A huge part of bridges has been designed and built during the 20th century. As they are, ageing, they are reaching their lifetime. The demand of new infrastructures is increasing worldwide, new infrastructures have to be built and ageing infrastructures need to be replaced. Nevertheless, a systematic replacement of these infrastructure is not sustainable and not costeffective. Furthermore, replacement often leads to practical issues such as traffic congestions. Fortunately, civil infrastructures are designed safely, and reserve-capacity exists. Assessing this reserve-capacity through structural identification allows to extend bridge’s life. New technologies using sensors give the opportunity to interpret model-based data by using structural-identification methods such as Error-Domain Model Falsification (EDMF). The performances of EDMF method depends on the sensor placement relevance. Nevertheless, this is often performed qualitatively by experienced engineers, and no framework exists to obtain an optimal sensor configuration, considering a complete set of criteria. In this master project, a new framework is proposed to improve the sensor-network performances based on the comparison of multi-criteria decision-making approaches which criteria are as follows: information gain, cost of monitoring, robustness of information gain to sensor failure, installation constraints and ability of the network to detect outliers. A new framework was developped to improve the efficiency of the previous sensorconfiguration methods based, only on redundant or uncomplete set of criteria.