At a time in the wireless industry where average revenue per user (ARPU) is flattening and the cost of sizing networks to meet an ever-increasing demand for data isn’t going down, operators find themselves in the unenviable position of trying to maintain earnings before interest, taxes, depreciation and amortization (EBITDA) in an environment with less revenue and higher costs. On top of that, the industry is beginning to focus on building out 5G, a costly endeavor without a clear killer application to drive new revenue and EBITDA in the same way the transition from 3G to 4G did. As the perfect storm of stagnating revenues, increasing traffic and associated costs and a new technology deployment puts enormous strains on capital expenditure (CAPEX) and operating expenditure (OPEX) budgets, it’s more important than ever for operators to run their networks as efficiently as possible because network efficiency equates directly to financial efficiency.
According to wireless industry expert Chetan Sharma, over the past decade, global operators have spent $1.7 trillion in CAPEX, building out networks not only to provide wide areas of coverage but also to onboard enough capacity to keep up with the demands of data-hungry users. An insatiable appetite for media-rich content, impacted most heavily by streaming video such as YouTube, drives CAPEX spending, 90% of which is invested primarily in expanding capacity in the radio access network (RAN). The relentless growth of data traffic has led to the construction of massive networks that require significant financial care and feeding — requiring CAPEX to acquire new spectrum, new site builds, carrier additions and other site modifications and OPEX to pay tower leases, urban real estate leases, licensing fees to OEMs, backhaul cost and utilities. And the forthcoming 5G deployments will only add more costs at every point in the ecosystem.
Bigger Networks Equal Bigger Costs
Looking ahead, the picture doesn’t get much better. Cisco estimates mobile traffic will grow to 175 exabytes per month globally by 2022, up from 20 exabytes in 2017. That’s over an eight-fold increase in just 60 months. The cost to maintain and grow these networks to support the coming tidal wave of traffic is significant, and with flat ARPU at best, EBITDA will be negatively impacted in a big way for years to come. Since operators can’t simply stop adding capacity or choose not to deploy costly new 5G technology, making their current networks more efficient is one answer that can address traffic growth and, at the same time, stop the CAPEX and EBITDA bleed.
Modern machine learning software holds the key to improving network efficiency in ways not possible just a few years ago. Today’s advanced software tools can, in real time, organize video traffic on mobile networks, which can significantly increase traffic flow over the same infrastructure and spectrum without sacrificing user quality of service. The net result is a more efficient network where capacity is increased with software instead of hardware. And the advantages of this approach are significant. Instead of growing a wireless network by adding more equipment, machine learning software can do the same thing, is significantly cheaper, far quicker to deploy and is, by its nature, adaptable as conditions in the network change. It’s technology agnostic and has the same effect on 3G, 4G and 5G networks. These software optimized networks require less new spectrum, fewer new site builds and fewer carrier additions as traffic continues to grow, resulting in significant CAPEX and OPEX savings.
Improving Your Wireless Network For Years To Come
At this time in the wireless industry, if operators want to achieve significant financial and network performance benefits, they must seek out and implement true machine learning software. The ability of machine learning software to significantly optimize wireless network performance independent of human intervention marks a profound change in the way networks are managed and, as a result, costs are contained. For operators, the shift from a traditional “hands-on” approach to network management and one where smart algorithms optimize network and financial performance independently can be daunting and complex. Operators, overwhelmed by ever-increasing data volumes, should develop an understanding of what this new technology is capable of to gain a competitive advantage and take their businesses to the next level.
Demand on wireless networks only continues to increase year over year, as do the costs required to build and maintain them. As operators struggle with continued downward pressure on EBITDA from network OPEX for the foreseeable future, maximizing the efficiency of existing network infrastructure is key to reversing this trend. Like no other tool available to operators today, deployment of modern machine learning software from one of many available sources increases network efficiency, and from that grows capacity at a fraction of the cost of installing new equipment to do the same thing.
Given the perfect storm of cost and operational challenges operators are facing, the path forward to improved financial and network metrics will be found through the deployment of machine learning software, which will bring greater financial performance by creating more efficient networks.
