How Technology and Regulation Impacts Business Strategies

All businesses are influenced by technology but even more by the regulations that govern its deployment. These technologies not only enable new products/services but also help improve the operational efficiency of existing businesses - making it more profitable. In current times, no sector has been left untouched by the "technology-led innovation", and this is especially true for the telecom sector which has to deal with "bottleneck resources". The objective of this note is to investigate the impact of technology and its deployment rules on the operator's business. In the process, we will also analyze and the strategies adopted by various technologies to deal with "bottleneck resources" (spectrum etc), and how it influences the operator's ability to offer services.

Wireline

In the beginning, calls were made using copper wires. These lines remained dedicated to the user and not shared with anyone else. The calls emanating from such subscribers were aggregated in a "switch" used for connecting different users. These switches were sufficiently equipped to handle all calls made during the busy hour. A wireline system is not limited by the size of the switch but by the high cost of laying cables - preventing faster roll out and a huge wait time for connections (sometimes many years). This issue got resolved with the advent of wireless technologies. It used a shared radio spectrum to connect different users. Hence, the ability to support a large number of calls depend on the capability of the technology to enable smooth coordination between various users which get significantly influenced by the "deployment rules". This is what I plan to discuss going forward in this note.

GSM

GSM (Global System for Mobile communication) was first launched in Finland in 1991. Before the advent of 3G, 4G it was the most popular mobile technology throughout the globe. In India, it was introduced in the year 1995 when the first cellular licenses were issued. This technology works fine with fragmented spectrum and is capable of handling blocks as small as 0.2 MHz (200 KHz), which is further subdivided into eight (8) time slots (periodic "slices of time" perceived as a continuous connection by the user). A time slot is what connects the users to the switch just like the copper wires do in a wireline network. One can get more of it by aggregating more spectrum, but all can't be used simultaneously, as the radio waves from the adjacent base stations interfere with each other. To prevent this the adjacent base stations always run of different time slots, which are then reused in base stations sufficiently spaced out from each other. This has three implications, first, a base station is able to use only a fraction of the total available spectrum for carrying traffic, second, any spike in traffic in any of the base stations can cause congestion, third, the network is best suited for carrying voice and not data (optimal data rate needs aggregation of multiple time slots, making the overall system very inefficient). This is the reason why a GSM operator cannot offer broadband and has difficulty in offering "unlimited voice". But, for the regulator to manage spectrum is very easy, as he doesn't have to offer it in contiguous blocks. Hence, the 900 and 1800 MHz band (used initially for deploying GSM) was so fragmented, which later had to be harmonized for supporting the newer technologies.

CDMA

CDMA, on the other hand, adopted a totally different approach compared to that of GSM. It was first deployed for mobility in India in early 2000. It needed a min block of 1.25 MHz of spectrum to work. Unlike GSM (which sliced users in small time slots), the users here are assigned the full block of 1.25 MHz simultaneously, i.e they overlap on top of each other. So the receiver sees nothing but "noise" - a signal full of interfering users. But, since these users are marked individually by orthogonal codes the receiver can separate them out from the background of noise. The receiver's ability to do so is limited by the intensity of the background noise. Hence, controlling the background noise is critical to ensure proper functioning of the system. The system does so by reducing the power of the transmitting users located close to the base station and increasing it for those who are far away. Also, the base stations are sufficiently spaced out (compared to GSM) to keep the background noise under control. That is why the total number of base station deployed in CDMA is much less. Therefore, a CDMA operator is able to offer voice at a much cheaper rate compared to a GSM operator due to lower opex and capex needed. This we saw in India when the calls rates got drastically reduced after its (CDMA's) introduction in the year 2003.

