Why virtual routers?

Most of us are familiar with routers. The little black boxes sitting in the corner of our home or office with messy wires sticking out. 

What is not common knowledge is that routers come in many shapes & sizes. Home Wi-Fi routers that are good for small number of devices and large enterprise routers that serve thousands of users. Costs also vary significantly with home routers costing a few dollars while enterprise models run into the thousands. 

No matter which category of router you fall in, it’s always hard to differentiate from the myriad choices. This is where virtual routers come in, with no hardware purchase and costing next to nothing. 

What are virtual routers?

Virtual routers are software routers that use a normal PC (or server) to do their job. 

With the advent of software defined networking (SDN) many functions of the modern enterprise are being virtualised to save cost and vendor lock-in. This is also referred to as NFV or network function virtualisation. In simple terms, functions that were performed by dedicated networking devices are now performed by software running on commercial-off-the-shelf servers (COTS).

 

The VSR

Inventum’s VSR is a virtual router that can replace any hardware router. The VSR software automatically detects the PC hardware and sets it up to function as a router. Connect your Internet & LAN to the relevant ports of the PC and voila! 

Performance

Performance of your virtual router really depends on your PC hardware. Depending on the number of CPU cores, memory and network cards virtual routers can easily handle multi-gigabit traffic. Advances from Intel® in multi-core processors and the DPDK framework have enabled network fast-paths on commodity servers.

Inventum has combined these developments with its own optimised software to build virtual routers that can deliver millions of packets per second (mpps) on Intel® Atom® & Xeon® processors. 

Go Virtual

If you haven’t yet dabbled with network function virtulization (NFV) the virtual router may be the first thing to try.

Fire up a machine on the cloud or use a hypervisor on-prem, the virtual router can address every use case that a physical router can. No lock-in to proprietary hardware and no expensive maintenance contracts. 

Download and try the VSR today!

Why Free Basics & Airtel Zero are bad for India

Free-Basics
“Free Basics is the new colonialism, its the new East India Company…”, said someone on television the other night about Facebook’s charitable initiative. I’m thinking to myself, there are no free lunches, so why is this even a debate?

Free Basics is an old debate about Net Neutrality. Its been defeated by regulation in the West, but in India we like debate. Facebook has made it so easy to sign a petition supporting Free Basics that hovering over the banner in your FB feed is enough! Many have accidentally succumbed.

Policy makers, telecoms, COAI & even TRAI are favouring a trial. They cite Delhi’s odd even car trial to reduce pollution as a precedent. Exactly how toxic pollution that is killing people compares with Free Basics eludes me.

To understand Free Basics, its first important to understand the Internet eco-system.

First The Basics

I see the Internet as a large reservoir of water. There are the content providers (Facebook & Google) who pump water into the reservoir and then us, the consumers, who tap in with little pipes drawing what water we need. The plumbing is owned by the telecom operators who charge both the content providers and subscribers for their respective pipes. The content companies make money from the subscribers who receive a service that they value and life comes a full circle.

Neutrality is in the fact that everyone gets the same water, just more or less, faster or slower. Everyone pays and none of us get champagne even if we can afford it!

Strange Bed Fellows

Circa 2000, the new new content companies were fighting tooth and nail with telecom operators. They wanted lower bandwidth prices, fatter pipes, unlimited always on Internet. The underlying motive was to drive faster Internet penetration leading to greater valuations which were based on unique user visits. Slow growth in Internet connectivity directly impacted their future.

AOL’s merger with Time Warner was an attempt to solve the slow growth problem and create a behemoth that built its own pipes and content. It has since become a corporate disaster case study.

Is Free Basics a similar attempt? Of course it is, just that the theatre is the third world. India still needs to connect billions and Facebook doesn’t have the patience to wait for our telecom providers and government to deliver connectivity. It wants to guarantee it’s market dominance in India’s future. Not charity, just plain old profit motive.

