In 2025, what directions in the communications industry are worth paying attention to?

Looking back on the past year, I personally feel that the entire communications industry is actually quite dull and boring. There are no big surprises or big shocks. Occasionally, there are a few big news (such as the R18 freeze, the start of 6G standardization, and the domestic Starlink launch), which are all planned.

So, how will the communications industry develop in 2025? Will it be as dull as in 2024?

In today’s article, Xiaozaojun will talk about my views on the coming year from the perspective of industry self-media.

Overall Trend

The reason why 2024 is dull is mainly because the entire industry is in a cyclical trough.

The peak of 5G and gigabit optical network construction has ended, a round of technology iteration has been basically completed, and user needs have been met to a certain extent.

At this stage, there are generally two things to do:

The first is to observe whether users are satisfied with the new technology and whether the effect of technology implementation is achieved. The second is to explore, collect new user and scenario requirements, conduct technical pre-research, and prepare for the next round of iterations.

To be honest, the situation facing the communications industry today is still quite complicated. The promotion of new technologies has not been as good as expected, and the bottleneck effect of basic technology theory has also added uncertainty to the future.

The prosperity and growth of the industry depends on user demand, which comes from scenarios and applications. The entire communications industry is trying to find popular applications to gain new profit growth and maintain survival.

At present, the rapid rise of AI has brought significant development opportunities to the communications industry.

Whether on the cloud side or the end side, AI technology has put forward higher requirements for communication capabilities and shows great dependence. This is expected to inject new development momentum into the communications industry and is likely to bring about a prosperous cycle.

In addition to AI, communications are also working hard to expand in capabilities and space.

The integration of communication and perception is a typical example of capability expansion. Low-altitude air-to-air networking, ATG (ground-to-air communication), and satellite Internet indicate that communication is accelerating towards the integration of air, space, and the earth, and realizing true ubiquitous full-space connectivity.

These fields currently have very broad application prospects and are attracting a lot of attention.

Next, let’s take a closer look at what are the popular development directions in the communications industry.

Traffic Management

Let’s start with the C-end (personal and family consumption) that everyone is most familiar with.

The search for popular applications on the C-end has actually reached a dead end.

After games, videos and live broadcasts, no new traffic applications have emerged in the consumer field. VR/AR is still not recognized by the market and has not yet formed a scale. Holographic communication is even more out of reach.

You will notice that our monthly traffic consumption has entered a stable state. The speed of tens of Mbps and the traffic package of 10GB can already meet the needs of most users. The total value of traffic related to consumer applications on the Internet has slowed down.

Without popular applications, users will not be able to feel the difference in experience brought by network upgrades. The early promotion of 5G was too high-profile, which made users "the greater the expectations, the greater the disappointment", which in turn generated negative emotions and even raised doubts.

This sentiment also extends to 5G-A. Users cannot understand: "Why is there a new generation when 5G has just been built? Are equipment manufacturers fooling operators? Is there duplicate investment and waste?" In a sense, this is indeed a backlash of over-promotion.

Feedback from users seems to have influenced the decision-making of operators. 2024 is the first year for 5G-A commercial use, but it seems that only China Mobile is actively promoting it in China. China Unicom and China Telecom are obviously slower in their actions, with a hint of wait-and-see attitude.

5G-A can reach speeds of over 5Gbps in the case of 3CC carrier aggregation. However, users are no longer interested in speed tests. Unless there are real apps that can use this speed, it is difficult for them to have a rigid demand for 5G-A and pay for it.

When the path of increasing speed is not feasible, a new trend has emerged in the mobile communications industry, which is to guide the market from traffic management to experience management, and to generate new consumption by providing differentiated network experience.

Simply put, it is just like selling economy class and business class tickets on airplanes. Through exclusive packages or data packages, high-end customers or customers with special needs (live broadcast UP hosts, game users, etc.) are provided with higher priority QoS (service quality) guarantees, increasing the operator's ARPU (average revenue per user).

The communication industry has a clear idea: since there is no higher requirement for speed, then find a way to solve the problem of coverage and experience. On the one hand, through actions such as signal upgrade, the signal quality in subways, stations, airports, tunnels and other areas is enhanced. On the other hand, for high-end users, paid priority optimization and guarantee are provided.

