5G, the fifth generation of wireless telecommunications, is making its debut across the globe. Communication Service Providers (CSPs) are in a race to be the first to offer complete 5G service within their markets. The stakes are high, but so are the potential business benefits. The technology promises faster connections, more reliable data streaming and a host of new capabilities that bring new opportunities.
Deploying 5G will be a challenge. Dense and complex, 5G networks have potentially high operating and maintenance costs. For certain applications, like connected cars, the latency must be ultra-low and totally predictable. Plus, CSPs must also consider potential security implications.
One area that 5G will directly affect is the design and architecture of the Radio Access Network (RAN). Simply put, the RAN is a collection of edge located functions that connect a mobile device to the CSP’s core network. The latency requirements and network load of 5G will put a great deal of strain on the RAN and the traditional ways of deploying RAN equipment are not well suited for the new needs. A new, cloud-based Virtual RAN (vRAN) approach provides the necessary functionality for 5G; performance, flexibility, and cost-efficiency that isn’t available in fixed-function RAN equipment.
Mitigating the Complexity of 5G Distributed Networks
5G networks will be categorically more complex than anything currently in production. Many factors contribute to this situation. For one, 5G networks promise to be incredibly dense compared to 4G. A single standard 4G macrocell can serve an area of approximately 25 square kilometers, whereas, with 5G, it will take 20 or more cells, with their respective antennas and RAN hardware, to cover the same area. When 5G traffic goes indoors, the coverage requires even greater density.
The Massive Multiple In Multiple Out (Massive MIMO) approach to wireless networking allows for the use of multiple antennas in data transmission. Massive MIMO gives 5G networks greater speed per antenna. It increases the number of transmitting antennas at the base station and mobile user endpoints (UEs). The result is the maximization of data transfer through simultaneous sending and receiving.
The density of the 5G network thus drives an increased need for switching and associated traffic management. It also requires a lot of compute power at the edge due to the following:
- MIMO works through algorithms at both the base station and UEs.
- “Beamforming” optimizes bandwidth use but further stresses compute resources.
- 5G duplex (two-way) data traffic architecture and high speeds create volumes of data that dwarf those in 4G.
Keeping Operating Costs Under Control in the 4G to 5G Transition
As one might imagine, adding 20X or more pieces of hardware to a network generates significant operating expenses. Beyond the equipment itself, there are cost centers that span configuration work, network testing and management, and software updating.
Meeting Ultra-Low, Deterministic Latency Requirements
5G networks need low latency to function and some emerging edge devices in the 5G network will require ultra-low, deterministic latency.
This new market of devices is defined by the need for ultra-reliable and low latency communications (URLLC). These requirements only allow a single millisecond for one-way communications across the entire network infrastructure. That’s much faster than legacy systems can accommodate. Examples of applications with these criteria include those used for autonomous vehicles, industrial robotic control, and safety. The CSPs that can successfully support the needs of URLLC devices stand to dominate the markets of the future.
New Capabilities Mean New Markets
One exciting aspect of 5G is its potential to transform whole industries beyond telecommunications. These industries will become huge new CSP customers.
For example, 5G is expected to form the backbone of autonomous vehicles and vehicle-to-vehicle (V2V) applications. It’s also poised to transform the way industrial sensors and other Internet of Things (IoT) devices function and communicate. 5G is even envisioned as playing a role in augmented reality, which could, in turn, influence the “smart city” phenomenon.
Who will these industries turn to for 5G services? Most likely, they will want to work with established CSPs. Industrial companies, “smart city” governments, automakers, and others will likely prefer to outsource the management of 5G’s higher bandwidth and massive endpoint count. In other words, it’s a big revenue growth opportunity for CSPs.
At the same time, these new avenues of growth come with customers who will present unfamiliar requirements. Setting up a massive industrial IoT environment, for instance, promises to be different from the regular business of running a CSP. Each opportunity is uncharted territory that comes with unique challenges, for example:
Industrial and robotics: Industrial companies want 5G to serve as the basis for an intelligent and secure edge platform, one that enables autonomous operations and distributed decision making. One difficulty with this comes from vision-based systems, i.e. robots that can “see.” They need low latency. Otherwise, the robot may not be able to react in time to visual inputs.
Automotive: Original Equipment Makers like Detroit’s “Big Three” and others want to make autonomous vehicles a commercially viable reality. One goal is to produce a Level 4 (L4) autonomous car that carries people around as if they were cargo. The car does everything, perhaps not even having controls for a person to use. L4 vehicles need, what appears to be, an unsustainably costly amount of onboard compute power in order to function safely. This problem can be solved with a small-footprint edge computing solution that provides on-vehicle compute while offloading compute to an edge cloud.
Aerospace and Defense: This sector is largely interested in 5G for its relevance to Urban Air Mobility as well as the potential to maintain global technology leadership and ensure Defense communications are optimized and secure.
New Security Issues
5G, like any new, accelerative technology, will produce its share of new security risks. Given the distributed and sometimes remote nature of 5G far edge nodes, some unique capabilities need to be considered to mitigate that risk, such as:
- TPM (Trusted Platform Module) to secure edge site hardware via cryptographic keys
- QAT (Quick Assist Technology) with Key Protection
- Secure UEFI boot
- Zero Touch updates to keep infrastructure current and secure.
5G networks will need ongoing security monitoring at all levels, from hardware to application software, operating systems and switches, and issue patches when necessary.
5G promises exciting advances for Communications Service Providers (CSPs), but the 5G rollout is going to be challenging. CSPs must rapidly build out dense, low-latency edge networks in ways that are affordable, secure, and easily maintainable. CSPs are looking toward open source, container-based network infrastructures that meet the 5G latency, reliability, and flexibility requirements while being inexpensive to deploy and maintain. And of course, security is a factor to contend with as well.
Paul Miller is Vice President of the Telecom Market Segment at Wind River, responsible for the company’s strategic vision and market support for telecommunications TEMs and CSPs. With more than 25 years of telecommunications and advanced applications technology leadership at both large companies and successful startups, he is currently focused on Wind River’s NFV and SDN enabled products for the carrier market.