What is different this time is that Indian telecom providers want to join forces with Free Basics. Let’s try and understand their game.

How Internet Providers Makes Money

The provider game is simple. Lay a big fat pipe to the reservoir, say 1 gigabit (1,000 megabits). Then we split this pipe between a 1,000 paying subscribers, each sold a 10 megabit connection. Hold on, 1,000 x 10 is 10,000 megabits or 10 gigabits, so how do we fit all of them into the original 1 gigabit reservoir pipe? By betting on the fact that not all 1,000 subscribers will download at the same time. This is what we call oversubscription and this is how we make money.

Subscribers only see the ill effects of oversubscription during peak hours when everyone starts downloading and consequently everything slows down. Oversubscription to my mind is fair game, so long as it doesn’t mess with the end deliverable.

Content Differentiation & Deep Packet Inspection (DPI)

So how does a service provider differentiate content? How does your ISP create differential pricing?

DPI technology has been widely used by ISPs and telecoms to look inside your Internet traffic. Most people don’t know that their ISP is already differentiating P2P & torrent content, throttling it down to improve oversubscription and pack in more paying customers.

Free Basics, Airtel Zero and other similar initiatives all use DPI to control ALL your flows, throttle & (eventually) charge depending on what you access.

Think of it like a new toll gate for every website you wish to go to. Its only a matter of time that premiums will be linked to peak hour traffic!

How Airtel Zero Would Make Money

Airtel Zero is simply a differential pricing game. Zero charge for curated content, delivered faster, regular charge for everything else. An express lane of sorts, for the wealthy content owners to expand their market share while telecom operators add to their profits by charging the content providers for delivery. Of course, the real game all along is control over the medium and eventually the customer.

Imagine if your electricity company could charge you on the basis of which appliance you used in your house? Large appliance makers would subsidise electricity giving them an unfair advantage in the marketplace, making it tougher  for smaller appliance companies to compete. And in time, the nexus of the appliance and electricity companies would monopolise the market, affecting all consumers.

 

How Facebook Would Make Money

If Facebook is the only channel you can watch on TV, what would advertisers be willing to pay? You do the math.

With nearly a billion people waiting to come online in India, Facebook wants to accelerate connectivity, while ensuring they become the de facto medium with a captive audience.

No thank you, but India is done with colonialism. We can pay for our citizens to come online in good time.

Conclusion

The Internet is a great leveller because you cannot differentiate upon the end use or user. Governments should focus on building affordable broadband services accessible to all citizens with a view to empower and provide a level playing field to all stakeholders. TRAI & COAI should stop pushing agendas that are less than charitable.

Telecom operators must focus on building capacity, improving service quality and ridding themselves of a monopolistic mindset. Data is growing rapidly and companies like Airtel have benefitted greatly from 3G growth. If anything, they are responsible for driving voice tariffs to global lows and now trying to slice and dice the Internet is poor strategy.

Facebook should realise that free Facebook, WhatsApp & Instagram ain’t making anyone smarter. And basics without its competitor Google is a dead give away!

 

 

Full disclosure – I work for Inventum, one of the only carrier-class Indian router manufacturers. The company provides Broadband Network Gateways (BNG), DPI appliances, Routers, Billing & Provisioning systems for telecoms and Internet Service Providers. The views expressed here are my personal experiences as an entrepreneur in the Indian high-technology space.

 

Why Most Public Wi-Fi Hotspots Fail

Public WiFi hotspots have been around for over 15 years. New ones are being created everyday… and are shut down everyday as well. Various state governments in India are in the process of setting up public hotspots as part of their smart cities initiative. The Delhi Government went one step ahead and promised 30 minutes of free Wi-Fi to everyone, everyday. It’s a noble undertaking, but unless they really think it through, it’s unlikely to be more than an election promise that was delivered, but didn’t do anyone any good. Sort of like the BRT.

To put it simply, the problem is a combination of the user experience and funding model.