EVDO

Compared to GSM, CDMA also handles data more efficiently. For that, it uses a variant called EVDO (Evolution-Data Optimized) which is only used for carrying data. Unlike CDMA, it adopted a totally different coding technique in which the full 1.25 MHz could be made available to a single user for carrying data. This increased the peak data rate significantly. In the case of competition from other users, they are all statistically multiplexed. Then, why did the Indian operators did not offer EVDO services initially? This could have given them a huge competitive edge. Unfortunately, the spectrum assignment rules were responsible. The CDMA spectrum was assigned to operators in incremental blocks of 1.25 MHz, that too when a certain threshold of voice subscribers was breached (as defined by the regulator). This was done to enable a level playing with GSM operators - who were offered double the spectrum (compared to CDMA operators) as the GSM technology was inefficient - preventing a CDMA operator from rolling out EVDO services, as EVDO needed one full block of 1.25 MHz which would have restrained the CDMA operator's ability to offer voice. The CDMA operator's inability to unlock its potential, that too, in the beginning, was one of the key reasons for the downfall of this technology in India. Though GSM had its advantages in terms of cheaper SIM based devices enabled by NRE (non-recurring expense) amortized over a much larger volume compared to that of CDMA.

3G

3G was introduced in India in the year 2010. Unlike in the case of GSM and CDMA, the spectrum for 3G was auctioned. The auctions prices went very high primarily due to "demand-supply mismatch" (3 slots of 5 Mhz each were chased by 5 serious players) and difficult "auction closing rules". See my earlier note for a better understanding - "The Story of Spectrum Auctions". The technology was similar to that of CDMA, only differing in the carrier size. The high price of spectrum prevented the operators in rolling it out in remote parts of the country. Also, too much focus on voice (driven by subscriber based spectrum assignment rules for 2G) prevented the operators in developing the data market, thereby further restricting its take off. The coverage in urban areas was patchy, as the spectrum offered was in 2100 MHz band, which being in higher frequency, was not able to penetrate walls - resulting in poor in-building coverage. 3G is very efficient for aggregating spectrum in blocks of 5/10 MHz. The complexity of the receiver increases significantly when the block size becomes bigger (required for higher speeds), forcing one to use a totally different approach. This laid the foundation of LTE (Long Term Evolution).

4G

4G in India saw a tussle between LTE and WiMAX. WiMAX claimed that they offer much better in speed and reach compared to 3G, and therefore more suitable for rural broadband. This prompted the regulator to set the reserve price of BWA spectrum at 1/4th the rate of 3G. The argument was strange, as compared to 3G, the BWA spectrum has poor propagation characteristics, it being at a higher frequency. Even after the price concessions, WiMAX never got deployed in India for mobility. The reason became very clear after the 2010 BWA auction. At that time WiMAX was deployed globally in the 2.5 GHz band. The WiMAX lobby wanted regulators to open this band in India as well so that they can amortize costs over larger volumes. But, unfortunately, 2.5 GHz band was blocked by DoS (Department of Space) and only 2.3 GHz band got auctioned. This demotivated the WiMAX lobby and they pulled out, thereby paving the path for LTE deployment in India. The urge for deploying LTE was so strong that most operators kept their spectrum vacant for a long period of time for the LTE ecosystem to mature.

For coding, LTE uses a similar approach to that of GSM. Here also the block of spectrum is divided into smaller chunks (180 kHz each). This is further subdivided into time slots. But, these slots are much tightly coupled with each other using orthogonal codes - making them much more efficient. Just like 3G, LTE needs spectrum in continuous blocks - the higher the block size the better. Hence, the existing spectrum bands of 800, 900 and 1800 MHz band needed to be harmonized to enable larger blocks of contiguous spectrum. Unlike other technologies, LTE is optimized for data. Hence, the voice here has to be transported in the same packets which are designed to carry data. This creates some difficulties, as the bit rate of normal voice has to be increased many times - resulting in lower coverage for voice compared to the conventional systems. That is why voice over LTE (VoLTE) is best suited for lower frequency spectrum (700/800/900 MHz bands). Also, the high spectral efficiency of the LTE system makes it best for offering "unlimited voice". See my earlier note - "How "Unlimited Voice" Impact Indian Operators?"

