The deployment of 5G appears to be finally gaining speed and scale. Yet, communications service providers (CSPs) in Southeast Asia continue to remain cautiously optimistic about its cost versus revenue proposition.

There are, of course, good reasons for this. 5G’s speed and coverage capabilities rely heavily on network densification, which requires not only the addition of towers and small cells to the network but also an evolution in the transport network and transition to a cloud-native core.

These cost considerations are further compounded by short-term revenue concerns—in particular, the lack of killer use cases in the enterprise segment, and limited appetite in the consumer segment for the premiums demanded by 5G-enabled devices.

The report by Deloitte titled “Unleashing the potential of 5G in Southeast Asia” weighs the cost-revenue balance as CSPs navigate 5G deployment in the region, facing challenges related to network densification and limited use cases while considering a range of 3GPP migration options.


Given that CSPs are at different stages of maturity in their transition to 5G, it is difficult to pinpoint a one-size-fits-all roadmap for deployment. Recognising this challenge, the 3rd Generation Partnership Project (3GPP)—a global organisation uniting seven telecommunications standard development organisations—has developed a series of six different 5G migration options for the industry. Based on the degree of forward and backward compatibility between its 4G Long-Term Evolution (LTE) and 5G New Radio (NR) control plane and user plane procedures—as well as the accompanying support provided by its wider ecosystem of original equipment manufacturers (OEMs), and device and chipset manufacturers—each CSP will likely need to consider a different set of options (see Figure 1).

Current industry consensus

It is important to note that while all six proposed migration options will be supported, the 3GPP intends for the industry to align on a single approach to minimise fragmentation of the 5G ecosystem from hardware, software, and interoperability perspectives. Furthermore, adopting a consistent approach could also considerably reduce complexity, lower system integration costs, and lessen the amount of testing that will be required before a system can go to market. Based on current industry consensus, the most cost-effective and future-proof trajectory is likely to be for CSPs to migrate from Option 1 (4G LTE SA network with EPC) to Option 3 (5G NSA network with EPC and LTE-NR dual connectivity), and then to Option 2 (5G SA network with 5G Core).

There are two main reasons for this. Firstly, this approach would be in line with the pace of development and maturity of 5G use cases in enhanced mobile broadband, ultra-reliable low latency communications (URLLC), and massive Internet of Things (IoT). Secondly, this approach would also maximise the re-use of existing assets, while ensuring a steady, manageable migration process to deliver the full range of benefits from 5G.

Two key considerations

In this section, we will examine two key considerations for CSPs in Southeast Asia as they embark on their journey of executing a cost-effective 5G migration process according to the abovementioned trajectory:

1. Leveraging a cross-domain design approach For CSPs: the ability to cost-effectively estimate capacity demand and translate it into network capacity is a never-ending challenge— and this conundrum is only going to become even more complex in a 5G world, where there would be a proliferation in the number of use cases (logical networks) that a given physical infrastructure will need to support. The use of a cross-domain design approach to network planning is, therefore, critical to enabling CSPs to sweep away the inefficiencies of traditional organisational siloes and make more informed investments in network upgrades that will ultimately underpin all their service offerings. Indeed, recent research suggests that leading CSPs who have leveraged a cross-domain design approach have benefited from an average 45 percent reduction in time-to-market, 30 percent decrease in total cost of ownership, and 14 percent increase in end-to-end service reliability.

To understand how the cross-domain design approach works, it may be instructive to consider its implementation for a specific use case—in this instance, the planning and design of radio sites. Briefly, such an exercise will entail an assessment not only of the radio sites per se but also of the surrounding transport networks and edge data centres. The two steps in this process are broadly as follows. Firstly, we will conduct a traffic analysis. Using call records, we can create a highly granular 3D map of an urban area illustrating the call volume and data throughput for the different streets and buildings—even down to the specific data for each floor within a building. Secondly, we will conduct an analysis of the site build costs, including those required for power deployment and construction of fibre towers.

The result would then be a radio plan, with a highly detailed cost-benefit analysis of network coverage and cost for each site. This plan could, in turn, serve as the input for a data centre planning tool to be used to identify suitable locations for computing and storage capabilities (or edge data centres) across the network, while taking into consideration requirements for physical space, power, and cooling.

2. Determining trade-offs in migration priorities Without doubt, the 5G NR is expected to account for the lion’s share of a CSP’s migration effort in terms of capital expenditure, overlay, time, and effort. The 5G NR alone, however, is by no means sufficient for the CSP to realise the full range of benefits that 5G offers. Indeed, the order of priority for CSPs should be to firstly get to the 5G Core; secondly, ensure that their transport network can do the heavy lifting; and thirdly, focus on the deployment of the 5G NR.

In the migration from 4G LTE to 5G NR, CSPs should also consider re-using existing assets wherever possible, including spectrum, to keep their total cost of ownership under control. Having a solid 4G LTE network, therefore, is an important step to laying the foundation for a seamless migration to 5G NR—especially given that both are expected to co-exist with each other for many years to come. CSPs should also deploy the extensive toolkits that they have at their disposal to expand 5G network coverage and capacity, while balancing these priorities with their legacy investments in 4G LTE.

Furthermore, when CSPs successfully arrive at the Option 2 scenario under our aforementioned trajectory, 2G and 3G networks will most likely cease to exist. In addition, we could probably expect a logical split in the 4G and 5G architectures, which will be necessary both to improve the efficiency of 5G systems as well as maximise spectrum utilisation by harmonising the user-plane stack6. Indeed, the higher up the user-plane stack this harmonisation can be achieved, the larger the gains for CSPs. But a word of caution is in order: CSPs must carefully manage the trade-offs between 4G LTE and 5G NR, and determine the optimal level of co-existence that they would like to have between the two. Beyond a certain threshold, it would become counterproductive to forgo potential 5G NR gains in favour of interworking with existing 4G LTE networks. Currently, most 5G deployments across the globe are taking place in the mid-spectrum bands; the only exception is perhaps the US, where some CSPs have begun deployment in high-spectrum bands7 (see Figure 2).

We expect, however, all bands to eventually be used for 5G, either through dynamic spectrum sharing (DSS) or spectrum re-farming. Ultimately, to ease CSPs’ capital expenditure burdens and expedite rollouts, 5G deployments must also be made airwave-agnostic—and in this respect, regulators will have a vital role to play.

Access the full report here.