Please use this identifier to cite or link to this item: http://localhost/handle/Hannan/579498
Title: Unified Importance Sampling Schemes for Efficient Simulation of Outage Capacity Over Generalized Fading Channels
Authors: Nadhir Ben Rached;Abla Kammoun;Mohamed-Slim Alouini;Raul Tempone
subject: hazard rate twisting|naive Monte Carlo|outage capacity|bounded relative error|importance sampling|Asymptotic optimality
Year: 2016
Publisher: IEEE
Abstract: The outage capacity (OC) is among the most important performance metrics of communication systems operating over fading channels. Of interest in the present paper is the evaluation of the OC at the output of the Equal Gain Combining (EGC) and the Maximum Ratio Combining (MRC) receivers. In this case, it can be seen that this problem turns out to be that of computing the Cumulative Distribution Function (CDF) for the sum of independent random variables. Since finding a closed-form expression for the CDF of the sum distribution is out of reach for a wide class of commonly used distributions, methods based on Monte Carlo (MC) simulations take pride of price. In order to allow for the estimation of the operating range of small outage probabilities, it is of paramount importance to develop fast and efficient estimation methods as naive MC simulations would require high computational complexity. In this line, we propose in this work two unified, yet efficient, hazard rate twisting Importance Sampling (IS) based approaches that efficiently estimate the OC of MRC or EGC diversity techniques over generalized independent fading channels. The first estimator is shown to possess the asymptotic optimality criterion and applies for arbitrary fading models, whereas the second one achieves the well-desired bounded relative error property for the majority of the well-known fading variates. Moreover, the second estimator is shown to achieve the asymptotic optimality property under the particular Log-normal environment. Some selected simulation results are finally provided in order to illustrate the substantial computational gain achieved by the proposed IS schemes over naive MC simulations.
URI: http://localhost/handle/Hannan/181520
http://localhost/handle/Hannan/579498
ISSN: 1932-4553
1941-0484
volume: 10
issue: 2
Appears in Collections:2016

Files in This Item:
File Description SizeFormat 
7328688.pdf3.75 MBAdobe PDFThumbnail
Preview File
Title: Unified Importance Sampling Schemes for Efficient Simulation of Outage Capacity Over Generalized Fading Channels
Authors: Nadhir Ben Rached;Abla Kammoun;Mohamed-Slim Alouini;Raul Tempone
subject: hazard rate twisting|naive Monte Carlo|outage capacity|bounded relative error|importance sampling|Asymptotic optimality
Year: 2016
Publisher: IEEE
Abstract: The outage capacity (OC) is among the most important performance metrics of communication systems operating over fading channels. Of interest in the present paper is the evaluation of the OC at the output of the Equal Gain Combining (EGC) and the Maximum Ratio Combining (MRC) receivers. In this case, it can be seen that this problem turns out to be that of computing the Cumulative Distribution Function (CDF) for the sum of independent random variables. Since finding a closed-form expression for the CDF of the sum distribution is out of reach for a wide class of commonly used distributions, methods based on Monte Carlo (MC) simulations take pride of price. In order to allow for the estimation of the operating range of small outage probabilities, it is of paramount importance to develop fast and efficient estimation methods as naive MC simulations would require high computational complexity. In this line, we propose in this work two unified, yet efficient, hazard rate twisting Importance Sampling (IS) based approaches that efficiently estimate the OC of MRC or EGC diversity techniques over generalized independent fading channels. The first estimator is shown to possess the asymptotic optimality criterion and applies for arbitrary fading models, whereas the second one achieves the well-desired bounded relative error property for the majority of the well-known fading variates. Moreover, the second estimator is shown to achieve the asymptotic optimality property under the particular Log-normal environment. Some selected simulation results are finally provided in order to illustrate the substantial computational gain achieved by the proposed IS schemes over naive MC simulations.
URI: http://localhost/handle/Hannan/181520
http://localhost/handle/Hannan/579498
ISSN: 1932-4553
1941-0484
volume: 10
issue: 2
Appears in Collections:2016

Files in This Item:
File Description SizeFormat 
7328688.pdf3.75 MBAdobe PDFThumbnail
Preview File
Title: Unified Importance Sampling Schemes for Efficient Simulation of Outage Capacity Over Generalized Fading Channels
Authors: Nadhir Ben Rached;Abla Kammoun;Mohamed-Slim Alouini;Raul Tempone
subject: hazard rate twisting|naive Monte Carlo|outage capacity|bounded relative error|importance sampling|Asymptotic optimality
Year: 2016
Publisher: IEEE
Abstract: The outage capacity (OC) is among the most important performance metrics of communication systems operating over fading channels. Of interest in the present paper is the evaluation of the OC at the output of the Equal Gain Combining (EGC) and the Maximum Ratio Combining (MRC) receivers. In this case, it can be seen that this problem turns out to be that of computing the Cumulative Distribution Function (CDF) for the sum of independent random variables. Since finding a closed-form expression for the CDF of the sum distribution is out of reach for a wide class of commonly used distributions, methods based on Monte Carlo (MC) simulations take pride of price. In order to allow for the estimation of the operating range of small outage probabilities, it is of paramount importance to develop fast and efficient estimation methods as naive MC simulations would require high computational complexity. In this line, we propose in this work two unified, yet efficient, hazard rate twisting Importance Sampling (IS) based approaches that efficiently estimate the OC of MRC or EGC diversity techniques over generalized independent fading channels. The first estimator is shown to possess the asymptotic optimality criterion and applies for arbitrary fading models, whereas the second one achieves the well-desired bounded relative error property for the majority of the well-known fading variates. Moreover, the second estimator is shown to achieve the asymptotic optimality property under the particular Log-normal environment. Some selected simulation results are finally provided in order to illustrate the substantial computational gain achieved by the proposed IS schemes over naive MC simulations.
URI: http://localhost/handle/Hannan/181520
http://localhost/handle/Hannan/579498
ISSN: 1932-4553
1941-0484
volume: 10
issue: 2
Appears in Collections:2016

Files in This Item:
File Description SizeFormat 
7328688.pdf3.75 MBAdobe PDFThumbnail
Preview File