PARTNER CONTENT: Network planning and optimisation are critical to the efficient use of communications service provider (CSP) networks. These functions will ensure that CSPs obtain maximum RoI, manage the cost of operations and deliver optimal quality of experience to customers. However, the migration to a new network technology brings new challenges to these functions. This will be the case with the current migration of networks to 5G.

The arrival of 5G promises to provide new opportunities for CSPs to increase service revenue and achieve cost efficiencies. By using advanced capabilities such as Massive MIMO, network slicing, edge computing and network function virtualisation, CSPs can offer new services including smart home, autonomous driving, drone operations and remote healthcare services. CSPs can also see the cost of operating networks fall in the long term. Implementing these features and technologies, however, present challenges to CSPs’ planning and optimisation operations. These challenges include fulfilling the diverse set of network requirements needed for each service, limitations associated with spectrum bands supporting 5G, the co-existence of 4G and 5G cells, and the complexities that come with implementing Massive MIMO.

Lessons from existing 5G deployments provide insight into ways CSPs can address these 5G network planning and optimisation challenges. These include taking a service-oriented approach to planning 5G networks, using 3D modelling to plan and optimise 5G networks, and prioritising 4G network optimisation.

  • Taking a service-oriented approach: Given the broad portfolio of services that 5G will enable, 5G network planning should begin with assessing the services the CSP plans to deliver. This assessment will include identifying the location of the target customers, determining the consumption pattern of these services relative to time and the network requirements for each service. With this assessment, network planning teams can determine the baseline service performance indicators for these services, which can then be mapped to the corresponding network performance indicators that must be achieved. The network planning team can then create a network plan that fulfills these KPIs.
  • Leveraging 3D modelling: 3D-site modeling will be relevant to 5G network planning and optimisation. It will enable CSPs to take into consideration the beam-based coverage and transmission patterns of the Massive MIMO antennas used in 5G sites. It can also address the fact that target 5G users (devices and humans) will be indoors and at varying heights above ground level. These capabilities can be achieved by taking the 3D characteristics of the antennas and the surrounding environment to ensure site plans are accurate.
  • Prioritising the optimisation of 4G networks: With current 5G deployments based on 5G non-standalone architecture, the performance of the 4G anchor network will be critical to the performance of the 5G network as it caters for its coverage needs. Therefore, resolving as quickly as possible any coverage or network performance issues in the 4G anchor network will assure the performance, throughput and other quality indicators of the 5G cells. One of the ways this can be achieved is by creating multiple anchor-point frequencies within the frequency bands that the 4G network relies on to ensure that 5G users remain connected to these frequencies.

To implement these approaches successfully, operators must be ready to adopt best practices that will enable effective 5G network planning and optimisation. These include standardising processes involved in optimising networks and keeping site plans updated.

Investing in big data and AI and in the accompanying infrastructure to support the increased computation workloads for these technologies will be required. For example, the use of machine learning can support the analysis of current and future consumption patterns of AR- and VR-based services. These technologies can also support the forecasting of 5G network coverage and capacity and identify the optimal parameter combinations required to achieve high spectral efficiency and optimal network performance in Massive MIMO antenna systems.

Finally, CSPs should invest in the skills that align with the relevant technologies and practices discussed. How a CSP decides to build up these skills will vary depending on their level of maturity when it comes to deploying 5G networks and implementing technologies such as AI and big data. CSPs can consider working with peers already running 5G networks or work with vendor partners that have built experience in planning and operating 5G networks using the methodologies discussed.

By Adaora Okeleke, senior analyst, Telecoms Operations and IT, Ovum