In India, we are all excited. 5G is knocking on the doors. It will bring new opportunities and unlock new possibilities, some of which, I am sure will change our lives forever. But are we ready? If yes, that is great news. But if not, what are the bottlenecks? What should we do to clear them? This note not only lays out these issues but tries to find optimal solutions for overcoming these. Let’s begin.
Spectrum Availability
Are we good here? My read is yes. We did a great job in releasing a sufficient spectrum for 5G. Band-wise quantum :- 2x30 MHz in 600 MHz Band, 2x25 MHz Band in 700 MHz, 330 MHz in 3.5 GHz Band, & 2850 MHz in 26 GHz Band. With 4 operators on the ground, each can get 800 MHz. Assuming is fair and equal distribution. See Figure 1 below which describes the proportion of 5G spectrum offered in the auction compared to that of 4G.
Spectrum Price
In my view, 5G spectrum prices are significantly high, especially for the lower spectrum band (below 26 GHz). The reason:- The principle used to price the 5G spectrum bands has been extracted from that of 4G. This is fundamentally incorrect. Why? Per MHz price of 4G should NOT be directly used to extract the price of 5G bands, as 5G needs significantly more spectrum in all bands (explained here — wedding cake model) than its 4G counterpart to drive higher data speeds and quality of service. And such higher data speeds of 5G will NOT result in a proportionate increase in revenues for the operators.
Solution — Ideal solution is to delink the price of 5G bands from the 1800 MHz (4G band) quite distinct from what the TRAI has done now.
In order to value 5G bands, the TRAI uses the prices of the other 4G bands as references. In other words, the prices of 5G bands are not derived independently — ground up. This is also true for all existing 4G bands as well, except for the 1800 MHz band, whose pricing is done ground up using various valuation models. The prices of other bands get extracted from the value of the 1800 MHz and other bands with similar propagation characteristics.
Since delinking the price of the 5G spectrum from 4G will take time and will delay the auction. Hence, an interim quick-fix solution can be to reduce the license period of the 5G spectrum from the current 20 years to 10 years. This will help reduce the price of the 5G spectrum by 50%. Such an approach will also be in alignment with the international practice (South Korea, Finland & Australia) where the 5G spectrum is being given for a much shorter duration than the conventional 4G spectrum. This approach also has other benefits too, as due to rapidly changing technology, creating longer rights on the spectrum might be detrimental to the interest of both the operators and regulators.
Backhaul Spectrum
5G and Backhaul Spectrum go hand in hand. Why? Operators need a thicker pipe to aggregate and carry huge traffic emanating from 5G access (20x that of 4G). Hence, the current Microwave based backhaul spectrum (carrier size of 2x28 MHz each) will choke and fail. Why? Even if you get 10 such carriers (2x280 MHz), it will be too thin to carry traffic emanating from the 5G access pipe that will be 3 times thicker (2x800 MHz). Unfortunately, OFC connectivity to BTS sites is still very low (less than 30% of average). Therefore, the operators will need fatter means to carry 5G traffic. Fortunately, E Band (71–76 GHz and 81–86 GHz) has 7 GHz of spectrum. So the 4 operators on the ground can each be given 1.75 GHz each — sufficient to carry 5G access traffic to switching stations. But the DoT is yet to announce a policy for the assignment of the E Band spectrum.
Solution — a) E Band spectrum should be given not link by link (as TRAI had suggested), but exclusively on an LSA basis like the existing microwave access spectrum.
TRAI’s approach of charging the E band spectrum based on a link-by-link basis will create an accounting nightmare for DoT (audit and reconcile license fees) due to a large number of links that the operator will deploy, compared to what they have already done for the microwave spectrum. And on top of this, DoT will have to deal with the multiple disputes emanating from the operator’s end on account of the responsibility of interference management. This the TRAI has left to the operator to manage on its own and which the regulator simply cannot wash its hands away.