Hotels that bundle the cost of Wi-Fi in the room rent or large coffee chains that simply write it off as a utility expense like electricity or air-conditioning have a very well defined business model. They know that the revenue earned from their core business (room rent, food and beverages) can be enhanced, or at least sustained by making their location more attractive to customers with free Wi-Fi. Because they know the number of rooms they have, they know the number of visitors they can expect and hence, know how much it will cost to provide a certain quality of Wi-Fi. So, if the cost vs revenue ratio is acceptable, the service is provided and everyone’s happy.

The situation with most other hotspots isn’t so straightforward. Where there’s no clear way to recover the cost of Wi-Fi from other sources, operators attempt to entice customers with a few minutes of free Wi-Fi and then pester them to make an online payment or watch an advertisement. Here’s where the problems start.

Unless we’re just hanging around an airport waiting for a flight to be called or killing time at a coffee shop, most of us don’t have more than a few minutes to spare to sync emails or search the web for a phone number or whatever.

But the experience of connecting, registering, paying, logging in and finally using most public hotspots is so cumbersome and time consuming, that most people don’t bother. Unless the Wi-Fi service is far superior (i.e. faster, cheaper, reliable and convenient), most of us would rather just use the data plan of our mobile phones and be done with it.

The customers don’t use the Wi-Fi beyond the free period and the operators don’t make any money. Everyone loses.

There are ways to make it all work and it has been done before successfully. This article mentioned that Vodafone India might be moving in the right direction. So far, nothing has been announced officially. I guess we’ll just have to wait and watch.

Digital India, an entrepreneur’s perspective

Back in 1983, I fell in love with computing on the BBC Micro & the Sinclair ZX Spectrum+. Having access to a computer was a small miracle back then. What was cooler was that the computer at school, a SCL Unicorn (clone of the BBC Micro) was made by Semiconductor Complex Limited Chandigarh, Punjab. The micro was legendary & my fellow enthusiasts may recall that today’s popular ARM architecture had its genesis on one of these very puppies.

Fast forward 30 years and India’s electronics import bill is poised to over take oil, topping $400 billion by 2020.

How did we get here? How does a nation that launches missions to Mars become a relative nobody in the high-tech electronics (ESDM) space.

The Services Gold Rush

We can’t build them, but we can programme them. And programme we did. Infosys, Wipro & HCL are stuff of legend. They got the country hard earned dollars in dark times. The brightest emigrated, most ending up in silicon valley where alongside the Chinese they dominated.

And while we got the gold, at home we forgot about the hardware bits – Research, Design, Development & real innovation. I guess that was relegated to government run ISRO, DRDO & other similar institutions.

Mamla Risky Hai (Matters of Risk)

The Israelis have made a fine art of creating, growing and then harvesting multi billion dollar tech companies. At the heart of their success lies a strong will, the right eco-system, risk capital, experience and serious R&D. The Chief Scientists’ Office (CSO) of Israel plays a central role in startups, providing capital, forging joint ventures, incubating & nurturing. Above all, the CSO is very protective of Israeli Intellectual Property.

India has no CSO equivalent, but has excellent talent (IIT aside) and daring entrepreneurs. Then why aren’t we seeing world beating Indian product companies in high tech? I believe a lot has to do with the lack of high risk capital for investment in basic research which is crucial to invention, creative destruction.

The truth is that tech services is a sure shot, building networking equipment for telecom or defence is not. Our financial apparatus including banks, private conglomerates, venture funds & the government really don’t like risk unless they can own and control it.

A conversation I had with one of India’s telecom Tzars many moons ago for for investing in telecom equipment startups summarised the sentiment “telecom is about regulation, not technology… our job is to manage regulation, sales & marketing; the tech we have outsourced to the Americans & Chinese…”.

It is no wonder then that despite being the second largest mobile market on the planet, we do not have a single Huawei, Cisco or Ericsson of our own. The gain of our mobile operators from sweetheart deals delivered by foreign vendors and their Exim banks ensured that any Indian innovation in core telecom products was killed off. A tragedy, considering a large part of our $400 bn import bill in 2020 originates from telecom. Those iPhones really add up quickly.

