A single failure in a communication network may trigger many alarms. When the communication network uses optical fibres as a transmission medium and increases its capacity by using Wavelength Division Multiplexing (WDM) and Space Division Multiplexing (SDM), the number of alarms and the difficulty to locate the failure are considerably higher. In this scenario, a single failure may interrupt several channels and the quantity of lost information is larger. We propose an alarm filtering algorithm for the fault management of an optical network that supports multiple failures and works in the presence of passives elements. A passive element is an netwrok element which may fail but never generates an alarm (e. g. an optical fiber). Our algorithm avoids the use of failure probabilities because they are difficult to estimate. The algorithm also accepts alarm losses. It presents to the human manager a list of faults which may have caused the observed alarms. We present a definition of teh problem and its abstraction. A feature of the algorithm is that it does not need a global knowledge of the network topology. The algorithm will be applied to the optical network of the ACTS COBNET project.