The New Paradigm for Wireless Network Optimization: A Synergy of Automated Processes and Human Intervention

With the evolution of wireless technologies, the optimization process for the radio-access network has undergone a dramatic increase in complexity. This trend is driven by the introduction of multiple objectives for packet data services, the increasing network heterogeneity resulting from technology overlays, and the growing demand for non-disruptive optimization of operating networks. In a multi-disciplinary effort, Bell Labs has developed a new optimization methodology that tackles these challenges using automated processes. We show that the paradigm shift toward automation requires the development of several interdependent algorithms each designed to cope with a particular optimization task. In the application described, human intervention proves indispensable demanding a new kind of expertise that combines the understanding of the automated processes with the knowledge of network operation. Through examples, we illustrate the variety of tasks that have to be automated and where guidance is needed through human control. The development of these automated processes is an important step toward dynamic optimization routines that are integrated into the network and adjust the network in real- time manner.