This blog was originally posted in RCR Wireless on January 11, 2019.
As mobile network operators (MNOs) prepare to integrate 5G wireless technology and services, they are taking the opportunity to evolve their network architecture with three goals in mind: reduce costs; improve deployment efficiency, service flexibility and velocity; and open the market to innovation to assist in the delivery of new use cases and driving new revenue streams. By moving to an open, centralized radio access network (Open C-RAN), MNOs can accomplish these goals.
De-constructing the network
A mobile network base station currently consists of remote radio units (RRUs), baseband units (BBUs), fronthaul links to connect the RRUs to the BBUs and backhaul links to connect the Radio Access Network with the core network. Each base station is a discrete access node in the network with its own equipment along with power and air conditioning units to support that equipment. Traditionally, MNOs have relied on a handful of vendors (i.e., Ericsson, Huawei, Nokia, etc.) to supply RRUs and BBUs, and once they select a vendor, they are locked in because one vendor’s equipment doesn’t interoperate with another vendor’s equipment. In this market environment, prices are relatively high because once the vendor is chosen there’s no competition.
An Open C-RAN fundamentally changes the equation by centralizing much of the BBU functionality. This is achieved by splitting BBU functionality into central units (CU) which perform core non-real time processing and distributed units (DU) which perform most of real-time radio-link processing. The exact functional split between CU and DU has been the subject of much debate during the 5G standards process; however, it will be relatively straightforward to virtualize the CU functionality, so it can run as software on common, off-the-shelf servers. Later, possibly with use of hardware accelerators, even DU functionality could be virtualized. In addition, the Open C-RAN leverages standards-based software that greatly eases compatibility restrictions. With an Open C-RAN architecture, MNOs are free to select their own servers from a wide range of vendors, and to develop or purchase BBU software written in common programming languages. This not only reduces costs, it increases service flexibility since the MNO is no longer dependent on a network equipment vendor’s schedule for new service rollouts.
The near-term deployment outlooks
Over the next 3 to 4 years, MNOs will focus on delivering enhanced mobile broadband over 5G infrastructure. 5G’s massive increases in throughput (up to 10Gbps per user) and capacity will lay the basic groundwork for future developments such as vast numbers of connected devices (the Internet of Things) along with ultra-reliable and ultra-low-latency connections that will enable new applications such as remote surgery and self-driving cars as well as many industrial applications. Deployments began in 2018 with fixed broadband 5G wireless that complements fiber-to-the-home or fiber-to-the-business connections and will evolve soon after to encompass mobile connections.
This transition won’t be without its bumps and bruises, however. A new crop of vendors must emerge, and they must provide not only virtualized BBU solutions, but also services like integration and regression testing, which today are included in offerings from the major mobile equipment vendors. In addition, the fronthaul and backhaul links between virtualized BBUs and the RRUs at cell sites will need to be upgraded to fiber to deliver the required bandwidth. But new vendors are rising to the occasion. They include AltioStar, Amdocs, Aricent, ASOCS, Blue Danube, Ciena, Cisco, CommScope, Fujitsu, Intel, Mavenir, NEC, Netsia, Nokia, Radisys, Nokia and Samsung.
How an open C-RAN helps MNOs
Now, let’s see how Open C-RAN architecture will help MNOs achieve their basic goals.
Many people discuss 5G goals in terms of enhanced mobile broadband, massive machine type communications and ultra-reliable low-latency communications. However, it is often forgotten that a fourth target of improving network efficiency by 100 times was an original ITU goal in its 2020 vision statement. This network efficiency target is necessary if our industry is to deliver on the service promises of 5G in an efficient way. By virtualizing BBU functionality in a central office, MNOs can significantly reduce equipment costs (using a white-box server versus a proprietary piece of network equipment), reduce maintenance costs (BBUs are centralized, so the MNO reduces truck rolls to cell sites), and reduce operating costs (cutting or eliminating air conditioning and power costs at the cell site).
Deployment and service flexibility
By virtualizing BBUs, MNOs have the flexibility to deploy them wherever it makes sense. In some cases, this may be in a central office, and in others it will be in regional or edge data centers. Edge data centers can help MNOs reduce latency to enable forthcoming applications that require highly reliable, low-latency connections.
Moving to an open architecture also makes it possible for MNOs to develop their own services using standard hardware and software. Services such as software-defined networking in a wide area network or virtual private networks help MNOs improve average revenue per user, reduce churn and speed return on investment for the new infrastructure.
C-RAN also helps improve the flexibility of network operations: the virtualized C-RAN units and baseband units will begin to resemble mini-data centers, and MNOs will be able to move capacity around the network to serve different groups of subscribers with different needs at different times of day. For example, an operator might shift capacity from train stations during to the commute hour to office centers during the workday to suburban neighborhoods at night. This flexibility also helps MNOs minimize interference by coordinating transmission across several adjacent cell sites. Co-locating also offers operators the opportunity to improve transmission efficiency by pooling resources, making dual connectivity effective and delivering service functionality across networks.
An end to vendor lock-in
Because they will use open hardware and software for their infrastructure, MNOs are no longer tied to a specific vendor. Virtual BBUs should be compatible with RRUs from any vendor, and the competition will drive lower prices.
As 5G networks roll out, we will see more competition among vendors and MNOs, but the network will be more efficient, and MNOs will gain more autonomy and flexibility in how they build out their networks and offer services. The Open C-RAN thus becomes more than an upgrade; it’s a platform for a new way of doing business that drives new profits and delivers happier customers.