One of the techniques that has been recently identified for dealing with multi-user interference (MUI) in future communications systems is base station (BS) cooperation or joint processing. However, perfect MUI cancellation with this technique demands severe synchronization requirements, perfect and global channel state information (CSI), and an increased backhaul and signaling overhead. In this paper, we consider a more realistic layout with the aim of mitigating the MUI, where only local CSI is available at the BSs. Due to synchronization inaccuracies and errors in the channel estimation, the system becomes partially asynchronous. In the downlink of wideband code division multiple access based systems, this asynchronism stands for the loss of the orthogonality of the spreading codes allocated to users and thus, for an increase in the MUI level of the system. In this contribution, we propose a framework for mitigating the MUI which builds in three main steps: definition of a cooperation area based on the channel characteristics, statistical modeling of the average MUI power experienced by each user and a specific spreading code allocation scheme for users served with joint processing. This code allocation assigns spreading codes to users in such a way that minimum average cross-correlation between active users can be achieved. Interestingly, these steps can be performed with a limited amount of extra feedback from the user's side.