This means that the operator should be free to use this spectrum for both access and backhaul. DoT can decide to auction this spectrum or give it administratively. If DoT decides on the former, it should immediately send a reference to TRAI for pricing the E Band spectrum and decide on the rules of the auction.
b) Satellite Spectrum can be a good option for backhauling 5G traffic, especially for connection remote areas. DoT should immediately send a reference to TRAI to decide on the modalities of assignment (auction or administratively) of the satellite spectrum.
In the interim 4 carriers, each of 2x250 MHz in E Band should be assigned immediately to the winners of the 5G auction to be regularized later once DoT is ready with the policy. This helps the operators in harnessing the full capability of their 5G spectrum and will be in the interest of the consumers.
Spectrum Usage Charge
Fortunately, DoT has through its press note dated 15th Sept 2021 reduced the SUC of all spectrum auctioned in the future to 0%. In order to get the full benefit of this policy, the DoT has also removed the current floor of 3% in the WA (weighted average) calculation formula — used for arriving at an aggregated value of SUC by including all past auctioned spectrum (which was given at SUC of different rates). But the SUC rates of backhaul spectrum for satellite and E band are yet to be decided. Again the TRAI's approach of SUC for both E-Band and Satellite spectrum and inconsistent and faulty. TRAI has recommended a link-by-link approach of charging SUC for E-Band, and 1% of revenues emanate only from carrying traffic in the satellite spectrum which is to be accounted for separately.
TRAI’s approach of charging the E band spectrum based on a link-by-link basis will create an accounting nightmare for DoT (audit and reconcile license fees) due to a large number of links that the operator will deploy, compared to what they have already done for the microwave spectrum. And for that of satellite spectrum will create a possibility of arbitrage, and innovating accounting by operators to save license fees.
Solution — The solution actually very simple to implement. a) The SUC for all spectrum acquired through auctions should be in line with the cabinet press note of 15th Sept 2021, i.e at a 0% rate, and the combined aggregated rate can be derived using the WA formula. b) SUC for all spectrum acquired administratively should be @ 1% (at par with BSNL’s satellite service), but there should be a suitable cap (let’s say 3%) so that the SUC rate should not cascade and become a burden on the operators. This will also motivate the operators to acquire spectrum through auctions than to take it administratively.
Low-Frequency Bands
5G is not just about high-frequency bands (3.5 GHz and 26 GHz). It is about low-frequency bands too (600/700 MHz). Actually, low-frequency bands play an important role in enabling the operators to deploy standalone 5G services. But why do we need standalone 5G? A) To get rid of the 4G core, as if 5G traffic pass through 4G core then many of the 5G features like network slicing etc can’t be implemented (these are important for enterprise offerings in a private network setting); B) To migrate voice from 4G to 5G, as the current 4G voice suffers from poor quality and migrating it to 5G can improve quality significantly; C) To get 5G networks to penetrate inside the walls of buildings and closed enclosures, as higher frequency spectrum, 3.5 GHz & 28 GHz will get blocked by concrete walls, quite unlike the low-frequency bands, which bends across walls to reach out deep inside buildings.
But unlike 4G, 5G will need a large block of spectrum in the low-frequency bands to ensure consistency of 5G services throughout the network. If the low spectrum bands are thin then the consumer’s experience will suffer drastically at the network edge (inside buildings) where the signals from 3.5 GHz and 26 GHz will not be able to reach. The situation will be similar to driving on a 16-lane expressway, and then being suddenly forced to exit into a dirty and bumpy village track.
The problem is that though we have plenty of spectrum in 600/700 MHz bands for the operators to take, it is beyond their reach as these have been priced using the 4G’s per MHz rate as a reference. A 2x20 MHz spectrum block in 600/700 MHz will cost the operator a staggering Rs 78,540 Cr at the current RP.
Solution — A) Revalue the spectrum using a different and more realistic approach. B) Or in the interim reduce the license period from the current 20 years to 10 years. The latter will decrease the operator's outflow by 50%.
