The End is Near: Sunset of Legacy 3GPP Technologies in Support of 5G

While 2020 has been the year for 5G deployments, it has also been the year for legacy technology sunset, as a rising number of operators join their peers to announce plans to shut down 2G (GSM) and 3G (UMTS) networks during the year 2020. Even though there is no globally unified strategy to phase out 2G and 3G networks, there is a cross-continental pattern visible:

In Europe, Deutsche Telekom and Vodafone have announced their plans to phase out 3G networks by the end of 2020 or during 2021 at the latest. Likewise, Norwegian operator Telenor announced its 3G switch-off for the end of 2020, while it intends to continue 2G networks until 2025. North America saw the shutdown of the last remaining 2G networks with T-Mobile in the United States and Movistar in Mexico in 2020. While countries in the Asia-Pacific region (Australia, Japan, New Zealand, South Korea or Singapore) have been leading the way to phasing out GSM and UMTS, operators in developing countries like Bangladesh, India, Macau and Myanmar have announced their shutdown plans for mid- to late-2020s. It should be noted, however, that these plans are far from being bullet proof, as Verizon announced in January 2021, to halt its plans to turn off its 3G network indefinitely. 

While phasing out legacy generation of cellular connectivity is an important step towards successful 5G deployments, it presents CSPs with additional challenges that need to be overcome in order to successfully transition from 2G/3G networks to 5G deployments in the near future. The recent announcement of Verizon to cancel their strategy for 2G and 3G sunsets until further notice highlights the dilemma that CSPs are currently in.

The importance of 2G & 3G sunsets for 5G deployments can be summarized in two words: Spectrum scarcity. The vast majority of C-band (which is the sweet spot for 5G deployments, because of its optimal combination of bandwidth and propagation) is already allocated to CSPs, leaving very little spectrum blocks for future allocations. Furthermore, remaining spectrum blocks are scattered around the band, which makes combining different spectrum blocks difficult as it raises concerns about interference. 

In addition to this, the efforts of the telco industry to convince regulators as well as the World Radiocommunication Conference 2023 (WRC-23) to award additional spectrum resources, such as the 6GHz-band, to cellular technologies, are not expected to be overly successful: While industry associations like the 5G Automotive Association (5GAA), for example, has argued for the entire 6 GHz band to be awarded for cellular connectivity for C-V2X use cases (following their White Paper A visionary roadmap for advanced driving use cases, connectivity technologies, and radio spectrum needs) the U.S. American Federal Communication Commission (FCC) decided to award 30 MHz to cellular technology & the remaining 45 MHz to unlicensed technologies. Regulators in other national markets are expected to follow this approach.  While there are vast amounts of spectrum resources available in mmWave bands its poor propagation characteristics would make it uneconomical to be deployed as coverage layer, where spectrum on the sub-6GHz bands would be ideal.  

This puts CSPs and the telco industry in a difficult position. On the one hand, sub-6GHz spectrum scarcity will jeopardize the success of 5G, as it needs enough spectrum resources for a solid coverage layer. As spectrum assets on the sub-6 GHz frequency bands have already been awarded entirely, sunset of 2G (EDGE) and 3G (UMTS) is an important building block to free up much needed spectrum for a 5G coverage layer. On the other hand, 2G and especially 3G are still important networking technologies, predominantly in developing countries: Following ABI Research’s Network Technology and Markets Tracker, 2G and 3G networks still account for 15% of all generated mobile network traffic in the Middle East & Africa (MEA) region.

In short, what is required from CSPs is a difficult balancing act to free up enough spectrum resources for 5G deployments without jeopardizing the economic development of regions that still rely on 2G & 3G connectivity. 

In developing a durable and long-term reliable strategy, CSPs need to consider two aspects: On the one hand, successful 5G deployments will need additional spectrum resources, as C-band spectrum is almost completely allocated already and there is strong momentum towards dedicating the 6 GHz band to unlicensed technologies, like WIFI. On the other hand, CSPs need to bear in mind that in some regions, particularly 3G still is an important networking technology and a premature sunset would jeopardize economic development. To account for these considerations, CSPs need to develop and communicate reliable strategy for 2G and 3G sunsets considering the following aspects:  

• Impact Analysis: Firstly, CSPs need to assess the impact of shutting down 2G and 3G networks for the existing connectivity landscape and their existing customer base. To do so, CSPs need to look at the number of devices using 2G and 3G connectivity as well as the share of mobile data that is generated via 2G & 3G networks.

• Microeconomic profit & loss analysis: CSPs will also need to ensure safeguarding & increasing their own profitability by shutting down legacy technology. Therefore, any decision against 2G & 3G networks and in favor of 5G deployments should only follow a careful comparison of the experienced Average Revenue Per User (ARPU) in 2G & 3G networks with the expected ARPU in 5G networks to calculate economic profitability of a shutdown. 

• Assessment of Economic Development & Requirements: Last but not least, in designing spectrum refarming strategies, CSPs need to assess the stage of economic development and the future requirements towards cellular connectivity within the respective region. If these can be addressed by 2G & 3G connectivity, the incentive to switch to 4G or 5G connectivity will be low. Subsequently, sunsetting 2G & 3G would risk losing the existing customer base instead of upselling them to 4G or 5G connectivity.

In response to all of these aspects, CSPs should adopt a regional approach to sunsetting 2G and 3G: Mature markets like Europe, North America or certainly the developed part of the Asia-Pacific region (e.g., Australia, China, New Zealand, South Korea, Singapore, Thailand etc.) should be first to sunset 2G & 3G, while developing markets in Latin America, the Asia-Pacific region or Africa should only follow once there are sufficient indicators suggesting that CSPs as well as the respective national society would benefit from shutting down legacy cellular technology to make spectrum available for 5G.

In doing so, CSPs need to give special attention to enterprise deployments, as a sunset will affect a substantial proportion of existing M2M and IoT connections: In North America alone, there are more than 120 million IoT connections running on 2G and/or 3G networks. Poorly communicated or managed shutdown of 2G and 3G networks could result in unexpected network outage with severe financial consequences. To mitigate these and therefore protect CSPs’ existing B2B business, operators need to offer their support with specific transition plans. Primarily, these plans need to include considerations for retrofitting existing production infrastructure with new network technology and provide a reliable cost analysis. Furthermore, it needs to ensure a smooth transition without any network outage that could be harmful for production workflows.