A version of this article was first published on webscalenetworking.com.

5G is coming, and with it comes the expectation of wireless speeds that are 100X or more what we experience today with 4G. In fact, one of the goals of 5G is to achieve maximum download speeds of 10 Gbps per user. This influx of traffic won’t come without a cost to the underlying networks that support it.

During a recent webinar about 5G technology and its impact on wireline networks, Ciena’s Brian Lavallée talked extensively about how 5G will affect all parts of the network, from end users to data centers and everything in between. To succeed, mobile network operators (MNOs) will need more than just a new radio access network, they will also need fiber—and lots of it – to manage the massive increase in bandwidth that will come as billions more users, both human and machine, join the network.

This Q&A with Brian digs into some of the key questions around the future of 5G.

Q: Will the proposed 5G performance enhancements be offered to all mobile users and devices?

BL: No. 5G is expected to be deployed strategically in different locations, especially in the early days. If consumers are expecting all 3G and 4G networks to be replaced with 5G, they’ll be disappointed. 5G is expected to complement 3G/4G where it makes sense. And depending on where service providers believe applications and use cases will be most lucrative, they can roll out speeds of up to 10 Gb/s.

This means if you’re in a rural community, chances are you probably won’t get 5G in the early days. In cities and metro areas you’ll see potential applications like enhanced mobile broadband, self-driving cars, video broadcast services, and other use cases that will require high-bandwidth and/or low-latency. So, service providers will deploy 5G in geographic areas where it makes economic sense.

Q: Will 5G require new wireless spectrum?

BL: Yes. A lot of people are looking at what part of the spectrum will be required, and different frequency ranges are being proposed. To complicate matters, different frequencies are being suggested in different countries. It’s too early to say if this will cause problems for 5G radios, but there definitely will be new wireless spectrum required to achieve the download capacities and associated coverage.

Q: Will small cells be required to implement 5G?

BL: Definitely. This is mainly because 5G will use higher frequencies, which don’t travel as far and don’t travel as well through obstacles, such as buildings so prevalent in concrete jungles (cities). This means Mobile Network Operators must install 5G radios, and a lot of them, closer to end users. Small cells won’t be required day one, as they can be installed over time. But there’s no getting around the fact that rolling out 5G to large geographic areas will require a tremendous number of new small cells.

Q: What are the advantages and disadvantages of fixed mobile broadband access?

BL: In the ideal scenario, MNOs could ship a 5G fixed access node to an end user and have it up and running quickly. That user may have to install the node on their wall, but once it’s turned on, it auto-discovers the mobile network and it’s good to go. It also means using new spectrum, and if service providers introduce mobile broadband access into a home, it opens the network to every device, sensor and “thing” in the house using high-bandwidth applications such as HD audio, video, or virtual reality, which will place an enormous burden on the RAN side of the network. The same goes for installing a 5G fixed radio access point on the side of a building serving tens to hundreds of employees.

Q: How much fiber will each cell site require in 5G?

BL: Given the optical technologies available today, we can easily ramp the capacity of a fiber into the terabits per second, but the economics of doing so towards the network edge still needs to be worked out. MNOs concerned about diversity will probably use multiple fibers, but capacity can be delivered with just one pair. For MNOs less concerned with diversity, a single bidirectional fiber may also be an option, albeit likely a more unpopular one, given the amount of capacity a 5G-enabled cell can carry.

Q:  Will IPv6 have a place in 5G or is IPv4 sufficient?

BL: IPv6 will have a big place in 5G, primarily because of the Internet of Things (IoT), which will add tens of billions of new devices to this mobile network as it’s roll out. IPv4 cannot cope with the number of unique IP addresses that will be required, but IPv6 can. IPv6 also introduces the Neighbor Discovery Protocol (NDP) that enables multi-protocol interoperability between IoT devices. So, both the number of addressable addresses in IPv6 and the features of the protocol will be critical to the success of 5G.

Q:  How will 5G address or improve voice quality, and in what way?

BL: That’s a good question because the voice portion of cell networks is rarely discussed anymore. There are a lot of over-the-top (OTT) VoIP applications today that are at the mercy of the network because existing 4G networks operate at best effort. This means the performance of these OTT applications is often unreliable. If there was a guaranteed service level assigned to a VoIP service, users would get the same performance level as today with wireline networks. MNOs could deploy their own VoIP applications (to replace VoLTE) giving them a guaranteed level of quality of service using network slicing, and by taking advantage of the much lower latency of 5G networks.

The Road to 5G podcast image