The version presented here may differ from the published version or, version of record, if you wish to cite this item you are advised to consult the publisher's version. Please see the 'permanent WRAP URL' above for details on accessing the published version and note that access may require a subscription.

Although optical fiber sensor networks (OFSNs) attract more and more attentions, they are still facing the problem of lacking the methods for assessing network robustness quantitatively. This paper improves the recently presented OFSNs robustness evaluation model by proposing the robustness error assessment based on Monte Carlo method. Robustness simulations of several network topologies, including line, ring, star, bus, and double ring topologies, have been done with the model, proving that the concentric double ring topology proposed by Fernandez-Vallejo et al. is more robust than the other four topologies. Parameters required for robustness calculation in ring topology are listed in detail. Then, an experimental robustness assessment approach is established. According to it, eight-point fiber Bragg grating sensor networks with the four basic topologies are employed. Support vector machine is applied to process data. Both threshold and attenuation coefficient have been gotten from experiments. Comparison between experimental and simulative robustness verifies the practicability of the robustness evaluation method, and also proves that ring and star topologies are better than the others, whereas bus topology is unfit for improving robustness by adding sensors.Index Terms-Optical fiber sensor networks, topology, robustness, Monte-Carlo method, support vector machine.

No abstract