From network programming to programmable networks

— Shyam Mardikar

It was the spring of 1994 when I first stepped into the amazing and ever-evolving world of telecom. The telecom industry as we know it today didn’t exist then. DoT was about everything in telecom—from policy to regulation, and from execution to go-to-market. Teledensity was 0.8% and there was a ‘waiting list’ for eight million phones. Mobile phones did not exist. One joined the DoT through Indian Engineering Services and it took 18 months after the exam before one got on-boarded. With time in hand, I joined C-DOT, the ‘subsidiary’ telecom R&D organization of DoT, which conceived and delivered the famous Rural Automatic Exchange revolutionizing India’s rural telephony in the late 1980s and early 1990s.

Research in telecom was fun, and C-DOT thrived on its radically different environment and ethos than the parent DoT. The team was younger, the timelines were tighter and there was a driving underlying mission to create and apply technology. Software development, especially for telecom, was a complex rigmarole of cross compilers, machine language coding and endless cycles of optimizing the real-time operating system. All this, to get to the dial tone on the ‘make in India’ telephone exchange which could stand the ‘reliable grid challenged’ and non-air conditioned high-dust environment of rural India. The world owes some of the brightest software and hardware engineering talent to C-DOT, the bulk of which was handpicked later by who’s who of global telecom giants like Lucent, Motorola, Cisco, Nortel, Nokia, and Ericsson.

In DoT those times, telephony was fixed, with a copper plant footprint that was evolving from fat paper core copper cables to fatter jelly-filled copper cables for better reliability. Getting a telephone took anywhere between a year (you would have hit a jackpot) to a decade (you were lucky to have a telephone). The exchanges were analog (though electronic) and the era of digital was about to start. DoT sourced five worldclass digital technologies for their switching system and India saw its first brush with scale and state-ofthe-art technology. Alcatel’s OCB, Siemens’ EWSD, Lucent’s 5ESS, Fijutsu’s Fetex and Ericsson’s AXE gave the country as well as Indian telecom professionals, a glimpse into the programmable and digital world of future. Massive upgrades of the switching systems across the country created a new workforce in telecom— the new-age digital professionals, both from within DoT and from the equipment provider ecosystem.

Almost at the same time, NTP 1994 came into being, opening up telecom in India for private companies. The country was divided into 20 telecommunication circles for basic telephony and 18 circles for mobile services. The government threw open the bids to one private company per circle in the beginning. Mobile telephony was starting and what followed is now history. India had skipped the first generation of cellular telephony and was embarking on the 2G journey more than a decade behind the rest of the world.

What no one could envisage then that a country so late in joining the mobile bandwagon would get to the global center stage in a matter of two decades, showcasing a meteoric growth, and creating a legendary story. By the end of 2002, in the first seven years of private telephony, the customer base stood at a meager 10 million. Cut to 2012, this number would touch nearly a billion mark, a 100-fold growth towards making India connected. This was the first ‘V’ of Indian victories in telecom—the ubiquitous voice network.

A century unfolds

Erlang model for network capacity. Both of these meant opening up of the network to partners like Ericsson and Nokia, thus losing almost complete control on planning and managing the network—the only product a telco sells. While this meant putting forth an unflinching faith in the partner’s capability to create and scale the network, it also meant the creation of a novel and collaborative governance framework to ensure customer experience during the then evolving mobile ecosystem. The result— for nearly 6 years, Airtel created capacity exponentially, doubling year on year.

rolled out, Indian telecom transitioned into its second ‘V’ of mobile ‘video’ consumption.

3G networks in the country were short-lived, owing to a late start in India and a faster transition to 4G—a more spectrally efficient and broadband ready generation. While 4G networks started to get traction by 2013–14, it was not up to 2016 that the next radical innovation hit India—Reliance Jio.

The phenomenon called Jio

Flipping the ecosystem over from the voice factory to a data abundant market was Jio’s first contribution. By 2016, India was hovering around 145th position on global data scale; in the next 2 years, it would rise to number 1. Pan India data consumption in September 2016 was around 25 crore GB per month; in the next three years it would rise to more than 700croreGB per month, with Jio contributing to 430 core GB per month. Even on voice front, voice on Jio’s all IP-network is nearly 1000 crore minutes per day—a number equivalent to what all of India did in the month before Jio started services.

This mega-scale creation of a greenfield network and its success to amass the quantum of customers and traffic has been the latest telecom story that India is proud of. The gigantic scale of more than three million LTE cells, nearly 400 million customers on a pure-play 4G network, and 1,000 crore minutes of all-VoLTE voice is unheard of, even anywhere globally.

In addition to the creation of a telecom network which is hyper-scale and all-IP, something that probably no other operator in the world can boast of, the other big contribution of Jio to India is the upgrade of the device ecosystem. A price-sensitive market like India, now handles nearly half a billion 4G devices, thanks to the seeding of low-cost, homemade 4G devices by Jio. These phones not only served as the first 4G phones for millions of Indians but also pushed the mobile device manufacturers across the world to start building and supplying affordable and VoLTE-ready 4G devices in the country. Today, 100% of smartphones shipped in India are 4G ready.

