The rise of Open RAN is gaining momentum, with major network operators such as AT&T, DoCoMo, and Vodafone making concrete plans. However, there is a concern that the top network vendors may build AI enhancements into the RAN, which could potentially disrupt the Open RAN ecosystem. A recent report highlights the capacity impact of AI on the RAN, showing a 20-30% improvement in capacity. This is crucial, as mobile operators anticipate a “capacity gap” in dense urban hotspots by 2027-2029.
The integration of AI/ML techniques into the RAN can help close this gap, but it raises a significant issue: AI/ML models rely on hard data to refine and improve themselves, and the data formats may not be compatible between different vendors. This could lead to a situation where major vendors like Ericsson or Nokia use their AI/ML models to differentiate themselves, giving their preferred hardware the best performance possible. This could be detrimental to Open RAN, as operators may be tempted to use their incumbent vendor in urban areas where capacity matters most.
This dilemma highlights the need for standardized data formats and closer collaboration between vendors to ensure seamless integration of AI/ML models with Open RAN. As the industry navigates this complex landscape, it’s essential to be aware of the potential disruptions that could arise from the intersection of AI and Open RAN.
