Why the Operators Need to Choose a Specific Sub GHz Band for 5G?
Carrier Aggregation works best when the bands are separated out wide in the frequency domain, else it will not only increase the cost of the device but also decrease coverage.
We all know that for the operators to deploy 5G services effectively they need a large quantum of spectrum, especially in the low spectrum bands. The reason - the low-frequency bands have the capability of penetrating indoors - which is absolutely essential for ensuring ubiquitous 5G coverage. Also, without the low-spectrum bands, standalone 5G can’t be deployed - preventing the operators from leveraging the advanced features of 5G. But the problem in India is that these low-frequency bands do not have sufficient spectrum, and whatever we have, a large chunk of it is already assigned to PSU, Railways, and Defence, or occupied by legacy 4G and 2G services.
In India, 67% of the 700 MHz band is allocated to Defence, Railways, and BSNL. Therefore only 33% (15 MHz) is available for the Private Operators to use of which RJIO is already using 10 MHz. Fortunately, the 600 MHz is empty, but unfortunately, it does not have an ecosystem of handsets yet, and might take a long time before it gets developed. The other bands such as 800 and 900 MHz are heavily used for delivering legacy 2G and 4G services, and can’t be freed soon, as these services are likely to remain alive for a reasonably long period of time. Also whatever little is available in these legacy bands are in small slices (less than 5 MHz).
Given this situation, how do we find spectrum in low-frequency bands for deploying 5G services in India? One option could be to combine spectrum from different low-spectrum bands (like 700 MHz mixed with 800 MHz or 900 MHz etc.) with the intent to create larger virtual blocks through a process called “carrier aggregation”.
The purpose of this note is to describe the challenges that we will face in aggregating airwaves of bands that are close to each other in the frequency domain. But before I start analyzing this fact, let’s try to understand why 5G needs a large chunk of spectrum in the Sub GHz band.
The Wedding Cake Model
A typical 5G network looks like something as shown in the picture below.
Note, that both mid and high bands have a huge quantum of airwaves - 3500 MHz has 100 MHz or more, and 26 GHz has 800 MHz or more. Both of these bands are capable of outdoor coverage only. Hence for enabling coverage inside the buildings, lifts, basement, etc., we are largely dependent on the Sub GHz bands (i.e. Bands < 1 GHz). Hence, if the width of this band is less, let’s say ONLY 10 MHz or below, then a 5G consumer’s experience will be similar to a racing car driver who is forced to enter a bicycle track (Low Band) suddenly while experiencing 250 Km/hour speed on the main track (Mid/High Bands).
Which is why, the thickness of the “Low Band” has to be a minimum of 20 MHz, but having a chuck of 25-30 MHz is nothing short of “great”. But unfortunately in India, only one operator has just 10 MHz, and the other three have nothing (the 10 MHz that BSNL has in the 700 MHz is being used for deploying 4G). However, the problem is not limited to this alone, it is even more serious, as there are no roadmap for the operators to get more in the future.
Given this situation, can we not solve this problem through “carrier aggregation”? I.e. join 10 MHz of 700 MHz band with 10 MHz of 800 MHz, and create a combined block of 20 MHz?
Unfortunately, this is NOT viable for bands that are spaced close in the frequency domain due to significant commercial and technical challenges. Before we go into discussing these challenges, let us understand how this process (called carrier aggregation) actually works.
How Carrier Aggregation Works
The following is a simplified diagram that explains how carried aggregation works in a typical handset
Typically a handset is capable of simultaneously receiving signals from multiple spectrum bands, but processed one at a time. However, those supporting carrier aggregation, are capable of combining more, i.e. signals emanating from two different bands. Some are even more. This results in the following advantages.
Firstly, the device can now support higher data speeds which is proportional to the total integrated bandwidth the device sees at any point in time. We all know that speeds at the edge of the network are significantly less compared to what we see in the middle or near the BTS. The 800 MHz band, with better link budget capability, will compensate for the decreased speed of the 3500 MHz band, even when its (3500 MHz band’s) bandwidth is 10 times more than that of 800 MHz. Now if this bandwidth gets fragmented, so will the data speeds, as in handsets without “carrier aggregation” capability, the signals from ONLY one band will get processed at a time.
Secondly, the coverage of the network for the handsets with “carrier aggregation” capability gets stretched much further than the one without it. The reason - handsets can choose to transmit back to the BTS using the band with a better link budget (in this case 800 MHz), while they continue to receive signals from the higher bands (3500 MHz) with poor link budget capabilities (BTS transmit at much higher power compared to handsets).
Thirdly, the coverage holes in the network can be plugged. This improves the quality of the network significantly as seen by the handset with “carrier aggregation” capabilities.
Carrier Aggregation For Tightly Spaced Bands
Normally carriers are aggregated for bands that are spaced apart (like 700 & 3500 MHz) and NOT that are close (like 700 & 800 MHz). The key reason is to prevent the “diplexures” design from becoming complex. These diplexers are nothing but band pass filters, installed just after the antenna, with the purpose of extracting the RF signals emanating out of the two bands whose RF signals are required to be combined.
Now if these two bands are close in the frequency domain (not spaced apart), then the diplexers need to have sharp cut-off capabilities - driving up their cost. Not only that, it will also induce an additional loss of 1 to 2 db - pulling down the overall coverage of the Subs GHz network.
Fortunately, for the bands that are spaced apart (like 700 & 3500 MHz), neither of the above problems are experienced, as the diplexers need not be of high cut-off capabilities, and therefore the loss induced in the signal path is also significantly less.
The following diagram explains this pictorially.
As can be seen from the above figure, if the bands are not spaced far apart then there is a significant risk of the harmonics of one band entering the other’s receiver thereby contaminating the reception. If the bands are spaced apart, then they do not face this risk, as by the time the harmonics reach them, they can’t cause any damage due to their significantly diminished amplitude.
This is the reason why an operator has to decide which of the bands in the Sub GHz spectrum it wants to use for the purpose of providing 5G services. They can’t mix and match, as it will serve no useful purpose given the fact that aggregating carries across bands that are NOT spaced sufficiently apart (like 700, 800 & 900 MHz) will make the device significantly costly and drive the coverage of the network down, thereby significantly diluting the key advantage of the Subs GHz bands.
Conclusion
India is on an advanced path of deploying 5G networks. Very soon we will have a Pan-Country 5G coverage. However, these networks can’t be fully leveraged unless the operators supplement them with adequate deployments (at least 20 MHz) in a specific low spectrum band (< 1 GHz).
We all know spectrum in the Sub-GHz bands are scarce and suboptimally used. Hence, it is extremely important for the GOI to create a roadmap that ensures its optimal usage so that all the operators can get access to sufficient 5G spectrum in these bands. Not doing so will make India’s 5G deployment in India incomplete - both in quality and capability. Which is neither in the interest of the Country nor its Consumers.