Broadband to Empower Rural India


[The paper is based on and borrows from some recent articles on Telecom in rural India [1-3]]
Ashok Jhunjhunwala, David Koilpillai and Bhaskar Ramamurthi,
Professors, Department of Electrical Engineering,
IIT Madras,Chennai 600036,
India

1.Introduction

Urban India has been on the move over the last ten years and its growth has accelerated especially during the last five. However, the same cannot be said about rural India. Urban Indians are full of confidence, but rural Indians do not see much of a future for themselves. The only change in the lives of many rural people is availability of some television, which in fact, has created greater aspirations amongst them. They can now visibly see the difference between life urban and rural India and cannot understand why they are being left so far behind. In a democratic set up, where some form of election (central, state or local) takes place constitute almost ever year and a half, the feeling of deprivation amongst the rural people plays havoc, especially as rural people 70% of India.s population. Every politician would be forced to promise more and more to rectify this deprivation and Government policies would be forced to be populist. Yet there are a few constructive programs which can really change life in rural India. Governments will be over-turned frequently, as no populist measure would rectify the great divide that is getting accentuated with India.s urban growth. In fact, the policies required to sustain even urban growth would be under threat and would not be consistently pursued. The only answer is a quick and urgent focus on rural areas towards rural transformation.

2.Rural India today

Rural India consists of approximately 700 million people, living in 638000 villages. The average village comprises 250-300 households, and occupies an area of 5 km2. Most of this is farmland, and typically the houses are in one or two clusters. Villages are thus spaced 2-3 km apart, and spread out in all directions from the market towns. The market centers are typically spaced 30-40 km apart. Each such market center serves a catchment of around 250-300 villages in a radius of about 20 km. As the population and the economy grow, large villages morph continually into towns and market centers. The villages are characterized by low incomes. About 85% of the households have an income less than Rs 3000 per month (amounting to Rs 600 per month per person, assuming a family of five). Two-thirds of the households are dependent on agriculture for income, and even this is often seasonal and dependent on rainfall Rural India now has very little industry. Its people are mostly under-employed in agriculture. At the same time, agricultural growth in India has slowed down to 1% over the last decade, falling behind even the population growth. Fortunately most of rural India has some form of road connectivity today (even though often much of is may be in a bad state) and at least one bus would ply to a village every day. A railway station is also not very far off. Highways connect towns, which are rarely farther than 15 Kms from any village. Also, a significant number of villages are on the electrical grid. However the grid supplies power only during the period when the demand in urban (and industrial) India is low. During peak-demand period, urban India has the capability of sapping all the power produced and the rural areas are supplied only whatever is leftover. So even when the power flows into the rural grid (0 . 18 hours per day, depending on the State), the voltage could fluctuate between 90V (reflecting higher demand) to 440V (during nights when no one is using power). Decentralized power generation in rural India may be the only answer to this problem in the short run. Telecom technology has advanced very rapidly. Even though only a small percentage of villages today have reliable telecommunications connectivity today, the situation is changing fast. With the rural thrust, it is reasonable to expect that most villages in India will have mobile coverage and a broadband Internet connection within the next three to four years.