This is an important trend in the consumer mobile communications market. In 2025, we will see operators launching more related services.

AI on the edge

In addition to differentiated experience, another important trend in the consumer sector is the explosion of AI-side applications, which may lead to corresponding popular applications.

In 2023-24, AI's focus was mainly on large models. Now, the focus has gradually shifted to application implementation. AI terminals, AI apps, AI agents, and embodied intelligence have entered a stage of rapid development.

We need to pay close attention to the impact that the explosion of edge-side AI may have on communications. It may be edge computing and edge-side computing, or it may be traffic growth, or an increase in intelligent communication traffic. How communications can serve edge-side AI and how to achieve efficient data entry into the cloud and computing are both worthy of in-depth study.

New services such as 5G new calls and 5G messages on the consumer side are now also embracing AI. Especially 5G new calls, which used to be simple video and audio channel transmission, will be implemented after the DC channel standard and industry mature (most likely this year). You can pay attention to a wave of user feedback.

At present, if there is really a hit application in the consumer field, it is likely to be an AI application. The terminal form is not limited to mobile phones, it may also be embodied intelligence (robots) or other intelligent entities. The AI ​​era is really here, and it will be too late if you don’t get on board.

Cluster Network

The impact of AI on communications is reflected in wireless technology on the user side and in wired technology on the core side, or in optical communications.

This trend was already evident when AI first broke out in 2023. Xiaozaojun mentioned it in last year's summary article.

The training of AIGC large models depends on computing card clusters. Computing card clusters have a strong demand for ultra-high-speed (800G, 1.6T) optical communications, which directly promoted the explosion of related industries and also made great progress in optical communication technologies such as SerDes, silicon photonics, optical chips, DSP, coherent light, CPO/NPO/LPO, and hollow-core optical fibers.

Today, as the scale of large model parameters continues to grow, computing card clusters are also moving from thousands of cards to tens of thousands and hundreds of thousands of cards. Traditional single data center clusters can no longer meet demand, and multiple off-site data centers are needed to form larger computing card clusters.

This brings opportunities for long-distance optical communications related to DCI (Data Center Interconnect). The domestic backbone network 400G upgrade has made some progress, and the subsequent research on 1.6T related solutions is being stepped up.

AI will be the main driving force for high-speed optical communication technology. The market value of related industries still has room for further growth.

Network Intelligence

The Internet empowers AI, and AI also empowers the Internet in turn.

In recent years, all equipment manufacturers and operators have been studying how to introduce AI to improve the performance of existing communication technologies.

As we all know, existing communication technologies are already approaching the Shannon limit. Apart from improving frequency resources and the number of channels, it is difficult to further increase network speed. The continuous evolution of communication technology has made network architecture more and more complex, which has also brought great pressure to operation and maintenance.

The emergence of AI may bring a breakthrough to solving these problems. At present, there are cases of AI intelligent improvement in wireless resource management, network planning and optimization, business analysis and prediction, troubleshooting and processing, etc. AI can help improve network performance, simplify network operation and maintenance, improve efficiency and reduce costs.

Some companies have also proposed large communication models and network agents, trying to help operators better manage and control the network through these technologies.

In 2025, the application of network intelligence will become more and more common, and network autonomy will accelerate into the L4 stage.

New Internet (IPv6+)

At the network layer, it is worth mentioning the new generation of Internet represented by "IPv6+", also called the new quality Internet.

The upgrade and transformation of "IPv6/IPv6+" has been going on for some time. The penetration rate is constantly increasing, but users do not seem to have a very obvious perception of it.

Today, the new Internet based on the "IPv6+" technology foundation also takes AI into computing, within computing, and between computing as its own service goals. The entire Internet is too large, and the technology generation upgrade will be a long process.

In addition, satellite-to-satellite interconnection is also seen as one of the main scenarios for the future of the new Internet. As mentioned earlier, one is to serve AI, and the other is to serve ubiquitous connectivity.

50G-PON

In the field of fiber-optic broadband (wired access network), what we need to focus on is 50G-PON.

