A common scenario in wireless communication systems is one where several systems coexist in the same area and communicate using the same portion of the spectrum, possibly even using different radio interfaces. Herein, we consider non-cooperative spectrum sharing rules for decentralized iterative dynamic channel allocation, based on utility functions, for multi-user systems with multiple antennas at both transmitters and receivers. Using the link rate as a utility function often leads to a suboptimal high-interference situation. The key to obtaining good separation over the scheduling slots, leading to higher spectral utilization, is a proper choice of utility function for the users. To obtain a decentralized scheduler with these desired properties, we introduce the idea of modifying the utility function and investigate its affect on the sum rate. A numerical example illustrates the performance compared to a scheduler employing the link rate as utility function.