Leveraging 5G Services
India must be able to leverage 5G for the purpose of nation-building. As there is no point in pushing for 5G if it is only cosmetic and doesn’t do much immediately helping to remove the bottlenecks that we face today in deploying basic services like fixed-line Broadband — the demand for which has increased significantly after the Pandemic. Fortunately, 5G has the capability to emulate wireline-like capabilities overall wireless.
To a layman, 5G empowers the operator to serve their customers with independent dedicated links (as in the case of wireline), the speed of which (links) is not impacted as other users start to access the same BTS (simultaneously) in which the customer is hooked. (For a more elaborate explanation, please read my note — How is 5G different from 4G?). In other words, 5G technology empowers the operators to offer broadband connections over wireless (faster to deploy) at a dedicated speed, just like they offer it over a wireline connection.
The problem is, like any other technology, 5G will suffer from teething troubles like high equipment prices which might hamper its take-off, till it is able to reach a scale sufficiently large to amortize costs overall large volumes (economies of scale).
Solution — It is understood, that in order to expand broadband connectivity in the country, the DoT is thinking to propose the removal of the license fee for fixed-line broadband services for a period of 10 years. In my view, a similar incentive scheme for rolling 5G FWPs should also be announced. The reason is simple, wireless broadband can be rolled out much faster than wireline. And as the FWP connections increase, the same will drive economies of scale and make “Ultra Fast” broadband even more affordable.
Net Neutrality Rules
The operator’s ability to leverage 5G services will clearly depend upon clarity on TRAI’s NN rules. TRAI on Nov 28th, 2017 issued the following recommendation.
In this order, TRAI defined some services as “Specialised Services” which are explicitly excluded from the principle of non-discrimination, provided the provision of such services does not degrade the overall quality of Internet Access Services. The logic is that the character of these specialized services (such as IPTV, VOIP, telesurgery, etc) can be preserved only if these are prioritized over the internet. The list of such specialized services will be finally decided by the licensor. Hence, if “video on demand” & “Linear TV” is included in the list of specialized services then an operator can offer its consumer an increased data speed (over normal) while watching Netflix or Linear TV Channels over 5G based BB — enabling an operator to monetize his investment without impacting other users accessing the internet.
Solution — DoT should proactively notify a list of 5G services that can be included in the list of “specialized services”. These should include Linear TV, Video on Demand, etc. Since 5G emulates wireline kind of capability over wireless doing so will not pose any problems, as such services are already permitted over wireline without violating NN rules.
Conclusion
It is clear from the above that we will not be able to unlock the full potential of 5G unless we clear the bottlenecks listed above. Mere auctioning of access spectrum for 5G isn’t sufficient. More needs to be done on the policy front so that once the spectrum once assigned it can be fully harnessed by the operators for delivering services (high-quality voice, ubiquitous inbuilding coverage, broadband, Linear TV, Video on demand, private networks, etc). One can clearly see from the above discussion that finding solutions to most of these problems is not that difficult — actions can run in parallel with the auction process for the access spectrum.
As per TRAI (page 5, Rec Dated 31st Aug 2021), Fixed Broadband connection in India is among the lowest in the world at only 1.69 per hundred inhabitants. This looks pretty bad if we compare it with markets like France (44.5), Korea (42.8), Germany (42.7), Canada (41.2), the UK (40.5), the USA (36), Australia (35.2), and Japan (33). 5G FWP provides an excellent opportunity for us to expand and enhance this number. Pandemic has enhanced the demand for fixed broadband connections significantly. We simply can’t rely on wireline-based broadband to meet this enhanced demand (See Figure-5). 5G not only has the capability to emulate fixed wireline over wireless but it can be scaled much faster, thereby driving costs down. This will enable affordability, thereby helping with the objectives of NDCP-2018 of providing high-speed broadband connectively to 50% of India’s households. And once that is achieved, then we can start to realize our dream of digital India, which will not only elevate India’s image in the league of nations but also change the lives of its citizens.
(Views expressed are of my own and do not reflect that of my employer)
PS: Find the list of other relevant articles in the embedded link.