3.The Telecom situation today

The mobile revolution of the last five years has seen base stations sprouting in most towns, owned and operated by multiple operators, including the state-owned company BSNL. The base stations of BSNL, as well as those of the new operations are also networked using optical fiber laid in the last five years. There is a lot of dark fiber, and seemingly unlimited scope for bandwidth expansion. The solid telecom backbone that knits the country ends abruptly at the towns and larger villages. Beyond that, cellular coverage extends mobile telephone connectivity only up to a radius of 5 km, and then the telecommunication services simply peter out. Cellular coverage can and will grow, but this will depend on the rate at which infrastructure and operating costs drop, and the rate at which rural incomes rise. Fixed wireless telephones have been provided in tens of thousands of villages, but it would be safe to conclude that the telecommunications challenge in rural India remains the .last ten miles.. This is particularly true if one were to include broadband Internet access in one.s scope, since the wireless technologies currently being deployed can barely support dial-up speeds. This then is the rural India in search of appropriate broadband wireless technology: characterized by fat optical-fiber POPs within 15-20 km of most villages, fairly homogenous distribution of villages in the plains, poor rural cellular coverage, and low incomes. The last aspect makes the provision of basic telecommunications as well as broadband internet services all the more urgent, since ICT is an enabler for wealth creation.
3.1 GSM and CDMA dominate today but 3G will be available in time
Before we look at Broadband technologies for Rural India, let us take a look at mobile technologies of today (we will mostly focus on GSM, though CDMA systems are also present in India today). It may not be readily evident that the bottleneck in rolling out services to rural areas is not the cost of electronic equipment, but is actually due to the following:
i.The most significant cost component is the site preparation and the erection of the tower. Infrastructure like roads and electricity may have to be first set up. The towers are about 40 m tall, and require considerable amounts of expensive steel.
ii.The second highest contributor to the cost is the power infrastructure . RF cables running to the top of the tower, the power amplifiers, RF filtering and the transceivers. Roughly 55% of the cost of the base station equipment is in these RF components4.
iii.The maintenance of cell site infrastructure requires local personnel who should be trained to deal with the problems that arise in wireless equipment.
iv.Availability of ultra-low cost (ULC) mobile phones at costs below Rs 1500 with financing packages.
v.Proper distribution infrastructure for phones, SIMs, spares and accessories in the remotest areas, and availability of basic training to users so that they can use the phones properly.
vi.Billing and collection infrastructure for pre- and post-paid subscribers.
If one accepts these as the real bottleneck, then it is immediately evident that as soon as there is sufficient GSM voice coverage across India, we are already past the key hurdles for upgrading. The cell sites and towers are set up and maintenance, distribution, user training and billing/collection infrastructure put in place. One cannot afford to deploy any new cell sites, but only add electronic equipment at existing cell sites. To deploy 3G at a cell site, Node B equipment has to be installed (instead of, or in addition to, the GSM BTS equipment). The cost of such Node B equipment has been falling by approximately 40% each year over the last 4 years. Taken together with the fact that 3G offers more capacity than GSM, the 3G Node B is just 50% more expensive today than the GSM BTS to deploy the same voice capacity4. It has already been seen that 55% of the cost of base station equipment is in the RF. Since a single 3G channel of 5 MHz replaces many GSM channels of 200 kHz required achieving the same capacity . the RF costs of 3G systems should over time be lower than that of GSM systems. Thus 3G will eventually lead to cheaper equipment than GSM, resulting in Mobile Broadband infrastructure in India.

