Decoding spectrum: How frequency bands work in telecoms

All wireless connectivity is based on electromagnetic frequencies, with different parts of the frequency spectrum being used for different purposes, such as 5G or 2G. The precise way in which this is decided, allocated out and regulated varies between country to country, and there is some politics involved in that, as well as international corporation as well.

Different spectrum bands have different characteristics and hence make them better or worse for particular uses – for example the higher the frequency, the worse the propagation, or ability to pass through objects.

It’s a highly technical issue, but really one that lies at the heart of the telecoms industry. To look at how it all works, how it is regulated, and what might change as we approach the 6G era, we assembled a panel of experts to talk us through the science.

Spectrum bands and how they are used in telecoms

Giving a broad definition of what me mean by the term spectrum bands, Luciana Camargos, Head of Spectrum at GSMA says: “Mobile uses spectrum to provide voice and data connectivity. Spectrum relates to the full range of electromagnetic frequencies, of which a portion is used for communications by a variety of services and industries. A small number of frequency bands is harmonised for the use of mobile across:

•             Low bands: between 400 MHz and 1 GHz

•             Mid-bands: between 1 and 7 GHz

•             High bands: between 24 and 66 GHz.

Spectrum and download speeds are intrinsically linked. The more spectrum there is available for any connection, the higher the potential download speed.”

Dan Grossman, Senior Analyst – Telecom Strategies at TechInsights gives an analogy: “You can think about spectrum bands like lanes on a highway. They are there to maintain order in a shared space that would be chaotic unless somebody – the regulators – keeps it organized. First, transmitters and their intended receivers must tune themselves the same frequencies in order to communicate. More important, if two transmitters were to start blasting out uncoordinated signals at the same frequency, those signals would interfere with each other, and the intended receivers would get nothing but garble. One policy to prevent interference is to license certain bands of frequencies – those lanes – to a transmitter for its exclusive use in a defined geography.

“Since the licensee controls all the legal transmitters in that band in that area, they can coordinate their transmitters and avoid interference. This is how mobile networks and broadcast radio operate. Another approach is for the regulator to make a band available to anybody that want to use it, as long as they follow defined rules for sharing with other users.  This is how WiFi (and others) works.”

Enabling wireless connectivity

Steve Blythe, VP of Group Spectrum, Orange said: “Spectrum is the fundamental asset that is required to enable wireless connectivity. Any wireless service, such as TV, radio satellite and mobile broadband, requires spectrum in order to work. Spectrum is however a scarce resource as there are physical limitations on the frequency ranges that can be used for any wireless service. So the usable range has been sub-divided into specific spectrum bands with each one being designated for a specific service. Using this type of band structure ensures that different services are able to co-exist with each other by limiting the amount of interference that services in adjacent spectrum allocations could cause each other.

“Mobile broadband services use licensed spectrum. For all spectrum bands there are international frameworks that determine which services are used in each band, a process that is managed at a global level by the ITU (International Telecommunications Union), part of the United Nations. Harmonisation of spectrum bands globally is important as it enables an effective eco-system of user devices to be developed with only limited regional variations. In Europe, the bands that have been defined for mobile broadband use include 700MHz, 800MHz, 900MHz, 1800MHz, 2100MHz, 2.6GHz and 3.5GHz. Other bands that are currently not used as widely are 1500MHz, 2300MHz and 26GHz.”

What are different parts of the spectrum better at or worse at doing?

So that’s broadly what the spectrum is – how it is used by telecoms operators and kit vendors depends on how different parts of it behave. Describing which parts of the spectrum better at or worse at doing, Doug Castor, Senior Director, Head of Wireless Research Lab, Interdigital said: “In general, lower frequency spectrum experiences less attenuation in the atmosphere and it is less susceptible to obstacles in the radio wave path, yet it offers limited bandwidth, so it is typically used to provide wide-area network coverage. Higher frequency bands, on the other hand, boast greater bandwidth, enabling them to provide higher data rate and capacity, however these bands have a shorter range and are more susceptible to obstacles.

