Next generation access networks will be based on the Fiber-to-the-X (FTTx) paradigm,
since optical transmission provides the required bandwidth for future services and
triple play architecture. FTTx technologies are now mature and standardized; network
deployment phase has already started in many areas around the world.
The profitability of such an optical broadband access technology strongly depends
on the design of optimal topologies, minimizing initial investments (CapEx). The network
planning problem itself is difficult (NP-hard), however close to optimal topologies
are required even for large data sets, for city-scale service areas.
Therefore efficient heuristic approximations are needed, like the Branch Contracting
Algorithm (BCA) presented in this paper. It can handle real problem instances, provides
close to optimal solution, and works fast enough to support economic comparison of
different architectures, supporting decisions of a network operator.