At this time last year, I predicted that 50G-PON would be put into trial commercial use. My prediction was correct, but the scale of the trial commercial use was smaller than expected. The reason for this is the same as 5G-A, that the demand from users is not strong enough. Of course, the lack of maturity of the industry chain is also one reason.

In 2025, I believe that 50G-PON will be more commercially available. In addition to home users, 50G-PON may have more applications in the B-end and G-end markets such as campus networks.

In the field of fixed-line communications, the relevant industry chain has proposed the concept of F5G-A, but it does not seem to be very popular at present.

In addition to 50G-PON, Wi-Fi 7 was also fully commercialized last year. However, due to the limitation of the 6GHz frequency band, this technology cannot show the difference with Wi-Fi 6/6E, and is not considered successful, with a bleak future. In contrast, domestic protocols (such as Xingshan) and new private protocols (network-free communication) are worthy of attention.

Private network communication

Let’s talk about the B-end market.

The performance improvement brought by 5G/5G-A is not very attractive to C-end users. However, it is still very meaningful in vertical industry scenarios, especially industrial manufacturing, Internet of Vehicles and other scenarios with high requirements for network performance.

In 2024, Xiaozaojun participated in many "5G Mountain and Sea Tour" research activities and visited many companies' digital transformation. Most of these companies are large and medium-sized enterprises with good profitability, with both funds and technology, and a solid information foundation, and can also hold the arduous task of digital transformation.

However, digital transformation is not just the task of a few leading companies, but should involve more companies. This involves the low cost, feasibility and replicability of relevant ICT solutions.

Industry scenarios are too complex. Even in the same industry, it is difficult for different companies to simply copy solutions. This is the main problem that plagues most companies' digital transformation. This problem is expected to continue in the next few years.

Enterprises’ digital transformation requires private network communications. However, there is a problem here: enterprises do not want operators to dominate the construction and operation of their private networks.

Private networks are different from public networks. Public networks involve the entire city, so a public operator is definitely needed to lead the network construction. Private networks are only a small area and belong to the private area of ​​the enterprise, just like our homes. If we can use Wi-Fi, we don’t want operators to get involved. There are both cost factors and data privacy factors involved.

There are three essential elements for building and using a communication network: equipment, frequency resources, and engineering (installation and operation and maintenance).

For enterprises, they can directly purchase equipment from equipment vendors, and for engineering, there are many companies that can provide services. However, only operators have spectrum resources. If free frequency bands are used, there will be too many restrictions and conflicts.

Therefore, in recent years, everyone has been paying attention to the issue of private network frequency bands. If private network frequency bands are directly authorized to enterprises, it will have an impact on operators. For some operators, they don’t know enough about products and solutions. Since they always bring equipment vendors to negotiate, why don’t enterprises go directly to equipment vendors?

This is a huge challenge for operators. It all depends on whether policies can be changed.

RedCap

Let’s talk about the Internet of Things market.

After a wave of reshuffles in the IoT market in the past few years, the current technology development path has become relatively clear. The rise of Cat.1 shows that cost plays a decisive role in market share.

The scale of IoT is so large that users are very sensitive to module costs. As long as the price is cheap and the network coverage meets the requirements (Cat.1 directly reuses the LTE network), users will be willing to choose it.

RedCap is a 5G youth version based on 5G. The entire industry has high expectations for RedCap. At present, it seems that the cost of RedCap will further drop to below 150 yuan, or even below 100 yuan. If it drops below 100 yuan, RedCap still has a great chance.

For operators, RedCap's wide-area continuous coverage is also one of the driving forces for them to build 5G-A. At least that's what China Mobile thinks. In many cities, China Mobile has initially achieved RedCap's continuous coverage, meeting the needs of wide-area coverage scenarios such as garbage collection and security cameras.

2025 will be a key year for RedCap's commercialization.

As mentioned earlier, in the field of IoT, passive IoT (P-IoT) is also worth paying attention to. This has potential market demand in warehousing logistics and supermarket consumption scenarios.

Synaesthesia

In addition to providing stronger service capabilities in existing scenarios, communications are also working hard to find new application scenarios. Synaesthesia is the most typical example.

