From the editor: The three Long-Term Evolution (LTE) network layers are designed to operate independently but to cooperate with each other. In the second of four articles in our Telecom Insights guide, Building 4G wireless networks: Exploring LTE architecture and services drivers, telecom expert Tom Nolle, president of CIMI Corp., explores the special considerations in planning optimal technology choices within each layer. Nolle addresses the role of the radio access network layer for service, 3G migration and tower location issues, and the Evolved Packet Core for traffic and mobility management.
Wireless 4G Long-Term Evolution (LTE) infrastructure, has a three-layer network structure where components perform a set of structured missions that are independent but cooperate with the other LTE network layers. LTE technology diagrams clearly show this structure -- which includes the radio access network, the control layer and the evolved packet core (EPC). But in the real world, network planners often have to contend with pressures that operate not on LTE infrastructure as a whole but on a component that represents only part of a specific LTE network layer.
Beyond the physical LTE network, most operators divide their organizations and activities so responsibilities for the radio access network can be separated from tower interconnection or from service management. A critical task for operators is to plan LTE network layers independently without losing the cooperation.
LTE radio access network addresses services, 3G migration and tower location
In a practical sense, LTE planning starts with the radio access network. Ideally, operators should consider deploying software-defined radio (SDR) technology capable of supporting both 3G and 4G services in the spectrum bands available as soon as possible to make their radio access network technology LTE-ready. When new cells are added or when cells are modernized, this software-defined radio platform investment can pay significant dividends in ensuring that radio life is maximized.
Radio access network planning for LTE must also consider target services, target locations and migration needs. Nearly all LTE deployments will evolve from existing 3G services, and there is a temptation to think of LTE as replacing or paralleling 3G on a per-cell basis, which would create planning problems for the following reasons:
- The capacity of LTE cells is higher, and a smaller number of more widely spaced cells may work well if cell capacity limits aren't exceeded.
- Early customers for LTE services are probably going to be users who have specialized 4G LTE handsets or appliances or who have specialized needs. These customers may be more likely to use LTE service in specific locations, so providing it there first is important.
- Some areas may lack the tower connection and backhaul capabilities needed to fully exploit LTE, therefore providing facilities there may take time, suggesting that the sites be de-prioritized for now.
- LTE network demand may be highly variable through the day owing to the patterns of smartphone or tablet use. Intelligent antenna technology may be valuable to customize coverage in order to maximize service utility.
LTE service evolution is particularly important, and the question is: What handset capabilities will exist? Almost every operator with LTE plans will have parallel 3G and 4G services operating for some time to accommodate the installed handset base and to support existing roaming agreements. Supporting 3G/LTE handsets for new service customers will ensure that they can use their devices while outside the evolving 4G service area, but it may also perpetuate the use of 3G. Transition planning for handsets is critical to support transition planning for the radio network.
Traffic and mobility management in the Evolved Packet Core
Traffic management and mobility management are capabilities LTE creates in the next layer -- the Evolved Packet Core. The EPC builds a sub-network within a metro/core infrastructure that extends from the tower (eNodeB) through the Serving Gateway (SGW) to the PDN Gateway (PGW) in order to provide connectivity between user equipment and the packet data network.
The portion of metro/core infrastructure within the boundaries of this chain is controlled by LTE, and the associated service control processes (often using IP Multimedia Subsystem, or IMS). Maximizing this zone makes mobile services discretely controllable in terms of Quality of Service (QoS) but can also limit infrastructure reuse. Making the zone smaller (jumping off into the packet data network sooner) will improve overall efficiency but limit mobile service customization potential.
Operators also need to consider how the LTE Evolved Packet Core elements noted above will be mapped to real hardware. Some vendors offer discrete or specialized devices to support these roles, and others use intelligent line cards in switches and routers. The decision on which hardware to use may be made by default if the infrastructure to carry LTE traffic from the towers to the service point has already been deployed and doesn't offer a "smart card" option, or where the wireless operator is outside its wireline market area and doesn't own the lower network layers. In cases where the choice is available, it would be smart to review the capabilities of the integrated smart cards. Normally, these will reduce capital and operations costs and reduce traffic latency by minimizing box-to-box connections.
At the deepest level, where the Evolved Packet Core integrates with existing infrastructure, it is important to review vendor options for traffic offload. Smart traffic handling at or near the eNodeB can segregate Internet traffic from wireless service traffic and move it immediately onto the best-effort infrastructure used for broadband Internet connectivity. That will reduce traffic in the more expensive Evolved Packet Core components and improve service performance and operations costs. Offloading is especially critical for Internet video traffic, which can load the SGW/PGW and the associated tunnels significantly.
LTE network layers must link with service control and registration logic
A final point in the layer-by-layer review for LTE planning and design is to ensure that all of the components, especially the radio access network (RAN) and EPC components, properly link with the service control and registration logic. EPC components require service control input to manage traffic routing and to rebuild the connection map for mobile users. The links must be stable and secure while also being compatible in terms of format and interface standards with the equipment that operators plan to deploy. Specific lab testing is recommended.
Lab testing is a good idea for LTE layer planning in all cases. Operators report that the best approach is to have two distinct test models: one that represents the "final form" of an evolved infrastructure, and one that represents the evolution itself. When the end-state components have been verified in the first test bed and any longer-term interoperability issues resolved, the evolutionary test bed can be used to model each stage of deployment to ensure success.
About the author:
Tom Nolle is president of CIMI Corporation, a strategic consulting firm specializing in telecommunications and data communications since 1982. He is the publisher of Netwatcher, a journal addressing advanced telecommunications strategy issues. Check out his SearchTelecom.com networking blog Uncommon Wisdom.
Mobile appliance explosion changes LTE broadband network planning
Building a better Evolved Packet Core for 4G LTE networks
4G LTE network planners face five Evolved Packet Core issues
Review the basics with our Evolved Packet Core primer