cost C B(e), the call will be admitted and otherwise it gets rejected. More general information on similar frameworks can be found in [4], =-=[28]-=-. It has been shown in [27] that for average-cost problems with finite S and A, the optimal policy is stationary. Further, we are only interested in stationary policies which result in irreducible cha... |

23 | T and Koole G, “On Optimal Call Admission Control in a Resource-Sharing System - Altman, Jiménez - 2001 |

13 |
Performance Analysis and Overflowed Traffic Characterization in Multi-Service Hi- erarchical Wireless Networks
- Chung, Lee
(Show Context)
Citation Context ... call overflows to overlay, it stays at that layer until its completion, i.e., repacking is not allowed. This common assumption allows one to analyze the underlay independently of the overlay network =-=[17]-=-, [18]. Note that the no-repacking policy serves as an indirect scheme for classification and layer assignment of users with different velocities. In particular, calls made by a highly mobile user is ... |

10 |
Teletraffic analysis of hierarchical cellular network
- Lagrange, Godlewski
- 1995
(Show Context)
Citation Context ...al reservation policy. Another related research direction is the modeling of overflow traffic for hierarchical macro/micro-cellular networks. Different models with various complexities are studied in =-=[16]-=-–[19]. These models are built upon prior research on the the characterization of overflow processes for loss queuing systems [20]–[22]. Initially, the overflow traffic has been modeled as a Poisson pr... |

6 | Optimal and Structured Call Admission Control Policies for Resource-Sharing Systems
- Ni, Tsang, et al.
- 2007
(Show Context)
Citation Context ...m to solve MDP problems [6]. This can hinder the application of optimal CAC schemes in practical scenarios. Furthermore, in HWNs with their increased complexity, the “curse of dimensionality” problem =-=[7]-=- makes it impractical to directly apply admission control methods that are previously used in homogeneous wireless networks. A more effective use of DP-based methods is to obtain structural results fo... |

6 |
Handoff Analysis of the Hierarchical Cellular System
- Wie, Jang, et al.
- 2000
(Show Context)
Citation Context ...nterrupted Poisson Processes (IPP) and Markov-Modulated Poisson Processes (MMPP) have been used to model the behavior of overflow processes in the context of hierarchical cellular networks [16]–[18], =-=[24]-=-. A common assumption among these studies is that the traffic statistics of all micro-cells are equal, i.e., they have identical arrival and hand-off rates. This reduces analysis complexity by allowin... |

5 |
Overflow analysis for finite waiting room systems
- Guérin, Lien
- 1990
(Show Context)
Citation Context ...ows. A new or a hand-over call overflows to the overlay when it cannot be accommodated by the target mesh Access Point (AP) at the underlay. It is well known that this overflow traffic is not Poisson =-=[12]-=-. In order to characterize the overflow traffic from the underlaying mesh to the overlay, a Partially-Observable Markov-Modulated Poisson Process (PO-MMPP) traffic model is developed. This model captu... |

5 |
Approximation of loss calculation for hierarchical networks with multiservice overflows
- Huang, Ko
- 2008
(Show Context)
Citation Context ...servation policy. Another related research direction is the modeling of overflow traffic for hierarchical macro/micro-cellular networks. Different models with various complexities are studied in [16]–=-=[19]-=-. These models are built upon prior research on the the characterization of overflow processes for loss queuing systems [20]–[22]. Initially, the overflow traffic has been modeled as a Poisson process... |

5 | Performance analysis of microcellization for supporting two mobility classes in cellular wireless networks
- Maheshwari, Kumar
- 2000
(Show Context)
Citation Context ... These models are built upon prior research on the the characterization of overflow processes for loss queuing systems [20]–[22]. Initially, the overflow traffic has been modeled as a Poisson process =-=[23]-=-, however, it was later shown that this approximation is inaccurate, and that by failing to model the burstiness of the overflow traffic, it underestimates the blocking of overflow calls at the overla... |

3 |
On Optimal Admission Control for Multi-Service Cellular/WLAN
- Stevens-Navarro, Wong
(Show Context)
Citation Context ...coming from different layers. In [14], admission control for voice and data services in a HWNs is considered and an iterative algorithm is proposed to find the dropping and blocking probabilities. In =-=[15]-=-, the minimization of linear cost functions is used to find the number of guard channels. In both sets of results the reservation scheme is chosen a priori. This is in contrast to our work, in which w... |

3 |
Friedhoffer, “Mobility management
- Varma, Ramesh, et al.
(Show Context)
Citation Context ...me instant, and as such do not affect our analysis of QoS for guaranteed-service call requests. B. Call Handling Policy In [11], we assumed tight coupling between different layers of wireless network =-=[26]-=- in a cluster, where the management of different layers is centralized. However, when overlay can comprise mesh networks with multiple APs, it is impractical to assume centralized control, because the... |

2 |
QoS provisioning via admission control in cellular/wireless LAN interworking
- Song, Zhuang
- 2005
(Show Context)
Citation Context ...ON There are several studies on QoS for HWNs where the conventional guard-channel policy is used and the available channels are divided between handover and new calls coming from different layers. In =-=[14]-=-, admission control for voice and data services in a HWNs is considered and an iterative algorithm is proposed to find the dropping and blocking probabilities. In [15], the minimization of linear cost... |

1 |
Efficient structured policies for admission control in heterogeneous wireless networks
- Farbod, Liang
- 2007
(Show Context)
Citation Context ...e optimal CAC policy for a single cellular Base-Station (BS) is the well-known guard-channel policy, which is fully determined by a single threshold that can be computed by an efficient algorithm. In =-=[11]-=-, the authors study a simple two-tier HWN with a single underlay network and show that the optimal CAC policy has a two-dimensional threshold structure. However, such results are not directly applicab... |

1 |
Performance analysis for hierarchical multirate loss networks
- Chang, Hwang
- 2004
(Show Context)
Citation Context ... of P (i) (i) B , P D and λiT may involve iterations. For example, the Erlang Fixed-Point approximation has been used to find the traffic intensities and blocking probabilities in loss networks [22], =-=[29]-=-. We start with a set of values for P (i) (i) B and P D , and solve (7) to find the set of λi hw . Then, (6) is used to recompute a new set of P (i) (i) B and P D . This iteration is repeated until de... |