Your browser is not supported. Please update it.

27 August 2019

Indian MNOs need a whole new infrastructure model to escape a 5G crisis

If any country needs a new set of economics before the operators deploy 5G, it’s India. The MNOs have struggled since the 2G era with excessive bureaucracy, inflated spectrum prices, very low ARPUs, challenging terrain and market over-crowding, not to mention high pressure from government to achieve lofty but often impractical national broadband goals.

Some of these challenges have been reduced in recent years, mainly by consolidation of the market down to three major MNOs – Vodafone Idea, Reliance Jio and Bharti Airtel – plus the two state-owned companies, BSNL and MTNL. This has given the privately run survivors greater economies of scale, lower competition and greater nationwide reach. But the other factors remain unchanged, including proposals by regulator TRAI to set very high reserve prices for midband and sub-GHz 5G spectrum; and very low ARPUs combined with a strong consumer and government drive for better network quality and higher speeds.

To make matters far worse, new entrant Reliance Jio has leveraged its greenfield network, high degree of asset sharing, and its parent’s deep pockets to be very disruptive in terms of pricing and network architecture. It has a very automated network based on commoditized elements with a high level of virtualization (though not full cloud-native core or vRAN yet). It has no legacy network. All these factors enable it to follow a capex- and opex-light model, which in turn has allowed it to price its entry level services extremely low, sparking a price war, further M&A among the majors, and a very successful grab for market share.

Now, Jio’s challenge is to implement its strategy to become a digital services provider to consumers and enterprise, not just a connectivity firm in a market where it has helped to make connectivity a devalued and commoditized offering. It has invested in fiber to support both markets and in video content and the development of a suite of applications for the home base.

Its latest move is to start to build an NB-IoT network over its LTE footprint to support machine-to-machine services for businesses, and to enhance its own smart home offerings. It says it is targeting one billion NB-IoT connections within two years once the service is launched commercially in January. NB-IoT is itself a capex-light option. Since Jio’s LTE network is so new, the MNO will be able to deploy the M2M network purely as a software upgrade, and the technology greatly increases the range of an LTE cell, reducing the number of new sites that would be needed to achieve full IoT coverage.

Mukesh Ambani, chairman of Jio and its parent Reliance Industries, said at the recent annual general meeting that NB-IoT was an “INR20,000 crore per year revenue opportunity for Jio (INR200bn or $2.8bn). He added: “Within another two years, it is estimated that there will be more than 2bn connected IoT devices in our country. Jio expects to connect at least 1bn of these on its IoT platform.”

So Jio aims to remain a low cost operator in terms of costs, with plans for a fully cloudified 5G network, but to look beyond low cost models in terms of service pricing, adding enterprise and multiplay services to its portfolio, while topping up the user numbers by harnessing its network efficiency to support ultra-low entry level tariffs.

As 5G auctions loom, the other MNOs need to adopt a new level of sharing and network efficiency to stand a chance of halting Jio’s assault, which has already seen it overtake Airtel to become India’s second MNO by subscribers.

There are several barriers in the way of this strategy. One is the high level of regulation, which restricts freedom to share active networks and even to roam, limiting each operator’s coverage if they cannot acquire spectrum in every operating circle – which may be a massive investment if TRAI sticks to its guns on 5G pricing.

These conditions – and the operators’ low ARPU and profitability, married with high costs and debt – are making vendors wary of the Indian market, despite the opportunity in terms of its scale and the likely size of its 5G contracts. For some years, Nokia and Ericsson have been trying to extricate themselves from loss-making managed services deals and stop bidding for loss-leading equipment contracts, and a fair proportion of these deals have been in India.

Nokia’s latest run-in with an Indian customer is with BSNL, which has reportedly fallen behind on payments for products and services. These are payments associated with 2G and 3G, which does not bode well for the state telco’s ability to sign good contracts for 5G equipment and so fulfil its government’s mandate to make rapid progress towards next generation wireless broadband.

A report in the Economic Times of India quoted sources close to Nokia who said the Finnish firm was considering withdrawing services to BSNL.

“There is tremendous pressure on us. We have been asked to shut down services we offer to BSNL by the top management executives in the headquarters due to continued non-payment of our outstanding amount,” the source said, claiming the outstanding payments total about $112.9m (INR8 billion).

India is often the focus of vendors’ exasperation over very low-priced contracts and slow payments. Earlier this year Ericsson demanded that Anil Ambani, chairman of former MNO Reliance Communications, should be jailed for the company’s unpaid bills totalling $78m. The arrears were paid, after various court actions, and Ericsson’s refusal to give RCom more time, by Mukesh Ambani, Anil’s brother and chairman of Reliance Industries and Jio.

BSNL is in a particularly precarious position, with none of the deep pockets of Ambani or even the funding options available to Vodafone and Bharti Airtel, which are both part of international groups which have engaged in consolidation and private financing to improve their financial straits.

Successive Indian governments have toyed with proposals to breathe new life into BSNL and its fellow state-owned telco, MTNL. Plans from break-ups to public offerings to mergers have been mooted but none has been implemented and the companies continue to lag behind their private competitors.