Short range communications

“This makes them suitable for high-capacity, short-range communications like those needed in densely populated urban areas. In today’s cellular systems, we have three main bands. Frequency Range 1 (FR1), which is 0.41 – 7.125GHz, FR3, which is between 7 and 24GHz, and FR2, which is the high frequency millimetre spectrum between 24.25 and 71GHz. The numbering of these bands is based on the order they have been introduced to 3GPP standards, rather than ordered by frequency. For 5G, FR2 was newly introduced, enabling gigabit-per-second data rates without sacrificing network capacity, however it has drawbacks in propagation and power consumption. For 6G, FR3 is of intense interest as a perfect balance across range, capacity, and power consumption. However, the additional challenge is coexistence in those bands with other incumbent users, such as satellite.” 

Long range communications

Camargos added: “Spectrum bands have different characteristics, making them complementary and suitable for different purposes. Lower bands offer superior propagation characteristics, reaching longer distances, which makes them ideal for connecting rural and remote areas. The drawback is less available capacity, which impacts network speeds.  An additional benefit with low bands is that they have superior in-building penetration, which can be used to provide improved indoor coverage in cities. Mid-bands offer a good mixture of capacity and coverage. Those characteristics have made them the best option for city-wide capacity.

“The 3.5 GHz range has emerged as the pioneer band for 5G launches globally (approximately 70% of 5G networks now use this band, according to GSMA Intelligence). Also, most 3G and 4G networks were deployed in mid-bands. The last piece of the spectrum puzzle is high-band or mmWave spectrum. It offers the most capacity but with limited coverage. Especially in dense urban areas, high bands complement low- and mid-band spectrum with fibre-like performance. The roll out of commercial services in the US shows the ability to deliver gigabit speeds.”

Grossman described the importance of ‘loss’ when it comes to wireless signals: “It gets down to Laws of Physics. Visualize a radio transmitter as a light bulb. It throws light as a sphere. As you move from it, the light appears to get weaker – what engineers call “loss”. Loss increases by the square of your distance from the bulb. If you replace the lightbulb with a radio transmitter and move away from it again, there is more loss at higher frequencies than at lower frequencies. A weaker signal makes the receiver’s job more difficult, so it can decode less and less data, until it can’t tell the difference between the signal and background noise. Engineers say that the “range” of the radio is the distance at which the receiver stops performing well enough. Adding all of this together, you can see that higher frequencies have less range than lower ones. That’s why AM broadcasters have longer range than FM stations. 

“In the past, since high-frequency spectrum has short reach and doesn’t penetrate very well, it was less desirable than lower frequencies. Since there was little demand for it, large blocks of that spectrum were not assigned or used, giving the regulators in the present the opportunity to assign wide bands. Wider bands allow higher-speed radios.

“Thus, the trade-off: low-band spectrum has long range and can pass through buildings, but there is little of it available, and that only in small blocks, so the blazing speeds promised by 5G can’t be achieved with it alone. High-band (or mmWave) spectrum is plentiful and assigned in wide blocks, allowing those blazing speeds, but it has short range and doesn’t pass through buildings, which requires more base stations. Mid-band spectrum is a good compromise and is therefore most valuable to cellular operators.”

High frequency bands vs Low frequency bands

Kelvin Chaffer, CEO of Lifecycle Software added: “The portion used for wireless communication within the full spectrum range (3Hz-300EHz) sits from about 20 KHz to 300 GHz. Within this range, lower frequency bands like 700 MHz and 800 MHz (sub-1 GHz frequencies)  are crucial for providing extensive coverage and reliable connectivity as they can easily get inside buildings and walls. This makes lower frequency bands important for mobile and IoT device use cases in urban and rural settings as well as in any other operations requiring seamless connections, such as remote monitoring and smart agriculture.

“On the other hand, mid-band and high-frequency waves ranging from 2.5 GHz up to 28 GHz and 39 GHz can carry more data and provide exceptional speed for consumers and enterprises. These frequencies allow faster connection and high-speed internet access and enhance user experience on devices like 5G smartphones and tablets. Enterprises also utilise mid-band and high-frequency ranges for bandwidth-intensive applications, including virtual reality (VR) training and real-time video analytics. However, higher-frequency waves need a lot of power to travel, and they have limitations in coverage range compared to lower-frequency waves.”

