Mobile data from a mobile device and the server traverses through different types of intervening networks; the cellular network, the Internet and the Data Center Network. A typical cellular network consists of three distinct and separate networks namely, Access Network – providing connectivity to mobile over the air interface along with cell towers and equipment servers for Radio operations, Core Network – network elements responsible for authentication, access Control, Security, and mobility functions of, and lastly Service Network – which is typically a part of the Internet but controlled by a network operator to offer services to end users.
Optimization Approaches for Mobile Data Growth
With an exponential increase in the number of mobiles and bandwidth-hungry applications like video streaming or virtual reality based games, the bandwidth requirements would exceed the network capacity of the currently deployed cellular networks. Also, the bandwidth requirements are dynamic in nature with respect to time. The Data Center network generally has sufficient capacity with high-speed switches and is not a bottleneck. Also, the Internet along with service network are mostly with wired optical fibers where additional bandwidth can be added as and when required. However, access and core networks may become overloaded due to an increase in the mobile data traffic and hence can be optimized by adopting some of the following approaches:
The next section describes how optimization leads to monetization.
Monetization in Next-Gen Networks
The ever growing and variable network resources requirements in next-gen networks need flexibility which can be served by adopting SDN/NFV based architecture for cellular networks.
SDN and NFV along with a bundle of open source APIs and Test Framework help to leap towards the next generation network, though the success depends on a deep-dive into the use cases, planning, and migration. Hard-wiring will be replaced by software defined soft-wiring which is easily maintainable, scalable, and programmable, thus delivering the promise for better resource utilization.
How NFV helps?
Network elasticity: For example, dynamic scaling of the base station from a few virtual machines based on the load in the coverage area. This reduces the need for pre-provisioned resources.
Consolidation: vEPC and Virtual Base Station can run in the same data centre.
Resource sharing with third party network function creates a competitive environment for vendors to innovate network function at a lower cost.
Enemies for NFV?
Maintaining tight requirement for round trip time due to the higher interrupt latency and overhead associated with VM switching in some cases.
Automated, programmable lifecycle management of network nodes leading to a reduction of the capital expenditure for idle usage of the network.
Quick certification of new services.
The shift towards value-based services to enable the integration with over-the-top content providers.
A huge amount of data produced from the network, when synthesized and used as feedback to the SDN controller, can be a real service differentiator.
Analytics - THE MASTERMIND
Analytics can be performed on both types of network data namely “Data at Rest” like call data records for data mining techniques, as well as “Data in Motion” like traffic for real-time prediction of the state of the network. The analytics insights obtained can prescribe modification to the configuration of the network, offer new services, and improve real-time customer service and experience. “Closing the loop” plays an important role and consists of collecting and analyzing data to figure out how the network can be optimized, and then implementing those changes in an automated way. In Access Network, such changes can be triggered for load balancing and interference management, whereas congestion control and better mobility management can be achieved in backhaul & core network.
In Next Generation “2020 Network”; SDN, NFV, and Analytics would play key roles in allocating network resources, and enabling appropriate computing and networking functions as per dynamic real-time access behaviors of the volume of traffic, with optimum Quality of Service and Quality of Experience.