Birden çok ortak çalışan baz istasyonu içeren çok girdili çok çıktılı veri ağları için hüzme biçimlendirme tasarımı

New generation cellular systems are required to support high data rates. The performance of conventional systems with a single central base station is limited by inter-cell interference. Furthermore, the network densities grow exponentially and it is not cost-effective for operators to use a large number of fully equipped base stations in small areas. Using multiple cooperating base stations, each equipped with multiple antennas, inter-cell interference can be eliminated by a proper beamforming design. In this thesis, beamforming strategies for cloud radio access networks and integrated access and backhaul systems are studied. In these systems, the baseband processing is centralized, and hence by an optimized resource allocation, the coverage can be significantly increased. Furthermore, the usage of simpler base stations decreases the cost of deployment, resulting in efficient and cost-effective system designs. Wireless fronthaul/backhaul solutions further decrease the capital and operational expenditures of operators. The beamforming design problem has a combinatorial nature due to fronthaul/backhaul transmissions. Applying some relaxations and transformations, the beamforming vectors are designed and cooperation schemes of base stations are optimized to minimize total power consumption and maximize the total users served under fronthaul capacity and transmit power constraints. Heuristic search, smooth approximation, bound-based link selection, and convex optimization-based algorithms are proposed to find suitable beamformers under realistic channel conditions. Some novel theoretical performance bounds are also derived to observe the limits of the algorithms and make the optimality analysis. The detailed simulations show the performance benefits of the proposed techniques compared to the existing baseline methods.