Influenced by the rise of the low-altitude economy, synaesthesia is currently receiving a lot of attention. The essence of synaesthesia is actually the functional expansion of the communication network. Communication is no longer just communication, but has added the ability of perception. To put it bluntly, it is the combination of communication and radar.

This has opened up a very broad new field, with many more application scenarios. Low-altitude economy is just one dimension of synaesthesia. Synaesthesia can also be used for the perception of vehicles and ships, deformation monitoring of bridges, and so on.

From the perspective of low-altitude economy, the role of aircraft monitoring with integrated senses is only one aspect. Low-altitude aircraft flying requires a lot of basic services, including radar perception, low-altitude communication, low-altitude control, navigation guidance, etc.

Therefore, the industry proposed the "four networks" - "facility network", "air network", "route network" and "service network". All of a sudden, the story was told in a bigger way, and the market scale and application scenarios also became much larger.

Synaesthesia is a new field. Although both communications and radar are based on radio electromagnetic waves, they belong to different technical systems after all. Therefore, if the two are integrated, there are still many technical and industrial challenges. However, this integration and expansion is a rare opportunity and is likely to bring a long period of prosperity to communications.

ATG

In addition to functional expansion, synaesthesia is essentially a component of the larger framework of "ubiquitous communication".

Simply put, the ultimate development direction of communications, in addition to increasingly powerful performance and more and more functions, is that the network is everywhere. In every corner of the universe, as long as there are human activities, network coverage is needed. This is "ubiquitous communication."

Low-altitude economy and synaesthesia are aimed at low altitudes. Then, there must be corresponding solutions for medium altitudes, high altitudes and space.

For high altitude network connections, mainly for civil aircraft, we can pay attention to ATG (air-ground communication).

In 2024, operators actually have a lot of actions in ATG. On the one hand, they are promoting aircraft model modification and certification. On the other hand, they are also deploying ATG base stations along key routes.

I believe that in 2025, operators will have some important releases in ATG.

Satellite Communications

When it comes to ubiquitous communications, satellite communications must of course be mentioned.

Satellite communications, especially high-throughput satellites and low-orbit Starlink, have been hot for many years and the popularity continues to grow.

As Musk's Starlink is booming, we are paying more attention to the development of domestic Starlink. In the past year, a hot topic in satellite communications has been the direct connection of regular mobile phones to satellite communications. Related businesses launched by brands such as Huawei have attracted widespread attention from the public. Service providers such as Tiantong Satellite have also greatly increased their presence.

Not long ago, China Star Network finally began to officially deploy satellites, which is still worth paying attention to. The domestic Starlink national team has begun to take action, but to what extent it can achieve and whether it can replicate the success of Beidou remains to be seen.

We have an advantage in ground communications, but Starlink is very different from ground communications networks. We have gaps in commercialization, potential users, market size, launch costs, and many other aspects, and we have a long way to go.

The significance of Starlink is not just to provide emergency communications in special circumstances. It is essentially a multifunctional network with ubiquitous coverage capabilities, of which communications are only one aspect. It is very necessary to invest in space layout and occupy space positions and spectrum resources.

I personally hope to see the rise of China's commercial aerospace industry. If we want to grow bigger and stronger in the field of Starlink, we should support private enterprises more. Simply put, the satellite industry also needs companies like Huawei.

Final Words

There are many technologies worth paying attention to in the field of communications, such as quantum communication, semantic communication, intelligent metasurfaces, etc. I think these technologies will take some time, and it is estimated that it will be difficult to see very significant progress in 2025.

In 2025, the communications industry will continue to move forward by trial and error. After Trump takes office, uncertainties in all aspects will increase greatly, and the probability of black swan events will also increase significantly.

The overall situation of the communications industry is not optimistic, but AI is an important opportunity that must not be missed. The expansion of emerging fields such as low-altitude economy and satellite Internet is also worth looking forward to.

There are so many paths, and I hope that one or two of them can be successful, bringing some rain to the long-dry industry and also bringing some hope to us practitioners.

In the new year, I hope everyone can keep their jobs, stay healthy, be happy and live a happy life!