Risk capital is not available to Digital India entrepreneurs period. What you can get is high interest debt. Schemes of the government are dysfunctional, impractical targeting a few lacs to get you started, but how does one compete with Cisco & Huawei? Almost all tech VC funds are foreign and the flavour of the month seems to be E-Commerce. So how do $1-5 million revenue companies in Digital India access capital without pledging their homes?

Government Stimulus

Would India be a market leader in auto components had it not been for Maruti? Probably not. And what of India’s IT boom without the early income tax exemptions. And what of our crony capitalist, none of whom would have existed without government patronage.

So why is ESDM the bastard child? Maybe ESDM was missing its crony!

The most recent meaningful regulation in ESDM has been PMA. DeITy notified the Preference to Domestically Manufactured Electronics Goods (PMA) in 2013. The policy promotes Indian goods by reserving part of government tenders for local companies. No concession on specifications or quality, only exemptions from tender conditions like “must have 5,000 cr net profit for the last 3 years”; after all which Indian company has that! PMA was instantly opposed by COAI, probably because private telecom operators were made party to it. Their chief argument was that Indian companies cannot ever make telecom gear. It eludes me as to why potentially the biggest sponsors of telecom ESDM would oppose PMA. And while PMA was diluted to apply only to PSU purchases, it did yield immediate results with two silicon fabs being announced.

For me, the government’s message to foreign vendors was loud & clear – India no longer wants to sell services at $50/hour, we want a piece of the Intellectual Property pie, to be registered, manufactured & taxed here.

What Indian companies need are many more policies like PMA. More government stimulus, capital & industry sponsorship. Defence must participate too.

Make in India

Nothing is easy to make in India. Its hard enough to set up an office, let alone build a product. Being an Indian entrepreneur reminds me of the video game Asteroids, fragments of menacing celestial bodies coming at you in full force from all directions. Patience is crucial to persevere.

And then Modi’s Make in India happened, Digital India happened. He put sexy back into making hardware. He did in 3 words what we had struggled to do for decades. There is hope.

I believe the road is going to be long and arduous. We need to sort out our labour laws, manufacturing, incentives, logistics, taxes, ancillary industries, human resources, IPR laws and so much more. And while foreign vendors like Cisco are already taking advantage, its the smaller Indian entrepreneurs that need to be empowered to drive this revolution.

With Mr. Modi on his way to Silicon Valley to woo the diaspora, my question is, who is listening to the little guys at home.

 

 

Full disclosure – I work for Inventum, one of the only carrier-class Indian router makers. The company provides Broadband RAS, Service Routers, Billing & Provisioning systems for telecom, enterprises, education & hospitality. The views expressed here are my personal experiences as an entrepreneur in the Indian high-technology space.

Food, Clothing, Shelter and Free Wi-Fi

With Wi-Fi being politicised by the Aam Aadmi Party (AAP) in the recent Delhi elections, it seems our politicians have found a new sop to entice the young voter. Free Wi-Fi is the catch phrase being canvassed across the country, synonymous with smart cities, real estate projects and even the railways.

Will these projects be delivered? And is Wi-Fi even the right technology for smarter cities and an enlightened population? In the case of the former, let’s just say most projects will remain election promises; the suitability of Wi-Fi as a technology to reach Indian masses however, requires more thought.

Wi-Fi Basics

Wi-Fi was invented to take your Internet wire at home or office and make it airborne. For the uninitiated and puritans do forgive me, but this analogy works every time – Wi-Fi does to your home Internet connection what the cordless telephone did to your landline. The idea had always been to deliver fast data over short distances, freeing us from our desks.