4.Broadband for Rural India

When considering any technology for rural India, the question of affordability must be addressed first. Given the income levels, one must work backwards to determine the cost of an economically sustainable solution. The 200 odd households in a typical village having disposable incomes can spend on an average Rs 50-100 per month for telephony and data services. Assuming an average of two public kiosks per village, the revenue of a public kiosk can be of the order of Rs 5000 per month. Apart from this, a few wealthy households in each village can afford private connections. After providing for the cost of the terminals, it is estimated that a cost of at most Rs 15000 is sustainable for the connection. This includes the user equipment, as well the per-subscriber cost of the network equipment and infrastructure (towers) linking the user up to the optical fiber POP.
4.1 Coverage, system gain, and cost of towers
It has been mentioned that one needs to cover a radius of 15-20 km from the PoP using wireless technology. The .system gain. is a measure of the link budget available for overcoming propagation and penetration losses (through foliage and buildings), while still guaranteeing system performance. Mobile cellular telephone systems have a system gain typically of 150-160 dB, and achieve indoor penetration within a radius of about 5 km. They do this with Base Station towers of 40 m height, which cost about Rs 500,000 each. If a roof-top antenna is mounted at the subscriber-end at a height of 6 m from the ground, coverage can be extended up to 15-20 km. When the system gain is lower at around 135 dB, as in any line of sight system, coverage is limited to around 10 km and antenna-height at the subscriber-end has to be 10 m in order to clear the tree tops. This increases the cost of the installation by about Rs 1000 per connection. Thus, roof-top antennas in the villages are a must if one is to obtain the required coverage from the fiber POP. A broadband wireless system will also need a system gain of around 150 dB if it is to be deployed with 6m poles. This system gain may be difficult to obtain at the higher bit-rates supported by emerging technology, and one may have to employ taller poles in order to support higher bit-rates at distant villages. There is an important relationship between coverage and the heights of the towers and poles, and thus indirectly their cost. The Base Station tower must usually be at least 40 m high even for line-of-sight deployment, as trees have a height of 10-12 m and even in the plains one can expect a terrain variation of at least 20-25 m over a 15-20 km radius. Taller Base Station towers will help, but the cost goes up exponentially with height. A shorter tower will mean that the subscriber-end installation will need a 20 m mast. At around Rs 20000, this is substantially costlier than a pole, even if the mast is guyed and not self-standing. The cost of 250-300 masts of this type is very high compared to the incremental cost of a 40 m tower over a 30 m one. With 40 m towers, poles are sufficient at the subscriber-end, and need rarely be more than 12 m high. In summary, for a cost-effective solution the system gain should be of the order of 150 dB (at least for the lower bit rates), a 40 m tower should be deployed at the fiber POP, and roof-top antennas with 6-12 m poles at the subscriber-end. The cost per subscriber of the tower and pole (assuming a modest 300 subscribers per tower) is Rs 3000. This leaves about Rs 12000 per subscriber for the wireless system itself, inclusive of both the infrastructure and terminal sides.
4.2 What constitutes broadband?
The Telecom Regulatory Authority of India has defined broadband services4 as those provided with a minimum data rate of 256 kbps. At this bit-rate, browsing is fast, video-conferencing can be supported, and applications such as telemedicine and distance education using multi-media are feasible. There is no doubt that a village kiosk could easily utilize a much higher bit-rate, and as technology evolves, this too will become available. However, it is important to note that even at 256 kbps, since kiosks can be expected to generate a sustained flow of traffic, 300 kiosks will generate of the order of 75 Mbps. This is a non-trivial level of traffic to evacuate over the air per Base Station, even with a spectrum allocation of 20 MHz.
4.3 Suitability of broadband wireless technologies
One of the pre-requisites for any wireless technology for it to cost under Rs 12000 is that it must be a mass-market solution. This will ensure that the cost of the electronics is driven down by volumes and competition to the lowest possible levels. As an example, both GSM and CDMA mobile telephone technologies can today meet the above cost target, (however, an even lower cost is needed for a non-broadband technology since the services provided are limited). The third-generation cellular telephone technologies will probably continue to meet this cost target while offering higher bit-rate data services. However, they will not be able to provide broadband services as defined above (as at most they will provide 64 kbps to each user). If one were to turn one.s attention to recently standardised broadband technologies such as WiMAX-d (IEEE 802.16d)5, it is found that at present volumes are low and costs high. Of these, WiMAX-d has a lower system gain than the others (which are all around the required 150 dB). All of them will give a spectral efficiency of around 2 bps/Hz/cell (after taking spectrum re-use into account), and thus can potentially evacuate 40 Mbps at each Base Station with a 20 MHz allocation. However, high cost due to low volumes is the inhibitory factor with these technologies. It is likely that one or more OFDMA-based broadband technologies will become widely accepted standards soon. One such technology is WiMAX-e (IEEE 802.16e)6 that is emerging rapidly. These will certainly have a higher spectral efficiency, and more importantly, if they become popular and successful, the cost will be low. However, it will be several years before a widely-adopted technology derives the benefits of market size and the cost drops to affordable levels for rural India. The obvious question is whether there are alternatives in the interim that meet our performance and cost objectives.
4.4 Near-term Technologies
4.4.1 Broadband CorDECT technology
