In March, the China Academy of Information and Communications Technology (CAICT), together with China Mobile and Huawei, publicly reported a terahertz wireless transmission test claimed to reach 1 Tbps over a distance of around 300 meters, with the terahertz link interfaced into an existing 800G optical transport network. Independent technical reports on terahertz prototypes from major vendors have so far described lower rates over comparable or longer distances, so the specific figures should be treated as a vendor-reported announcement rather than a peer-reviewed result. Either way, the development is significant for one reason that is often missed in coverage of the news: the test is not a story about replacing fiber. It is a story about how strongly 6G will continue to depend on fiber optic cable infrastructure.
For network operators, telecom integrators and infrastructure planners, the more useful question is not "how fast is the wireless link" but "what does this mean for the optical layer underneath." This article looks at that question.
Why 6G Still Depends on Fiber Optic Networks
Every generation of mobile network has made the radio side faster while pushing far more traffic onto fiber. 5G accelerated this trend by densifying base stations and shifting most of the heavy lifting - fronthaul, midhaul, backhaul, transport - onto the optical layer. 6G is expected to extend the same logic, only at a steeper slope.
According to the ITU-R IMT-2030 framework, 6G targets six usage scenarios: immersive communication, hyper reliable and low-latency communication, massive communication, ubiquitous connectivity, AI and communication, and integrated sensing and communication. None of these scenarios can be carried by the radio link alone. Each one assumes a dense, low-loss, high-capacity optical transport network behind every radio site, every edge node and every data center.
This is the essential point that the recent terahertz announcement actually reinforces. The test is described as "terahertz radio interfaced with an 800G all-optical network." In other words, the value of the wireless breakthrough only materializes if there is already an 800G-class optical layer waiting to absorb the traffic. The faster the radio gets, the more demanding the fiber underneath becomes.
What the 1Tbps Terahertz Test Means for Optical Cable Infrastructure
Setting aside the headline number, the technical claim with the largest implication for cable infrastructure is the integration between the terahertz link and an existing optical transport network - without intermediate protocol conversion. Carriers have been moving in this direction for years, with the goal of removing electrical-domain bottlenecks between the radio site and the metro core.
For optical cable planning, three points follow:
- Higher per-site capacity, not fewer sites. Higher-frequency radio (mmWave, sub-terahertz, terahertz) attenuates quickly in air and through obstacles. To deliver the rates 6G is targeting, networks will need denser radio sites - which means more fiber optic cable feeding each base station, not less.
- Higher fiber count per route. When each site demands tens or hundreds of gigabits, the metro and aggregation network has to carry a multiple of that. Cable types optimized for high fiber count, such as ribbon designs, become more relevant.
- Tighter optical performance. 800G and emerging 1.6T transport pushes coherent optics into a tighter loss and dispersion budget. Standard outdoor cables that were "good enough" for 10G/100G may not be adequate for long-haul links operating at 800G with tight margins.
Fiber Backhaul, Midhaul and Fronthaul Requirements in the 6G Era
Mobile transport is usually split into three segments. Each one is affected by the move toward 6G in a different way.
Fronthaul: from base station antenna to baseband
Fronthaul is short-reach, latency-sensitive and often runs in tight outdoor or in-building paths. Today this is dominated by CPRI/eCPRI links riding on dedicated fronthaul cables. As 6G radios push toward higher symbol rates and tighter timing, fronthaul fiber must offer low loss, predictable latency and mechanical robustness against bending, vibration and weather. FTTA (fiber-to-the-antenna) cable is the workhorse here, and 6G densification will pull more of it into both macro and small-cell deployments.
Midhaul and aggregation
Midhaul aggregates traffic from clusters of cell sites into the metro edge. With 6G traffic profiles, this segment will move from 100G/200G toward 400G and 800G in many networks. Aggregation rings are typically built with aerial or duct-based outdoor cables; in environments where there is no available duct or it is uneconomical to dig, ADSS fiber optic cable is the default choice for stringing aggregation along power and transport corridors.
Backhaul and metro transport
Backhaul carries aggregated mobile traffic to the core and into data center interconnect networks. This is where the 800G all-optical network referenced in the recent test lives, and it is also where coherent transmission distances and span budgets matter most. Operators planning for 6G are increasingly specifying low-loss G.654-class fiber for new long-haul builds, since it directly improves the reach and capacity of 800G coherent optical modules.
What Types of Fiber Optic Cables Will Support 6G Networks?
