It becomes clearer and clearer that 5G is different from all “G” predecessors in a lot of ways. 5G is not simply a network, but rather an ecosystem that supports vertical applications and industries, enabled by the three use cases, eMBB, URLLC, and mMTC.
This brings with it the potential of unleashing unparalleled opportunities for telcos, as well as enterprises with Internet of Things (IoT), and many other technologies. But all these come at the cost of significant increase in the complexity required to support services with such diverse QoS/QoE requirements, that can be completely different, and even contradictory.
Then, there are the 5G unprecedented performance requirements, with high bandwidth, low latency, connected device density (including humans, M2M and IoT), and coverage area. All of this is accomplished with high energy efficiency and low battery consumption constraints.
Last, but not least, unlike voice and data centric predecessors, 5G is designed at its core with human and machine context-aware QoE service delivery.
However, 5G can meet its requirements and achieve the technical and business potential for which it has been designed, only if 5G backhaul can cope with its unprecedented challenges.
The unprecedented 5G backhaul challenges of capacity
The first 5G backhaul challenge is the increase in capacity of x1000 per site. This capability cannot be feasible without a much denser mobile grid than that of legacy networks. In addition, the introduction of mmW frequencies for RAN require smaller coverage areas per cell site, and consequently, a denser grid. This requirement comes with several 5G backhaul challenges:
-
A denser backhaul link will highly limit the frequency reuse, as the links get closer to each other, and therefore require better utilization of the wireless backhaul spectrum
-
There will be unprecedented requirements for cell site synchronization, and therefore, much stricter accuracy requirements than LTE-A (i.e., 1.5 μs to approx. 0.5 μs)
-
Long distance reach, represented by how far a cell site can get backhaul support from the core network, will require high levels of quality of service. Massive backhaul traffic aggregation at the super cell, can generate congestion, and collapse the backhaul networks
-
Mass deployment requires high capacity non-line-of-sight wireless backhaul links, as well as, quickly installed, low footprint, low-power consumption equipment. Therefore, there will be an expected increase in deployment costs
Massive Traffic and Maintaining Quality of Service
The second 5G backhaul challenge is peak throughputs and speeds of 10Gbps. This will have a significant impact on the media type (e.g. Ethernet, Fiber, Air) used for the backhaul.
There will no doubt be high impact on the backhaul media, with the high availability and very low latency required to support mission-critical services, such as autonomous vehicles, tactile Internet, and many other machine-to-machine applications. For these deployment examples, the risks of network failure are very high. Therefore, the biggest 5g Backhaul challenge is to support massive traffic, while maintaining the required quality of service and lower latency requirements.
Supporting network slicing
Finally, but not the least of which, is the need to support network slicing. 5G uses network slicing to cost-effectively support all the services, by meeting the most onerous requirements for each service. Network slicing allows multiple logical networks (slices) to be created over a single physical network, with each slice built to address a set of service behaviors, such as latency, availability, bandwidth, etc.
Network Functions Virtualization (NFV) and Software-Defined Networking (SDN) technologies are required to support network slicing. NFV enables service providers to flexibly add network computation and storage when and where they are required by a service. SDN provides the ability to add capacity and connectivity in real-time. So, as services move around the network, that capacity and connectivity can follow the services.
Service orchestration is key to making network slicing operate end-to-end across the entire mobile network. This orchestration provides an understanding of the service requirements, and translates these in real-time into the network requirements for each part of the network, including RU, fronthaul, backhaul and Core.
The impact on the 5G backhaul network, is that it must support the slices as defined by the service orchestrator. IP transport has supported virtual private networks (VPNs), each with different performance characteristics, for decades already. Then NFV has been deployed in mainstream networks for some years already as well. However, the 5G change comes with the need for decoupling the control and data plane to allow control of the network from a centralized controller and/or orchestrator. Working together, NFV and SDN can ensure that each service shares the backhaul as efficiently as possible, while optimizing per-application/service QoE that is scalable and flexible.
Although 5G backhaul challenges exist, backhaul solutions are still in discussion. Different operators will likely opt for different approaches, such as fiber and/or wireless (e.g. mmW, free space optics similar to fiber, but using an invisible beam of light for transmission). They will likely be based on the self-backhauling (integrated access and backhaul) concept, which can address some of the 5G backhaul challenges, including:
-
Higher spectrum efficiency (reuse of time, frequency and space resources between access and backhaul)
-
Higher cost efficiency with shared radio hardware
-
O&M management systems
-
Higher performance based on dynamic optimization of resource across access and backhaul
However, these benefits can only be safely enabled through well-designed testing and monitoring of the mitigations taking place at the new access-backhaul interface. Additionally, ongoing careful assessment of the complex scheduling of channel resources between the two domains, and the potential limitations on the end user experience (e.g. rate, latency), due to the sharing of resources between access and backhaul.
5G backhaul comes with a wealth of unprecedented challenges. So far, several issues are still open, and final solutions are not completely decided. Infovista is a provider of testing solutions for 5G backhaul deployments. Our network planning and optimization and Service Assurance solutions can be used towards this scope. For example, dimensioning of 5G backhaul can benefit from Infovista’s 3D mmW planning and traffic prediction modeling. Our planning solution can be used for interference detection, in the case of integrated access-backhaul using the same frequency range. Infovista Service Assurance solutions can be used for network slicing, SLA evaluation and monitoring. They can also optimize the network slices’ resource allocations, and monitor the channel resource scheduling between the two domains.
To find more details about our products and how these can help visit https://www.infovista.com/.
