The Electronics and Telecommunications Research Institute (ETRI) of Korea recently released a research report stating that after telecom operators adopt artificial intelligence (AI) networks, overall total cost of ownership is expected to fall 7.1%, with network construction and operation costs decreasing 17.75%. The estimate is based on the assumption that network-related costs account for about 40% of operators' total costs. The study shows that virtualization of radio access network and mobile core network equipment is the main source of cost reduction, with these two categories accounting for about 40% of network costs. After adopting open radio access network and cloud radio access network technologies, related equipment costs can fall 25%, equivalent to a 10 percentage point decline in total network costs, accounting for more than half of the overall network cost reduction.
Other cost reduction items include AI energy optimization, transport network automation, and fault prediction. Adjusting base station power consumption according to traffic changes and leveraging equipment idle time to reduce energy consumption can lower network costs by 2.25 percentage points; transport equipment and operating personnel costs can each be reduced by 2 percentage points; and predicting faults and reducing maintenance burden can lower costs by 1.5 percentage points. However, actual results depend on the virtualization level of existing networks. If the cost reduction ratio from wireless network virtualization falls from 25% to 15%, the reduction in operators' total costs will drop from 7.1% to about 5.5%.
Global AI Network Equipment Market Expected to Grow Significantly
The study projects that the global AI network equipment market will grow from $13.8 billion in 2030 to $95.6 billion in 2037, expanding about 6.9 times, with a compound annual growth rate of 31.9%. Of this, the AI radio access network equipment market is projected at $6.6 billion in 2030, while the AI autonomous network equipment market covering mobile core, transport, and backbone networks is projected at $7.2 billion. The study also notes that while network virtualization and operational automation can save costs, new investments in AI accelerators, edge data centers, and high-performance switches may rise.
The relevant calculations do not account for the net financial impact after new investments; actual economic viability will also be affected by factors such as equipment replacement cycles, AI deployment speed, and investment payback periods.