The latest plan was a recently proposed package worth about $10bn, which would aim to reduce staff and costs and enable the operators to expand their mobile broadband coverage and capacity. The proposals include a voluntary retirement scheme (VRS) for employees, a reduced retirement age (from 60 to 58), and the allocation of additional 4G spectrum, worth INR200bn ($2.91bn), to BSNL plus up to INR130bn ($1.89bn) in financial aid to support 4G expansion.

That would give the two operators an improved network on which to provide a viable alternative to the big three MNOs. Currently, the capacity and quality of service on their networks are poor, which is driving customers to the private networks, and they have not been allocated 4G spectrum (except for some legacy broadband wireless airwaves, previously used for WiMAX).

However, it is questionable whether the latest proposals would be enough to save the two operators (MTNL has licences in the two largest cities, Delhi and Mumbai, while BSNL operates everywhere else). BSNL recorded a loss of INR138bn ($2bn) in its last fiscal year, while MTNL’s loss was about INR34bn ($495m). The larger operator has been in local news recently for delaying salary payments as well as payments to suppliers.

But while the situation for these two companies is extreme, all the operators are struggling with the prospect of affording 5G build-out and spectrum prices, amid such price competitiveness. This situation will only get worse for the Indian MNOs if they are not able to buy from Huawei and ZTE. The Chinese firms, with their rich reserves of financing backing from state-owned banks, have been willing to bid for Indian contracts at competitive prices and embed themselves into major networks via their services organization.

And ZTE, in particular, has said it will make India a primary focus following sanctions and bans in the USA and Australia, which are also threatened elsewhere. This focus is sure to come with some attractive deals for MNOs – if they are allowed to purchase from the firm.

Xiao Ming, ZTE’s president of global sales, said recently: “India is one of our key markets. It is the second largest telecoms market in the world [after China] and ranks number one in the world in terms of data consumption. We are the main vendor for BNSL and Airtel and we also work very closely with Reliance Jio, although not yet in their radio access network. We already have a good position in India and increased our market share following the merger between Vodafone and Idea Cellular. That is a hugely positive sign for ZTE and we are hiring a lot of staff to support that growth.”

Little wonder, then, that the Chinese government has reacted strongly to reports that India might block Huawei, and possibly ZTE too, from its 5G networks, on the grounds of national security fears. The Indian government needs to strike a delicate balance between good relations with two important trade and political partners, the USA and China; but this is not just about the current geopolitical climate – India has had national security worries of its own about Chinese suppliers for years. At the start of the 3G and 4G roll-outs, it tried to restrict use of Chinese kit, but was forced by operator backlash to rely on certain rules such as the compulsory placement of source code in escrow.

Now, the private operators are lobbying hard against any restrictions on their choice of supplier, but comments from BSNL, ZTE’s flagship customer in India, reflect the pressure on the state-owned telcos. Chairman  Anupam Shrivastava recently said about Chinese suppliers: “When it will come to rolling out 5G network, a cautious call will be taken. Should there be a risk in technology, certainly, we won’t go ahead, and as long as it is indigenous and has no hazard in putting it up, we can advance.”

The Chinese foreign ministry last week warned of “reverse sanctions” against Indian firms in China if Huawei is blocked – it is not clear whether proposed restrictions would also apply to ZTE.

According to Reuters, the warning was delivered to Vikram Misri, India’s ambassador in Beijing, last month. According to Indian telecoms minister Ravi Shankar Prasad, Huawei is one of eight vendors which has applied to take part in upcoming 5G technology trials, ahead of auctions late this year.

Swadeshi Jagran Manch (SJM), a right-wing group that is closely aligned with the ruling Bharatiya Janta Party (BJP), has demanded a ban on all Chinese companies from telecoms networks, saying in a recent statement: “China today controls a significant section of India’s telecom networks even though information dominance is at the core of China’s military strategy, posing an unacceptable security risk.”

The SJM also supports the government’s ‘Make in India’ programme to increase the use of domestic suppliers, or at least vendors which manufacture their equipment in India, for advanced telecoms networks.

Variations on that policy have been in place in successive governments over the past decade or more, but while they have encouraged vendors like Huawei to manufacture more equipment or devices in India, they have often been stymied by the limited progress of a homegrown telecoms network business. While Indian smartphone companies, such as Spice and MicroMAX, are gaining in market share at home and in some emerging economies, there is no significant equipment industry, certainly not one capable of delivering the advanced enablers of 4G expansion and 5G, which ZTE and Huawei are offering at tempting prices.

This is becoming urgent as the MNOs consider how to make their 5G networks affordable as well as capable in the next few years. ZTE has placed great emphasis on its ability to address the Indian operators’ predicament over low ARPU and high performance expectations. Ming told journalists: “Operators need to be able to offer something affordable to the population but at the same time they need to be able to make a return on their investment. Operators are struggling.”