There is no single "6G cable." Different layers of the network have different physical, mechanical and optical requirements. The table below summarizes the main mappings:
| Network segment | Typical role in 6G | Cable types commonly used | Key fiber characteristics |
|---|---|---|---|
| Tower / antenna | Fronthaul to active antenna units | FTTA cable, hybrid power-fiber composite cable | G.652.D or G.657.A2; bend-insensitive; rugged jacket |
| Aggregation ring | Cell-site aggregation, metro edge | ADSS, aerial figure-8, duct cable | G.652.D / G.657; high tensile strength; environmental rating |
| Long-haul backbone | Inter-city and DCI transport, 800G+ | Loose-tube outdoor, direct-bury, submarine | G.654.E low-loss single-mode fiber |
| High-density routes | Metro core, data center, cloud edge | Ribbon fiber optic cable, micro-duct air-blown | High fiber count (288, 576, 864+); mass fusion splicing |
| Data center and AI cluster | Server, switch and GPU interconnect | MPO/MTP assemblies, indoor multi-mode and single-mode | OM4/OM5 or single-mode for 400G/800G; ultra-low insertion loss |
The pattern is consistent: 6G does not change the fundamental cabling categories, but it raises the performance bar in each one. A network that meets 5G specifications today will still need to be progressively upgraded over the next decade, especially on the long-haul and aggregation segments.
6G, All-Optical Networks and the Future of Telecom Cabling
The broader industry direction is toward an end-to-end all-optical network: the optical layer carries traffic from the access edge to the core with as few electrical conversions as possible. Operators have already been deploying 400G and 800G in metro and DCI. ITU-T G.654.E low-loss fiber, optical cross-connects, ROADM technology and coherent pluggables are being normalized into standard transport architectures.
6G accelerates this. The integrated sensing-and-communication scenarios in IMT-2030, AI-native traffic patterns from large model training and inference, and ubiquitous connectivity (including non-terrestrial networks) all push more traffic into the same optical backbone. The terahertz radio test announced in March is one of many signals that the industry is preparing for this load - but the actual capacity is being built in glass, not in the air.
For an extended look at how the optical layer is evolving in parallel with mobile generations, see our deeper analysis of 6G and fiber optics in ultra-high-speed networks.
Practical Implications for Network Operators and Cable Buyers
For operators, integrators and project owners planning network expansions in the 2026-2030 window, four practical takeaways follow from the current trajectory:
- Specify with the next upgrade in mind. Cables installed today on backbone and aggregation routes will likely carry 400G to 1.6T traffic within their lifetime. Choosing low-loss fiber and adequate fiber count up front is far cheaper than re-trenching.
- Account for site densification. 6G radio physics means more sites per square kilometer in dense urban areas. Plan duct, sub-duct and aerial routes accordingly.
- Treat fronthaul as a discipline, not an afterthought. As radio interfaces tighten, FTTA, hybrid power-fiber composite cable and short-reach high-precision assemblies become more critical to RAN performance.
- Align cable choice with all-optical strategies. If the operator's roadmap includes ROADM, OXC and end-to-end optical switching, link budgets must support that, which has direct implications for fiber type selection.
FAQ
Q: Does 6G Replace Fiber Optic Cables?
A: No. 6G is a radio-access generation, not a transport technology. The radio layer ultimately connects to fiber. Higher 6G capacity increases - not reduces - the load placed on the underlying fiber optic network.
Q: Why Does Wireless 6G Still Need Fiber If It Is So Fast?
A: Terahertz and sub-terahertz radio attenuates quickly with distance and is easily blocked by obstacles. To deliver the rated speeds at scale, 6G needs many small, dense radio sites, each one connected back through fiber for fronthaul, midhaul and backhaul. The faster the radio, the more fiber capacity must sit behind it.
Q: What Fiber Cables Are Used For 6G Base Stations?
A: At the antenna and tower, fronthaul typically uses FTTA cables and, where remote radio units need both power and signal, hybrid composite cables. Aggregation from cell clusters typically uses ADSS aerial cable or outdoor duct cable. Long-haul backhaul into the metro and core uses low-loss single-mode fiber such as G.654.E.
Q: What Is The Relationship Between 6G And 800G All-Optical Networks?
A: 800G is a transport-layer line rate that is currently being deployed in metro and DCI networks. 6G mobile traffic, especially in dense areas, will be aggregated onto these high-rate optical links. Vendor announcements that interface a terahertz radio link directly into an 800G optical transport network reflect this convergence.
Q: Will 6G Change Which Type Of Optical Fiber I Should Specify Today?
A: For long-haul and high-capacity routes, many operators are already moving from G.652.D toward G.654.E low-loss fiber to extend the reach of 400G and 800G coherent systems. For access and FTTH, G.657 bend-insensitive fiber remains the standard. The 6G transition is unlikely to introduce a brand-new access fiber type, but it will continue to push backbone networks toward lower loss and higher fiber count.
Summary
The reported 1 Tbps terahertz test in March is one data point in a longer industrial roadmap that points to commercial 6G around 2030. For optical infrastructure, the more durable conclusion is structural: 6G amplifies fiber demand at every layer of the network - fronthaul to antennas, aggregation between cell sites, backhaul into the metro core, and the optical fabric inside data centers. Operators and network builders who plan their cabling with that trajectory in mind will avoid stranded investment as the next decade unfolds.