Future mobile access networks, like LTE-Advanced, target wireless speeds of 1 Gbit/s per cell. Such data rates require to increase the cell density. This, however, also increases interference, which limits the mobile network performance and prevents to reach the targeted data rates. Therefore, interference management has become the key factor to determine the overall wireless system performance. The most promising method for managing interference is to coordinate neighbor cells. This approach is called Coordinated Multi-Point (CoMP) transmission/reception. Multiple BSs form a cluster and cooperate by exchanging, e.g., signaling and/or UE data. To enable CoMP, the backhaul network must fulfill stringent capacity and latency requirements, which are on the order of several Gbit/s backhauling capacity per BS and round-trip latencies between BSs down to 1 ms. This thesis addresses CoMPs feasibility from the backhaul network perspective. It first evaluates whether current and future backhaul infrastructures can support certain CoMP techniques. It turned out that even optical technologies, like Passive Optical Networks (PONs), do not allow CoMP in a satisfactory way. Therefore, I investigate mechanisms to deal with limitations caused by the backhaul shortcomings and mechanisms that improve CoMP feasibility. Finally, the influence of the metro-wide network topology and architecture is evaluated.Based on the gained insights, several conclusions can be drawn. First, CoMPs feasibility is strongly limited, even with optical backhauling technologies. This requires a cross-domain approach, which decides when, where, and how to cooperate based on wireless and on backhaul network properties. The presented techniques and architectures extend CoMP support to a metro-wide scale while even reusing existing deployments to a large extend.