Stable interaction of multiple optimizations in wireless networks

The ability of base stations in wireless access networks to regularly and autonomously self-optimize their parameters has become a key requirement from network operators. A number of specific optimization use cases have been discussed so far, and the federation of the entire available set of parameters has been approached by separating optimization use cases in groups that are treated consecutively and that have negligible or no interactions between each other. In this paper, we introduce various separation strategies and discuss their suitability to avoid conflicts. One such strategy we propose is a separation by the amount of measurements needed to make a reasonable decision to modify the parameters. For this particular case, we statistically derive insight on the relation between the tolerance of parameters and the number of observations that are necessary to trigger a parameter modification. Recommendations are provided towards a stable holistic autonomous solution for wireless access networks.