In mobile internet, mobile operators are facing both increased capital expenditures and operating expenses, as well as low income growth. Candidates for the next generation of network access technologies to solve the operator problem are cloud radio access networks (C?RAN’s) with high energy and spectral efficiencies. C?RAN has been supported by both the research community and industry due to its potential benefits. In C?RAN, all computational processing is performed in the central baseband unit (BBU) pool, while radio operations are carried out in the remote radio heads (RRHs). The central BBU pool is connected to the RRHs by fronthaul links. Therefore, by separating the central baseband processing unit and developing distributed radio units, the clustering structure of RRHs can be designed to adapt to network changes. Coordinated multipoint algorithms are also used in the C?RAN structure to reduce the interference effects,. Due to the high volume of channel state information required in C?RAN, the training overhead and channel estimation must be controlled. One of the challenges of ultra-dense C-RANs is how to cooperatively allocate radio and computing resources to users. This dissertation deals with the problem of weighted sum rate maximization and resource allocation in ultra dense C?RAN. To reduce complexity, we use user-centric clustering and appropriate training resources allocation. High channel training overhead is also required to estimate complete channel state information in cooperative transmission. As a solution, an incomplete model of channel state information is considerd, in which only intra-cluster channel state information is estimated. The lower bound of the user data rate is obtained by taking into account the incomplete channel state information and channel estimation error, and its convergence to the exact rate value is measured by simulation. By replacing the lower bound of user data rate in the main problem, the beamforming vectors under the constraints of computational and radio resources are designed in three steps. In the first step, the weighted sum rate maximization problem is solved under the maximum radio transmitted power constraints by weighted minimum mean square error method. Then, in the second step, using the snapsack algorithm, computational resources are allocated to users. In the third step, the maximum fronthaul capacity constraints are applied by a greedy algorithm. In the simulations, the effect of BBU capacity parameters and superstructure capacity and RRH power and clustering size on system performance and the number of required BBUs are investigated. In simulations, the effect of BBU capacity, fronthaul capacity, RRH power and clustering size parameters on system performance and number of required BBUs are investigated. Keywords: Cloud Radio Access Network, Remote Radio Head, Cooperative Radio and Computing Resources Allocation, Channel State Information, Ultra-dense Networks, Weighted sum rate maximizatio