Mobile video and the internet of things, which will add tens of billions of new devices to the network, are stretching the ability of our wireless networks to handle them. Cisco estimates that by 2020, smartphones will generate more traffic than PCs and that non-PC devices (including smartphones, tablets and machine-to-machine links) will generate 71% of all traffic. A new generation of wireless networks will be needed both to handle these existing trends and to propel completely new services.
A new network traffic report from Sandvine shows how new applications, taking advantage of better networks, drive traffic growth. Real-time entertainment, such as Netflix and YouTube, is still the largest traffic generator by far. But cloud storage of files and photos from Dropbox, iCloud and Google Cloud, to name the largest providers, has passed BitTorrent and is now the second-largest component in upstream wired traffic, at 23%.
Meanwhile, video chatting has also vaulted into the top tier of bandwidth usage. A decade ago we predicted that video chatting would be a huge generator of network traffic. But for years, it didn’t quite catch on. Wireless networks just weren’t fast enough to deliver a satisfying real-time interactive video experience. Remember the disappointment of FaceTime when it was first released Now, however, apps such as FaceTime, Skype, WhatsApp and Snapchat are exploiting faster networks and creating the type of video chatting wave we projected. Real-time communications, according to Sandvine, now accounts for 17% of upstream traffic on mobile networks.
Facebook, including its Instagram and WhatsApp properties, now accounts for 22% of all North American mobile traffic. But where would Facebook or FaceTime be if we hadn’t invested hundreds of billions of dollars in 3G and 4G networks and spectrum
Likewise, connected cars, virtual reality, mobile health and the internet of things will depend on wireless networks that are faster and more reliable, capacious and ubiquitous than today’s.
Today, around 544 MHz of spectrum is deployed for mobile use. This is a drop in the bucket. By contrast, a single fiber-optic thread boasts nearly 60 THz of potential bandwidth — 100,000 times more than all available mobile spectrum. Most of today’s licensed spectrum falls between 400 MHz and 2.5 GHz. It is augmented by unlicensed spectrum, used for Wi-Fi, Bluetooth and other short-range signals, in the 2.5 GHz and 5 GHz bands.
The new spectrum being considered is located at much higher frequencies — at 28, 37, 39 GHz and beyond. Higher frequencies have strengths and weaknesses. They can transmit more data, but the signals don’t travel as far. The higher frequencies will thus be crucial to deliver rich new services to billions of new devices, but they will also require the construction of new networks. These new networks will be composed of millions of new small cells, often hundreds of meters apart, to complement the existing network of cell towers, which are kilometers apart.
The key to the higher frequencies is their immense bandwidth. Bandwidth is the defining constraint on information transmission. More bandwidth, more data. All of today’s mobile signals are crammed into just more than 500 MHz of bandwidth. But it’s possible that just one of the new higher-frequency bands could offer that much. Taken as a whole, the new high-frequency airwaves could possibly quintuple the amount of spectrum available for mobile and wireless broadband.
The FCC order is a big step in the right direction. As usual, however, some firms are asking for special favors, hoping to steer the proceeding away from simple, neutral rules and toward complicated bureaucracy. They are attempting to block competitors from acquiring new spectrum, which could delay the arrival of 5G networks and thwart a vast array of new services. The FCC should thus resist the temptation to apportion spectrum to particular firms based on political favor.
Connected cars, virtual reality, mobile health and the internet of things will depend on wireless networks that are faster and more reliable, capacious and ubiquitous than today’s. 5G networks are a crucial foundation for the next 25 years of U.S. innovation. If we get spectrum policy wrong, the whole economy will lag.
Bret Swanson is president of Entropy Economics LLC and a visiting fellow at the American Enterprise Institute’s Center for Internet, Communications, and Technology Policy.