One of the earliest and most widely deployed examples of such re-engineering is the corDECT Wireless Access System developed in India7. It provided toll-quality voice service and 35/70 kbps Internet access to each subscriber without the bandwidth having to be shared. The next-generation Broadband corDECT system has also been launched recently, capable of evacuating 70 Mbps per cell with a 5 MHz bandwidth (supporting 144 full-duplex 256 kbps connections simultaneously). With this system, each subscriber will get 256 kbps dedicated Internet access, in addition to toll-quality telephony. These systems are built around the electronics of the European DECT standard, which was designed for local area telephony and data services. Broadband corDECT incorporates added proprietary extensions to the DECT physical layer that increase the bit-rate by three times, while being backward compatible to the DECT standard. Thus, the spectral efficiency goes up three times when compared to conventional DECT. Additionally, dual-polarization antennas have been used to exploit polarization isolation while till operating within the DECT MAC framework, and further double spectral efficiency. More importantly, all this has been done while retaining the use of the low-cost DECT chipsets. The system gain in Broadband corDECT for 256 kbps service is 125 dB. This can be increased by a few dB, where required, by increasing the antenna gain at the subscriber-end (which is 11 dBi now). This is sufficient for 10 km coverage under line-of-sight conditions (40 m tower for BS and 10-12 m pole at subscriber side). A repeater is used, as in the corDECT system, for extending the coverage to 25 km. The corDECT system, and now the broadband corDECT system, both meet the rural price-performance requirement comfortably, but with the additional encumbrance of 10-12 m poles and one level of repeaters. The first-generation technology is proven in the urban and rural Indian environment, and much is known about how to deploy it successfully. The Broadband corDECT system works with the same deployment strategy. It is being deployed in significant numbers beginning 2007.
4.4.2 WiFi rural extension (WiFiRe)8: A new WiFi-based wide-area rural broadband technology
In recent years, there have been some sustained efforts to build a rural broadband technology using WiFi chipsets. The WiFi bit rates go all the way up to 54 Mbps. The system gain is about 132 dB for 11 Mbps service, and as in corDECT, one requires a 40 m tower at the fiber POP and 10-12 m poles at the subscriber-end. The attraction of WiFi technology is the de-licensing of spectrum for it in many countries, including India. In rural areas, where the spectrum is hardly used, WiFi is an attractive option, provided its limitations when used over a wide-area are overcome. Various experiments with off-the-shelf equipment have demonstrated the feasibility of using WiFi for long-distance rural point-to-point links. The more serious issue with regard to the 802.11 standard is that the commonly supported MAC protocol is a Carrier Sense Multiple Access (CSMA) protocol suited only for a LAN deployment. When standard WiFi equipment is used to set up a wide-area network, medium access efficiency becomes very poor, and spectrum cannot be re-used efficiently even in opposite sectors of a base station. A solution for this problem is to replace the MAC protocol of 802.11 with a MAC protocol more suited to wide-area deployment. Such a new MAC, christened WiFiRe, has indeed been defined8, and carefully, such that a low-cost WiFi chipset can still be used, and the in-built WiFi MAC in it can be by-passed. The new MAC can be implemented on a separate general-purpose processor with only a modest increase in cost. With the new WiFiRe MAC, it is estimated that using a single WiFi carrier, one can support about 25 Mbps (uplink + downlink) per cell. This would be sufficient for about 100 villages in a 10-15 km radius. Repeaters, possibly operating on a different frequency, will be needed for covering more villages over a greater distance.
4.5Tomorrow.s Broadband Technologies
In about few years, we would see significantly better broadband technologies, which could provide the 150dB system gain, even while measuring data speeds of 256 kbps or more for each connection. The three most promising technologies are all standard-based and are therefore expected to meet the price targets required for Rural India. These technologies are: IEEE 802.16 m 3GPP . LTE 3GPP2 . UMB We present each of these efforts in brief.
4.5.1 IEEE 802.16 m6
This is an OFDMA based standard emerging out of efforts of IEEE. The earlier version of the standard is IEEE 802.16 e which was finalized in 2006 and is popularly known as WiMAX. This broadband wireless standard, using state of art modulation, coding, scheduling and multiplexing would use multiple smart antennas at least at the base station side, to enable peak data rates of 100 Mbpsec for mobile users in 20 MHz spectrum. The working group finalizing the standard aims to finalize the requirements, channel model and evaluation methodology by May 2007 and make a proposal to ITU-R Working Party 8F (WP8F) for IMT-advanced requirements by March 2008. The principle stakeholders driving this effort are vendors developing 802.16 products, licensed carriers using 802.16 products and members of WiMAX ForumTM.
4.5.2 3GPP LTE9
A Third Generation (wireless) Partnership Project (3GPP) Long Term Evolution (LTE) was started with feasibility study on evolution of Universal Terrestrial Radio Access Network (UTRAN) in 2004 and grew with the recommendations for delivery of mobile broadband services by Next generation Mobile Networks (NGMN) initiative in 2006. A Technical Report (TR 25.913) provides detailed requirements, which include downlink peak date rate of 100 Mbps within a downlink spectrum of 20 MHz using the OFDMA technique. The uplink peak data rate is expected to be 50 Mbps with a 20 MHz uplink spectrum using SC-FDMA technique. It is proposed to support 200 users in active state in each cell. The users are expected to get high performance with mobility as high as 120 kms / hr. MIMO is expected to be used and an enhanced multimedia service is expected to be a part of the standard.
4.5.3 3G-PP210
The CDMA Development group (CDG) is collaborating with Third generation Partnership Project 2 (3GPP2) to define an Ultra Mobile Broadband (UMB) standard as an evolution of CDMA 2000. The systems requirement document was approved in May 2006 and uses scalable bandwidth up to 20 MHz. The forward direction peak data rate is expected to be as high as 500 Mbps in fixed and 10 Mbps in mobile environment using OFDMA. The reverse direction data rate is to be 150 Mbps in fixed and 50 Mbps in mobile environment using qusi-orthogonal transmission based on OFDMA, together with non-orthogonal user multiplexing with layered superposed OFDMA (LS-OFDMA). The reverse link also supports CDMA for control and low-rate, low latency traffic. The advanced air interface agreement has been reached by Technical specification group C (TSG-C) based on a consolidated framework proposal submitted by several operators and equipment vendors worldwide. The detailed technical specification of air interface framework is expected by end of second quarter of 2007 and the Technological Evolution Framework (TEF) outlines the evolution strategy beyond the 2010 time frame.