Spectrum allocation, reallocation, and regulation

Of course, firms and organisations cannot simply go their own way when it comes to building wireless connectivity – there has to be organisation and corporation on what spectrum gets used, by whom, and for what.

Describing the process of spectrum allocation to operators, as well as any subsequent reallocation, Camargos said: “Auctions remain the most common approach globally. They work best when there is excess demand for spectrum, and they help to select the mobile operators most likely to use their spectrum assignment to benefit society. Administrative assignments, on the other hand, may be suitable in areas where there is less demand for spectrum and may allow authorities to compare the range of policy objectives offered by candidates.”

Ari Kynäslahti who heads up Strategy and Technology within Nokia Mobile Networks, highlighted the scarcity of spectrum as a resource: “Spectrum is a scarce and valuable resource that should be optimally used for the benefit of the society. The rules and processes for allocating spectrum to operators vary from country to country, but typically, the available spectrum is split between operators and licensed for a certain time frame, following an auction.

“Licensed spectrum provides exclusive rights of use without harmful interference, certainty for investments, and should continue to be the priority for regulators to ensure that networks provide guaranteed quality of service. There are many variables that determine the economies of spectrum allocation, including (but not limited to) exclusivity, timing to release new spectrum, availability of contiguous blocks of spectrum, duration of licenses, and pricing.”

Grossman said: “It’s politics, with lobbyists at every step. Mid-band spectrum is an extremely valuable resource, and there is never enough to go around. An incumbent user is loath to let go of it, while prospective users clamour for it. It’s doubly hard when the incumbent user is the national defence establishment. The regulators have to rob Peter to pay Paul. That means a legislative and regulatory tussle. If the non-incumbents win, the incumbents have be paid to move their signals elsewhere. In the US, as in many other countries, the regulator then auctions off blocks of spectrum to operators or releases it for use by the public.”

Chaffer point out the costs of acquiring spectrum: “It's also worth noting that each generation, from 2G to 5G, utilises specific frequency bands. The shutting down of 2G and 3G networks allows for the freeing up of spectrum. Spectrum is physically limited and one of the most expensive resources for a mobile operator. The ongoing cost-of-ownership is usually lower for 4G and 5G networks compared to 2G and 3G, while spectrum efficiency is higher.”

National spectrum regulations

Blythe highlighted the internation corporation that is involved in all of this: “The ITU manages the spectrum frameworks for spectrum band identifications globally. These frameworks are not static as some services that use spectrum are in decline whilst other require more spectrum as demand increases. Every 4 years the ITU organises an international conference to decide on how these frameworks will develop going forwards – this is called the World Radiocommunications Conference, with the last one occurring in 2023 (3000 delegates, 190+ countries attending). At these events, the country members of the ITU decide on any changes to the international spectrum agreements based on proposals from members.

“Any agreed changes can then be incorporated into national spectrum regulations. These changes could for example identify a new spectrum band for a service, one that was reused from a declining service. It would be normal for a national Government to award any new spectrum bands identified by this process through a competitive award, with mobile operators bidding against each other to acquire new spectrum licenses (and also pay). For example, this is the process that has gone on in most countries in Europe and beyond with spectrum that is required for 5G technology where part of the spectrum in question was used previously for the Terrestrial TV service.

“Spectrum licences normally have a finite duration - 15 to 20 years generally. Mobile technologies get updated every 7-10 years and so when a new technology is adopted the usage of the previous technology declines. This decline allows, over a period of time, for the spectrum to be re-used (or refarmed) for the newer technology. This allows for greater capacity for the newer technology as more spectrum bandwidth can be used, and increases the efficient usage of the spectrum. This is a continuous process and most operators are going through a process of turning off legacy 2G/3G networks which will further facilitate this process.  At the end of the licence period, the National Regulator can either extend the licences for further payments or investments conditions, or re-award the spectrum through another process. The former is preferred as the latter is associated with high risks on operators in ensuring service continuity.”

