The ever-increasing growth of traffic in backbone networks is expected to exceed the available capacity provided by the fixed-grid wavelength division multiplexed (WDM) technology. Orthogonal frequency division multiplexing (OFDM)-based networks, often called flexible optical networks (FONs), have been recently proposed by the research community to address this bandwidth crunch. Flexible-grid networks provided by FONs can now handle traffic demands by the elastic allocation of spectrum contrary to the fixed grid utilized in WDM networks.

In FONs, optical fibers transfer enormous amount of data over small periods of time. Hence, important security issues arise, since even short attacks can still compromise large amounts of data. Therefore, it is important to ensure that security is considered in these evolving communication networks. Physical layer security, also known as optical layer security (OLS), offers various benefits to the entire network. For instance, implementing a secure protocol in the optical layer could relax security measures in the upper layers, thus improving throughput. Further, if the optical layer is not properly secured, any secure protocol implemented at the upper layers would still be susceptible to attacks through the lower layers and thus, valuable information can still be extracted. Therefore, by enabling OLS, the information of all higher layers will be hidden to an eavesdropper, making the confidential connections more secure.

The project “Resource Allocation in Flexible Optical Networks” (REALFON) aims at designing and developing innovative optimization algorithms for resource allocation in FONs taking into account several design parameters. The optimization algorithms proposed by REALFON is based on four technical pillars:

a) Attack-aware routing and spectrum allocation (RSA) in FONs with traffic demand variations
b) Attack-aware RSA with physical layer impairment (PLI) considerations in spectrally spatially (SS)-FONs
c) Survivable and secure FONs using network coding
d) Secure anycasting in FONs

These innovative solutions will reach a significant advance beyond the current state-of-the-art and will fulfil the requirements of future optical networks regarding scalability and security, as well as important savings on the spectrum utilization for network operators. Furthermore, reduction of the total cost of ownership and power consumption represent other significant results. Integer Linear Programming (ILP) formulations, relaxation techniques, and meta-heuristic algorithms will be proposed in order to deal with open issues of resource allocation in flexible optical networks.