Operators can’t escape carrying wireless traffic on their metro networks unless they don’t have any wireless traffic in the first place.
So 4G network offload strategies that use the mobile backhaul network are focused on moving traffic to its destination at the lowest possible cost.
Internet traffic is the primary mobile backhaul focus because operators get the lowest return per bit on IP, and they need to make sure that services that consume the expensive wireless infrastructure are the ones that bring in high revenue per bit.
Mobile backhaul strategy could change Evolved Packet Core design. The architecture for LTE backhaul is defined by the Evolved Packet Core and how its components relate to the rest of the metro network. The Evolved Packet Core can be visualized as tunnel-managed connectivity between the LTE serving gateway (SGW) and the public data network gateway (PGW).
The Evolved Packet Core can handle (QoS), but it will likely present premium costs for the ability to offer traffic-handling options. If the majority of 4G network traffic is regular IP rather than premium wireless traffic that brings in more revenue, it makes sense to move the two gateways closer together (or combine them into a single device). This would shunt traffic onto the metro network near the wireless network edge, thus reducing custom EPC equipment and handling costs.
Moving 4G gateway boundaries can help with network offload. Compressing SGW/PGW boundaries reduces the area of the metro network where the Evolved Packet Core has to manage QoS, which means that the “new” PGW must be able to move premium traffic onto metro routes already designed to handle traffic like applications under SLAs, for example.
If the metro network already carries other premium services traffic and supports classes of service, it may make sense to add 4G network offload traffic there rather than manage an EPC-based set of handling options for wireless traffic in addition to the QoS options already available for wireline services on the metro network.