Vish Iyer, Director, Video Architecture, Cisco Asia
The Indian economy has been growing steadily over the last few years. In order to sustain this growth there is a need to ensure a more well-rounded development. Building an extensive broadband network across the country and enabling universally accessible, inexpensive broadband can ensure efficient delivery of technology-enabled healthcare, remote education and improve access to government services besides helping increase GDP.
Broadband scenario in India
Broadband penetration in India is extremely low. Most broadband customers in India today are primarily concentrated in urban areas; semi-urban and rural areas are virtually excluded. Cost constraints, lack of compelling local content, limited availability, and a fragmented group of industry operators are all barriers to broadband penetration. Another barrier to widespread penetration of broadband in India is availability. Unlike other countries, India has not experienced a smooth progression of technologies that lead the way towards an obvious evolution to broadband. DSL for example is the most common evolution path in other countries, but India does not have large enough base of wired households ready to upgrade.
It is accepted that broadband access is required in India and that it should be provided to far greater numbers of subscribers than today. There are many technologies that can help: Telephone line based (ADSL2+); wireless (4G) or cable ( cable TV based). While all of them can increase broadband spread, their implementation is not that easy. 4G based services will not have enough spectrum (bandwidth) to service the cities (too many customers for bandwidth available) but can provide service in regional areas and a fill in (mobile) service in the cities (same spectrum but fewer customers).
Telephone line based technology needs twisted pair copper cabling to be in place and it is acknowledged that twisted pair based technologies will form a large part of broadband deployment in India. Cable TV is widely deployed today with somewhere between 60 and 120 Million subscribers spread across the footprint of many thousands of local operators - the exact number of subscribers will fall out of the digital TV upgrades currently occurring.
Cable has the disadvantage that the plant is what is termed one way (head office to the customer video distribution) and is often of poor quality in material and construction. Such factors limit the ability of the existing cable plant to pass DOCSIS signals and support broadband. But notwithstanding such limitations, cable technology provides the most promise, because it can leverage what is deployed now with light touch plant upgrade to support a basic DOCSIS service.
Cable digitization is expected to change the cable landscape significantly over the next few years. The onset of digitization will provide the cable sector with an industry status to attract institutional financing and become organized. Investment in the sector will help to upgrade existing infrastructure and provide value-added services to customers.
What DOCSIS 3.0 and 3.1 bring
DOCSIS 3.1 is an advanced version of DOCSIS technology. Of greater significance is that DOCSIS 3.0 already has standardized the upstream spectrum to provide much higher upstream (sending data from homes) data capacity than what is available in legacy plants seen in mature HFC markets. India has a great advantage here - if the existing plants are upgraded, they can take advantage of the increased spectrum on day one because there is potential to provide higher bandwidths. The fact that noise - the bugbear of all HFC plants - can be reduced using this new technology is an added advantage.
Deep fiber versus DOCSIS
There are a lot many arguments that say ' fiber to the home' is a future proof solution. The fact is that most (if not all ) of the national based FTTH rollouts are public funded not privately funded. This means the cost of such a rollout is very high and difficult for a business to make a case for. While there is no dispute that it is an ideal 'future' solution, there are more gradual and phased approaches to achieving the same end game that allow operators to provide broadband service and gain monetarily.
The comparison of a DOCSIS and HFC based broadband solution with that of fiber based solution is comparing where over subscription occurs (over subscription being the ability to sell more bandwidth than you actually have in place). The higher the ability to over sell, the more money can be made but at the expense of the service provided to the subscribers. Over subscription levels are market driven. If it is too high, the customer will migrate to another service - 'churn' in broadband parlance. In the end, everything is driven by over subscription levels: CAPEX, OPEX are all crafted to meet customer service expectations and maintain enough revenue to make a business case.