Will 6G need different spectrum?

So that’s the picture of how the telecoms industry uses spectrum in different ways to make the connectivity that underpins the modern world work. But will anything change as we approach 6G, which while this hasn’t been fully defined, is already being talked about as ushering in new capabilities.

Castor says: “The demand for more spectrum with each generation is driven by higher data rates, capacity, and user requirements. –   6G will be no different. Social media applications on our mobile phones, for example, are demanding higher data rates and lower latency to deliver high resolution pictures and videos. 6G technologies are also considering new applications such as immersive media and leveraging spectrum for radar-like sensing for more accurate positioning. Today it is difficult to say how much more spectrum will be needed, but new technology and applications are clearly emerging that will be hungry for as much spectrum as regulators can enable for mobile operators to purchase.”

The need for more spectrum

Grossman says they’ll always be demand for more spectrum: “The mobile telecom industry always needs additional spectrum, regardless of whether it’s called 5G or 6G, because their customers’ usage keeps growing. The US is presently in a big tussle between the Pentagon, the WiFi interests, and the mobile operators over some mid-band spectrum that is now reserved for military uses. The Pentagon’s argument is national defense. The operators’ argument is national technology leadership via 5G Advanced and 6G. Both arguments, whatever their merits, have emotional appeal to the political establishment and regulators. As to 6G, some of the futuristic visions include use cases, like holographic communications and digital twins, that could demand higher speed and lower latency than 5G Advanced can provide.

“In addition, 6G will be able to use extremely high frequency “sub-THz” spectrum, which is above the 90 GHz mmWave band. Sub-THz radios and their components are still in the research phase and will require some big technological advances before commercialization.”

Padma Ravichander, CEO of Tecnotree says: “The development of 6G technology necessitates a nuanced spectrum strategy tailored to its advanced requirements. While specifics are subject to evolution, 6G networks are expected to leverage a mix of existing and new spectrum bands to fulfil demanding use cases. These may include ultra-high-speed data transmission, ultra-low latency communication, and massive IoT connectivity. While mid-band and mmWave frequencies from 5G networks may persist, 6G may also explore higher frequency bands for enhanced data capacity. Innovative spectrum sharing approaches could further optimize spectrum efficiency and utilization, aligning with the dynamic landscape of wireless technology evolution.”

Blythe adds: “When a new mobile technology is introduced initially the previous generation network is normally still heavily used, and it may take a number of years for the utilisation of this older network technology to decline sufficiently that some of the spectrum supporting it can be refarmed (see point 3 above). Also, as mobile broadband data continues to grow, more and more spectrum bandwidth needs to be used to support this growth. These 2 factors mean that new technologies normally require new spectrum bands to be used for initial deployments. These will then be supplemented with refarmed spectrum as and when. It is likely that 6G will be no different, with the newly identified 6GHz band being the most likely candidate for Europe, Africa and the Middle East.”

Spectrum management

Finally, Kynäslahti said on the process of spectrum management in the years to come: “The traffic of the mobile networks will increase ca. 20-30% per year. We need new capacity to cater for the traffic growth. The new capacity is offered with the new frequencies. With this traffic growth average mobile network will run out of capacity at the end of decade. The introduction with 6G, deployments starting in 2029, will be very timely to introduce further capacity for the mobile networks.

“In December, WRC-23 identified new candidate spectrum to study for 6G between 4.4 and 15.4GHz. Specifically, the WRC identified the 7.1-8.4 GHz band as a future global candidate for 6G services. The propagation characteristics of 7.1-8.4 GHz would allow operators to build 6G on top of their existing infrastructure, and there will be plenty of capacity for multiple service providers in every country. Most significantly, many countries would be able to deploy 6G over the full bandwidth of these frequencies, realizing massive economies of scale and ensuring widespread interoperability.

“Eventually, refarming existing frequency bands to 6G will happen as well but this will take time as refarming frequencies from previous to new mobile generation requires to modernize the device population.”

For a wider view of 6G and what the industry expects from it, check out our deep dive what is 6G?