5G

There is a limit to the extent to which the efficiency of a technology can be increased using conventional methods. Thus, the only way to get more speed is by aggregating larger blocks of spectrum. But, this is possible only at higher frequency band. However, these bands have very poor propagation characteristics, and therefore one will need a totally different strategy to leverage them. Hence, 5G uses "beam steering" techniques to overcome the inability of the higher frequency bands in driving radio waves to a larger distance. To enable "beam steering" we need multiple antennas to get deployed on the handsets. The number of these antennas can run into hundreds. Fortunately, the size of the antenna is directly proportional to the wavelength of the radio signal - which is very less at these higher frequencies (frequency of the radio signal is inversely proportional to wavelength). This is why the 5G technologies cannot be deployed efficiently in the lower frequency bands - as it is difficult to accommodate so many antennas in the mobile handsets at the lower frequencies without making it unwieldy. In India, all the higher frequency bands being discussed for 5G are also candidate bands for backhaul spectrum. Hence, one needs to manage them properly to preserve the flexibility of using them for 5G at a later date. The following is a table which lists all the bands at the higher frequencies and their status in India and internationally.

WiFi

WiFi is deployed in the unlicensed bands. Unlicensed spectrum does not guarantee exclusivity and therefore, there is no coordination between operators/users using this spectrum. Hence, other than managing interference from adjacent access points (the equivalent of towers), managing interference from users of other operator's network also becomes paramount. Physical isolation is created - a) by identifying unlicensed spectrum blocks at higher frequencies than lower, as at these frequencies the RF signals tend to attenuate faster compared to those at lower frequencies, b) by mandating the operators/users to transmit at much lower power compared to their licensed counterparts (4 watts ERP vs 20 watts ERP). This enables multiple numbers of networks to coexists in the same frequency band very close to each other. However, uncoordinated use makes it impossible to eliminate the possibility of interference and hence, the technologies using unlicensed spectrum uses frequency hopping techniques (similar to GSM) and dynamically choose a block of spectrum (among multiple) with least interference. Thus in order to enable reasonable data rates, an unlicensed operation need a much larger quantum of spectrum compared to its licensed counterpart (typically a block of 100/200 MHz) of which only a fraction (20/40 MHz) is used at a time. This is the reason why lower frequencies (less than 1 GHz) are not good for unlicensed operations as it is difficult to find such large blocks of spectrum at these frequencies. Also, RF signals (even at lower power levels) in the sub-GHz frequency tend to travel far causing severe interference to uncoordinated networks operating nearby. This also explains why licensed bands are priced at much higher level and unlicensed bands are free. In India 2.4 and 5 GHz have been unlicensed to be used for deploying WiFi technologies.

Whitespace

Whitespace technologies are designed to use small slices of fragmented spectrum lying between signals transmitting at very high power with an aim to reach large distances. This strategy of transmission is used in broadcasting where the receiving equipment does not have to transmit back to the base station which limits the coverage of the conventional mobile systems. Whitespace systems are designed to work at low power and with filters with sharp cut-off to prevent interference into the broadcast bands. These special requirements make the "whitespace" system costly compared to conventional mobile systems. In this regard, I will like to point to a recent study done by IIT Mumbai. The study points out that more than 80% of the TV UHF band IV (470-585 MHz) is totally free and lying unused (see Table III of page 5 of the PDF in the link above). If we go with this study then there are no real "white spaces", but only free and unused spectrum that exists in India. In other words, the TV UHF band IV is as free as any other newly auctioned spectrum band (100% free and unused spectrum does not fall under the definition of "whitespace"). For more details, please read my earlier note - "TV Whitespaces: How white are these spaces?" Recently, India supported the TV spectrum to be used by IMT technologies in the WRC held last year. As IMT alone can ensure efficient usage of this band. The only issue is to manage the price this spectrum band optimally so that services offered in this band are affordable.

From the above, it is clear that every technology has been designed to solve a specific business problem and if used differently can lead to inefficiencies. Also, the regulators dealing with them can inadvertently pick winners and can be detrimental to the overall health of the market. As long as we acknowledge this we can use technology as a useful tool to solving complex business problems, and we all should profit from it.

 (Views expressed are mine and do not reflect that of my employer)