He argues that the right choice of 5G technologies, from the most cost-effective suppliers, can square the circle. “India remains one of the most competitive markets in the world but new technology can help monetize 5G, which is something that operators in India really need to do,” said Ming, giving Massive MIMO – a technology where ZTE has a strong position – as an example.

“Massive MIMO technology allows operators to increase capacity and efficiency, using the same amount of spectrum, for their 5G services. It allows them to dramatically reduce the number of sites, by between three and five times. These types of technologies can help operators lower the cost of their network deployment and increase efficiency and capability of their networks in such a low ARPU market. Massive MIMO is a fundamental technology for 5G,” he added.

Vodafone Idea is already taking this advice to heart and recently deployed Massive MIMO technology across its 4G networks, to increase capacity, coverage and spectral efficiency, and to lay the foundations for a heavily software-based upgrade to Non-Standalone 5G between 2020 and 2022. To some extent this delays the real upgrade pain – moving to Standalone 5G with a 5G core – but Vodafone is following in the footsteps of several 5G deployers, including its sister operators in the UK and Italy, in seeing Massive MIMO as a stepping stone to 5G, which also squeezes more mileage out of 4G spectrum, allowing the migration to be more staggered as well as smoother.

Vodafone Idea’s CEO, Balesh Sharma, told a recent conference in Delhi: “5G in India, as with everywhere else in the world, is going to launch on a Non-Standalone basis, which means it will be built on top of our existing 4G framework. This is an absolute imperative to get the 4G story right. It means we have to spread the 4G networks right across the country and make sure that they are robust enough to deal with enormous volumes of traffic.”

However, he complained of another issue which could compromise the performance of Indian 4G and 5G and present another heavy cost burden – inadequate backhaul. As reported by LightReading, he told the conference that, in a speed test of the new Massive MIMO technology, he had witnessed speeds of 39.8Mbps on LTE.

This, he said, was “pretty good. That’s because we have already deployed Massive MIMO technology on our network. Massive MIMO is a 5G technology but we have already deployed over 5,000 Massive MIMO sites across the country. That is the second biggest deployment of that technology in the world. It’s already here.”

But he went on: “So, why are the speeds stuck at 39Mbps and not going on to 100Mbps or 200Mbps? Its not a spectrum issue, its not a radio issue, it’s a backhaul issue. We have to have fiber everywhere as far as I’m concerned. We could even make it a shared fiber network, to ensure that everyone has access to that essential fiber that will power their 5G backhaul. Either that, or we need a massive release of microwave spectrum in the E-band for mobile backhaul. If you give me either of those things, we could see speeds of 100Mbps+ on our exiting 4G networks today.”

Sharma’s reference to a shared fiber network is a clear nod to the necessity for greater physical asset sharing in every network domain if India’s 5G is to have any prospect of being both high performance and profitable. That could be supported by the government, with an easing of current red tape around shared networks, as well as lower spectrum prices and easier access to state fiber.

But of course, while all the operators are investing in fiber, for access and backhaul purposes, the most active telco in this respect is Jio. Jio has about 300,000 kilometers of fiber now, almost double Vodafone’s 156,000.

Will Jio be willing to open up its fiber to others? Possibly. Despite its huge investments in network infrastructure, as well as content and spectrum, its parent will not throw money at its business forever without seeing continuing growth in profits. And in India, that will mean a tight control over capex and opex.

Its physical assets in sites and fiber will give it currency to trade in negotiations with others, and it already relies on sharing of spectrum and locations with other operators. If it is serious about plans to become a cloud player, it will need to invest in, or access a third party’s, cloud infrastructure too, and that will be important to support its software-driven network vision. Earlier this month it signed an important cloud alliance with Microsoft to help its drive into enterprise markets and higher ARPUs.

The cloud-native 5G network, if implemented well, should reduce opex significantly over a 10-year period, supporting lower hardware capex, higher automation, and lower cost to introduce and support a variety of services and users. A greenfield operator has significant advantages because it does not have to cope with the migration and with maintaining the legacy.

However, moving to a public cloud like Azure, not just for data center applications but for the RAN, core and transport, remains daunting for many operators. Jio will reduce its risk by investing in at least some of the data centers itself and becoming a cloud/network operator, not just a cloud user. That will help it to gain a share of an Indian public cloud market that Gartner predicts will be worth $2.4bn this year, up 24% on 2018.

Investing in physical infrastructure such as fiber and data centers will help Jio achieve economies of scale, while it mitigates its cost and risk through partnerships and asset sharing. And it will also use those assets to increase its value over time through further partnerships or divestments. It has already offloaded its tower business, Reliance Jio Infratel, to Brookfield Asset Management in a deal which will help to reduce its debts of INR365bn ($5.2bn), and the debt mountain of its parent, which stands at INR2.87 trillion.

The next set of assets up for sale might be Jio Digital Fiber, in order to reduce debt further but also fund expansion of the Giga Fiber home broadband, and enterprise, businesses. Earlier this year, Jio Digital Fiber raised INR270bn ($3.9bn) in loans from a consortium of three Indian banks but the operator may go a step further, especially if regulator TRAI sticks to its resolve to set high minimum prices for 5G spectrum.