Jio’s unique capability of driving serious researchbased indigenous products is not limited to devices. As of now, all of the voice traffic on the network, is processed on a completely home-grown TAS, the telephony application server of IMS. Jio’s TAS in its current scale and form is bigger and more software-centric than the largest of European and Chinese equipment providers serving both mobile and fixed voice in a truly convergent architecture.

Lastly, the platform version of network, which is fast becoming the de-facto architecture, is putting India as the biggest thought leader across the globe with futuristic capabilities to maximize both efficiency and performance by disaggregating the hardware and software components, by designing a heavy-duty user plane and a hyper-efficient control plane, and by creating a data-driven decision algorithm that runs the network in an automated, predictive, and zero-touch manner.

In the last 25 years, telecom in India has achieved more than what the rest of the world has managed in more than a century. Staring ahead is the golden decade of 2020s, where we can not only hold our global leadership in scale, complexity, and innovation, but also drive the technology outlook in telecommunication by redefining 5G, for individual human beings, for homes, for enterprises, and for machines and almost all sectors of life through one digital core that is ubiquitous and converged. This will be a truly digital India that the world will look up to.

If there is one consequence of liberalization of India’s economy—from the Nehruvian permitlicense- quota socialistic pattern—it is in the realm of telecommunications and internet. The change from a situation of “apply apply, no reply” to “hath hath mein telephone” and “gaon gaon mein internet” is the remarkable outcome of demonopolisation, regulated competition and exemplary entrepreneurship in our country. went through an almost crippling phase during 1994-99. The DOT was the licensor, a deprived service provider and a regulator laying down the conditions of license and regulating competition unlike anywhere in the world. The deprived DOT as operator-licensor-regulator put crippling conditions on the competitors. To mention just a few:

• While the universal practice is that the caller pays, the

By 1999 the P-Telcos were in dire financial straits. Fortunately, the Atal Behari Vajpayi government heeded the wise advice of telecom liberalisers like me and economists and as importantly, the bankers who funded the P-Telcos but were not getting the repayments according to schedule. The result was the National Telecom Policy 1999 (NTP-1999) ignoring the usual and characteristic, destructive criticism of leftists and their parties. Vajpayee’s government allowed migration of P-Telcos from upfront license fee payment regime to one of revenue sharing. The NTP 1999 also converted the DOT, provided telecom services into a corporation—the Bharat Sanchar Nigam Limited (BSNL) in October 2000.

NTP-1999 let the BSNL compete with the P-Telcos in every variety of telecom service, including internet. This revolutionary NTP 1999 was the outcome of the acceptance of the recommendations of the 1998 Prime Minister’s (Vajpayee) National Task Force on IT and Software of which, the late N Seshagiri, Prof Indirasan, myself and the unusual IAS officer N Vittal among others were members.

We as a group were responsible for making 108 recommendations, which were accepted within one month of submitting the report and within four months of constituting the Task Force. The distinction between telephony and data was ended because of the electronification, digitization and computer- processing of every type of telecommunication—voice, text, data and picture.

There was a sustained criticism from telecom liberalisers within the country and foreign companies investing in Indian telecoms, who were outraged by a competing government operator and licensor (DOT) regulating the private companies. To facilitate fair competition, the statutory Telecom Regulatory Authority of India (TRAI) was constituted in 1997. With a number of Telcos in the field, disputes arose between the licensor (government) and among Telcos in regard to interconnection, interpretation of licence conditions, etc. The Vajpayi government therefore constituted another authority, Telecom Disputes Settlement and Arbitration Tribunal (TDSAT) in 2000. The NTP 1999, the TRAI 1997 and TDSAT 2000 and the government Internet policy (1998) unleashed hyper competition leading to the provision of 19 million telephones in one month in the year 2010-11.

The PSU, Videsh Sanchar Nigam Limited, of which I was the first Chairman and Managing Director, was privatized in Feb 2008, again due to persistent campaigns by, among others, the Center for Telecom Management and Studies (CTMS) which I founded in 1998 (even when I was the CMD of VSNL) to wage a crusade for ending the telecom monopoly and ushering free market competition under proper regulation.

There are two things which have led to the end of the euphoric telephone and internet plentitude and descent of Telcos into distress. The primary reason which so far has not been highlighted is the immoral claim of ownership of radio spectrum (so essential for mobile telecommunications) by government and auctioning it as though it is its property to realise the highest revenues. Radio spectrum the Telcos use is not created by government or by God. It is created by the operating companies. It can be used again and again and is therefore an unexhausting resource the companies create. The only legitimate function of government is to regulate its use by slicing it into different bands and allotting them to different companies so that the use of the same spectrum in the same service area by more than one company does not lead to disorder. The

regulator can realize the cost of regulation-allocation of the spectrum and monitoring its use. this, they are now to pay Rs 92,000 crore to government. Government gifting Rs 82,000 crore to BSNL/MTNL is a drain on tax payer’s money.