Title :
OPEX simulation study of PONs based on network geometric and Markov cost models
Author :
Fernandez, Alicia ; Stol, Norvald
Author_Institution :
Dept. of Telematics, NTNU, Trondheim, Norway
Abstract :
Passive Optical Networks (PONs) are regarded as the preferred technology for broadband access networks. Lately, the dependability of PONs has become crucial as end-users expect access networks to be highly reliable. However, operators are more concerned with reducing the costs associated to failures. These costs include repair costs, payment of penalties due to excessive disconnection time and costs related to loss of reputation. This paper performs an exhaustive analysis of these dependability-related costs. To carry out the analysis, a Markov cost model, capturing the dynamic behavior of the system, is built based on a network geometric model describing the PON deployment area. This way, the impact of the network physical layout and infrastructure sharing is also covered by means of the geometric model. The whole model is evaluated by means of Monte Carlo simulations, and the performance, in terms of expected costs, of different protection schemes is assessed. Hence, the best suited protection scheme is identified, while the effects of variables such as number of repair crews, design decisions and user profile (residential or business) are also pinpointed.
Keywords :
Markov processes; Monte Carlo methods; broadband networks; passive optical networks; telecommunication network reliability; Markov cost model; Monte Carlo simulations; OPEX simulation; PON deployment; broadband access networks; exhaustive analysis; network geometric model; passive optical networks; Biological system modeling; Business; Maintenance engineering; Markov processes; Mathematical model; Passive optical networks; Ports (Computers); Markov cost model; Operational Expenditures; Passive Optical Networks; network geometric model; protection;
Conference_Titel :
Optical Network Design and Modeling, 2014 International Conference on
Conference_Location :
Stockholm