So what is the difference between deep fiber and DOCSIS solutions? The point at which over subscription occurs is different in each architecture. In a deep fiber solution over subscription occurs in the head office, with the plant deployed based on optical link budgets that are governed by the optical split levels; the optical split levels are engineered to access the maximum number of subscribers for the dollars invested. In a DOCSIS + cable plant the over subscription is crafted to suit subscriber's needs and occurs more in the plant than in the head office and more in the upstream (sending data to the internet) than the downstream where enough spectrum is available.
For emerging markets the end result is the same. Over subscription is required. The backbone to the global internet is not yet mature (read low capacity and expensive) and maximum benefit must be leveraged from what backbone bandwidth is available. Over subscription will and must occur somewhere for a business case to succeed.
With a DOCSIS plant - especially just after digital TV migration - downstream bandwidth is not a restriction. Even the current Indian digitalization of the cable plants will not use the bandwidth that is available. That said, it is easier to reserve spectrum today for broadband than to allow digital TV to use it all and fight to regain broadband bandwidth later on. Even if low peak speeds are to be offered per subscriber, it is very easy (and highly recommended) to provide 100s of Mbps of shared capacity accessible by subscribers on day one of any broadband deployment - more peak speed than can be provided per subscriber with a deep fiber solution.
Such bandwidth can also be expanded as needed due to the broadcast nature of the cable plant - add capacity at the head office with no plant changes on a pay as you grow basis. This contrasts with a deep fiber deployment where the fiber must be deployed with enough capacity on day one for future needs. Upstream is traditionally seen as a weak point in DOCSIS technology but this need not be the case in India. This weakness pertains to mature markets that grew out of pre-DOCSIS 3.0 technologies. Mature markets have cable plants built to earlier specifications and it is financially difficult to justify the upgrades to the newer technologies - increase in revenue does not justify the cost. India has the advantage of upgrading the existing cable plants to the latest technology and realizing full revenue from such an upgrade.
DOCSIS is not the only cable-based solution that is possible. Ethernet over coaxial cable as a solution grew out of intra-home broadband distribution over main power lines and can provide a broadband service using propriety systems loosely based on HomePlug AV and HomePNA technologies. Such technologies are widely deployed in apartment buildings in China for example. Even DOCSIS can be deployed using miniature DOCSIS head end devices deep in the HFC plant. Both options however suffer from a few fundamental flaws.
Each involve relatively expensive, active (electronically based) devices that are mounted externally in the plant. This creates a security (theft) and OPEX situation. Each device needs to be managed and given the scale of the potential Indian market managing tens of thousands of such devices becomes a major OPEX issue for an operator. Placement of DOCSIS head end devices in a secure office situation is much preferred. The number of devices is reduced - each device managing many more subscribers than the above alternatives. This helps leverage from plant over subscription to get more revenue from less hardware and the location can be air-conditioned and physically secure.
For broadband to become a mass medium in India, both industry and the government must take concerted action to address issues of concern. Operators (telecom and cable) must work to ensure ubiquitous supply by understanding the strengths and competitive positions of various broadband technologies and their applicability across different segments.
The government needs to provide more civic services online, establish a policy at the national/local levels to streamline access and advance the rollout of the fixed network infrastructure besides encouraging digitization of the cable industry.
According to the draft of the National Telecom Policy, the Indian Government aims to 'broadband on demand' by 2015 and achieve a target of 175 million broadband users by 2017 and 600 million by 2020. The draft policy proposes to revise the existing broadband download speed of 256 kbps to 512 kbps and subsequently to 2 Mbps by 2015 with higher speeds of at least 100 Mbps thereafter. Part of the policy is also to increase rural tele-density from 35% to 60% by 2017 and to 100% by 2020.
The Government spending on e-governance projects and SWAN initiatives is likely to drive the growth of the networking market further. Going forward, advanced wireless technologies like WiMax, 3G and 4G are expected to pace up the broadband growth in India. Wireless for last-mile connectivity in rural India and growth in broadband penetration will drive market growth and eventually help India transition to a fully networked economy.