Planejamento de redes WDM resilientes em malha com compartilhamento de recursos de proteção para conexões com requisitos de disponibilidade sujeitas a múltiplas falhas

Failures of fiber links can result in major loss of data in high speed optical communication networks. Survivability is of critical importance and assuring high levels of availability becomes an important issue. A typical approach to the design of resilient optical networks is through protection schemes that pre-determines and reserves protection resources considering single and double link-failure scenarios. This thesis proposes a heuristic planning for WDM networks that computes the resource capacity required to transport the traffic demand and to protect the optical connections meeting the availability requirements in scenarios of multiple link failures. To protect a connection against multiple failures, the model selects, for each connection, a service lightpath and protection lightpaths, where is the quantity of simultaneous failures to be considered. Thus, +1 lightpaths are selected and arranged in a sequence of activation, where the first lightpath of the list is the service lightpath, and the others are protection lightpaths. Each set of +1 lightpaths in a row of activation supports a connection. For the network model considered, the source node of the connection is responsible for switching the protection lightpath, then, when a connection is interrupted, the source node initiates the re-establishment of the connection, using the next protection lightpath. The proposed method is called MSB (Multiple Shared Backup). The MSB method proposes two algorithms, one for selecting the lightpaths for each connection, and another to calculate its unavailability. From a set of k shortest paths, the first algorithm selects the lightpaths of each connection and their order of activation. The second algorithm uses a method to calculate the availability of each connection (hours per year) for multiple failures. Both algorithms are used by a planning procedure to meet the demand between all pairs of nodes, ensuring a pre-defined level of availability, using the least possible amount of resources.