Please complete the form below to access the complete report and in-depth interview transcripts.
Mobile operators have to manage a volume of traffic that increases faster than their network capacity. As they strive to provide a consistently high QoE to meet the expectations of their subscribers, they have started to move beyond a best-efforts approach, which treats all radio traffic the same way. This is because alongside the volume growth, the traffic mix has become more complex, with multiple interacting factors – such as wireless interface, traffic type, application, device/OS, policy, and QoS – that determine specific requirements for high performance and QoE.
The varying wireless traffic composition deeply affects how mobile networks perform. Assuming a constant traffic load – i.e., a steady number of bits transmitted – network performance and QoE change depending on factors such as the percentage of video streaming (e.g., relative to browsing or messaging), the signaling load, the policy and traffic management adopted, and the RAN conditions.
To understand how network behavior and subscriber experience change as a function of traffic profile, operators have to shift to a testing approach that enables them to identify the impact on performance and QoE of the multiple factors that define mobile traffic. Tests limited to varying radio traffic loads are no longer sufficient in commercial networks on which traffic composition continually changes with the mix of devices, applications and services, with the concentration of subscribers, and with RAN conditions.
RAN conditions have traditionally been the focus of wireless traffic simulations for network testing. To capture the impact of traffic composition on network behavior and QoE, data simulations used for testing must also be able to reproduce the variety of traffic profiles that operators see – and expect to see – in their networks. Operators need the flexibility to define traffic profiles in which they can systematically and automatically change the parameters that matter to them – e.g., the split among devices, the activities of subscribers, or the utilization level of their network.
In this paper, we explore how the growing differentiation in mobile traffic characteristics affects performance and QoE, and how this is driving a new approach to testing. Wireless testing has to assess network performance under a wide range of RAN and traffic scenarios to give mobile operators a realistic and complete view of what their networks can do and how they behave under stress. To achieve this goal, mobile traffic simulations have to systematically reproduce the variability of radio traffic composition expected in commercial deployments, in addition to the RAN conditions on which network testing has focused traditionally.
The white paper was sponsored by PRISMA Telecom Testing