5.Implications for India

These next generation broadband wireless standards are important for India, as it would enable broadband wireless to reach urban as well as rural India, pretty much like GSM / CDMA mobiles do so today. A broadband Wireless Consortium of India (BWCI) has been formed (see cewit.org.in) between operators, equipment manufacturers, component suppliers, academia and contribute the standardization efforts based on OFDMA technologies defined above. While these technologies would be available beginning 2009, the operators are starting to use Broadband corDECT in small towns and rural India. The next-generation wireless system would have the capability required to deliver broadband to all villages. The bit-pipes would be there. The challenge beyond would be use the bit-pipes to transform the rural economy.

References

1.Ashok Jhunjhunwala, Next Generation Wireless For Rural Areas, IJRSP, Special Issue on .Rural Wireless Communication., Vol.36, Number 3, June 2007.
2.Bhaskar Ramamurthi, Broadband wireless technology for rural India, IJRSP, Special Issue on .Rural Wireless Communication., Vol.36, Number 3, June 2007.
3.G. Venkatesh and Ashwin Ramachandra, Can 3G technologies benefit rural India?, IJRSP, Special Issue on .Rural Wireless Communication., Vol. 36, Number 3, June 2007.
4.http://www.dotindia.com/ntp/broadbandpolicy2004.html
5.IEEE Standard for Local and metropolitan area networks, Part 16: Air Interface for Fixed Broadband Wireless Access Systems, IEEE Std 802.16.-2004 (Revision of IEEE Std 802.16-2001).
6.Draft IEEE Standard for Local and metropolitan area networks "Part 16: Air Interface for Fixed and Mobile Broadband Wireless Access Systems, IEEE P802.16e/D5, September 2004.
7.corDECT, Wireless access system, Technical report of Midas Communication Technologies, Chennai, India, December 2000.
8.Paul Krishna, Varghese Anitha, Iyer Sridhar, Ramamurthi Bhaskar & Kumar Anurag, WiFiRe: Rural Area Broadband Access using the WiFi PHY and a new MAC, IEEE Commun Mag (USA), Jan 2007 (accepted).
9.http//www.3gpp.org/ftp/pcg/Beijing workshop presentation
10.www.3g-pp2/CDG Press Release (Aug 15, 2006)