The Achilles heel for Wi-Fi is its range and radio interference, dictated by physics more than government regulation. The typical range of your Wi-Fi home router is that of a yesteryear digital cordless telephone. They do of course use similar frequencies (2.4 & 5.8Ghz) and have the same issues penetrating brick and concrete walls in average Indian homes. With Wi-Fi operating on an “open to all” radio frequency, it pre-disposes the technology to lots of interference. Just think of it like rush hour in Delhi, everyone coming at you from all direction, zero lane discipline, and a free for all.

If you have a large enough home to need two routers, you most likely have experienced the limitations of range and interference. Most people I know will have different SSIDs “Ground Floor” and “First Floor” jumping between the two as they walk around the house. Each time you hop, your web page or video stalls.

So I ask the question, is Wi-Fi the right technology to deliver a quality user experience to the populous in a cost effective (since nothing is free) manner?

Experts, mostly expensive Wi-Fi equipment vendors, will argue that their gear has evolved to provide better range and deal with the hopping issue. Many cite examples of successful city-wide projects like Wireless@SG in Singapore. US & European service providers are using Wi-Fi Offload to divert subscriber 3G data traffic over to hotspots in an effort to decongest their mobile networks.

And while these innovations may work very well in other countries, the real question in my mind remains – is it right for Delhi, is it right for India?

Having been part of a company that has contributed to building some of the largest Wi-Fi networks in the world, here are my arguments & learning:

  1. Wi-Fi access points including the outdoor expensive type have an average range of 500 meters or less. Range falls sharply with obstructions such as walls, trees and of course other devices on the same frequency including your 2.4Ghz microwave oven and your existing home Wi-Fi router;

  2. Each access point will need an Internet wire connection to take the airborne signals and switch them back to the ISP. If you live in Delhi, you probably already know how difficult it is to get a quality DSL connection, so getting a reliable wire to each (almost) access point will be a serious challenge;

  3. To cover 1 square kilometer you would need dozens of Wi-Fi access points, each requiring a pole to be mounted, uninterrupted power, Internet backhaul wire, protection from the elements and vandals; One look at the condition of our public CCTV surveillance systems and it will be a wonder if the access point survives the first week;

  4. And then there is the issue of reliable power for each of these access points, sprinkled all over the city. One of our customers (who had installed several outdoor power systems to fuel their Wi-Fi business) summarized the situation quite well “our biggest challenge was protecting the UPS batteries from being stolen from the pole…” ;

  5. Right-of-way from civic authorities to use the poles and permission from homeowners to install and then service the access points is another huge challenge. Unlike mobile towers, you need equipment every 100 meters or so and with prevailing health concerns, fancy getting those;

  6. And then there is the user experience. I have tried and mostly failed to login to the hotspots that exist. It works well in hotels and some airports, but tried a coffee shop recently? No wonder India’s 75 million broadband connections are really 60 million USB data dongles (TRAI Sep 2014).

So seriously, our government wants to do all this for free?

Let’s call the Singaporeans

Singapore recently revamped its decade old Wireless@SG Wi-Fi network, free for citizens & visitors. It was built almost entirely through government grants and operated by a chosen few small service providers. Ask any Singaporean and they will tell you how poor the network used to be.

The Singapore government has now turned to mobile operators (such as M1) to build and manage Wireless@SG. Are the people being served? Time will tell, but Singapore doesn’t have the answer yet.
So will Delhi succeed where Singapore failed?

I think the people would be better served by:

  1. Increasing landline broadband connectivity – believe it or not, India has only 15 million wire based broadband connections serving 304 million broadband users; With bigger bang for their buck in mobile, operators are simply not investing in wires. Wikipedia says 13 million wires against China’s 175 million

  2. Give every citizen of Delhi free 30 minutes worth of 3G data instead. Everyone already has a cellphone and USB data dongles are cheap for those wanting to connect their laptops. The mobile infra is in place and reliable.

  3. Develop every government office as a free hotspot for citizens

  4. Partner local businesses to develop small Wi-Fi zones & hotspots in public private partnership by incentivizing local businesses