India's mobile penetration, currently at 51%, is projected to reach 72% by 2016, with 367 mn mobile broadband connections. HSPA connections are projected to grow from today's 10 mn to more than 100 mn by 2014, making India the largest HSPA market worldwide. Bharti airtel has launched LTE services, and the other big players will soon offer 4G services, further fuelling the growth in data services at an unprecedented rate.
Why 3G Failed
While these numbers are encouraging, consumer adoption of 3G services has been poor. The top reasons are dismal speed and poor connectivity. The majority of consumers find little or no difference between the 3G and EDGE or GPRS experience. These are indicative of a backhaul bottleneck, which is usually the case wherever there is an explosion of data. With India's geographic and demographic diversity, the problem gets more complex and needs innovative solutions.
Earlier, RAN architectures could not support the exponential rise in bandwidth demands. Migration to all IP based packet transport networks is the only viable long-term solution. Within the IP backhaul space, the options are many but with no simple single solution.
Let's look at some specific geographic and demographic factors. The 3 largest cellular subscriber markets are China, India, and the USA.
What Do the Facts Say?
In India, as the numbers show, the market is very price sensitive and extremely competitive. In addition, India's population density has a wide variation from a low of approximately 150 per sq mile in the Andamans to 18,000 per sq mile in Delhi. At an average of 1,000 cell phone users per tower, this translates to roughly 1 tower per 6 sq miles in the Andamans and 18 towers per sq mile in Delhi.
Roughly, the data rates delivered by various cellular technologies are-GPRS-115 Kbps, EDGE-236 Kbps, 3G-0.2 Mbps, 3.5G-4 Mbps, and 4G-100 Mbps. With an over subscription ratio of 1:5, this translates to approximately 40 Mbps from each cell tower for 2G and 800 Mbps from each cell tower for 4G. Practically, the subscription ratio for 4G would have to be in the region of 1:3 or 1:2 considering the proliferation of applications and associated online requirements which mean 1.5 Gbps to 2 Gbps from each cell tower. This places a significant demand on the backhaul network.
The diverse geography of India includes large plains, hilly areas, densely inhabited metropolitan cities, smaller towns, rural villages, sparsely populated island states, desert regions, and forested nature reserves which are popular tourist locations, etc. Most of the tourist spots have an influx of subscribers for a few months of the year and a far lower subscriber density otherwise. Since the ARPU changes depending upon the geography and also depending upon the time of the year in some regions, the service providers have a substantial planning challenge.
The choice of a backhaul technology involves 2 different aspects of the technology-the media for transport and the method/encapsulation carrying payload. The predominant choices for the medium have been fiber, microwave, and satellite.
How Carrier Ethernet Becomes the Dominant Choice?
Currently, about 80% of backhaul in India is estimated to be supported over microwave. This works well for the legacy TDM based infrastructure where bandwidth demands are lower and the cost to start services is lower than that of laying out fiber. Voice services are still a major revenue component for Indian operators, but data usage is increasing especially among the urban subscriber base. A backhaul solution must manage both TDM and IP traffic. On the encapsulation side, the much lower cost and simplicity of ethernet compared to MPLS have made carrier ethernet the dominant choice.
This is the key to the current backhaul bottleneck. Microwave links are limited in the bandwidth they can deliver and are susceptible to varying climate and weather conditions. Carrier ethernet's scalability and QoS abilities are limited compared to MPLS. So, service providers grapple with serious questions about an expensive upgrade to the backhaul. Should microwave be gradually replaced with fiber? Should packet over ethernet give way to expensive MPLS services?
In light of this, it is encouraging to look at some innovations that may answer these questions
#1 Free Space Optics (FSO): FSO combines wireless and optical technologies by transmitting data over air using optical technology, as if over invisible lasers. It promises the bandwidth and fidelity of optical media minus the high fiber costs. Limited to a line of sight connection with a maximum range of 5 kms, this can be a viable choice for dense urban settlements.
#2 NLOS (Non Line of Sight) or nLOS (near Line of Sight) Solutions and Small Cells: Small cells facilitate setting up of non line of sight (NLOS) and near line of sight (nLOS) microwave backhaul networks, where point-to-point visibility between antennas is unclear or disturbed. Since this is capable of operating in the unlicensed band, cost can be further reduced. These are particularly useful in urban mountainous areas and regions with heavy monsoon showers or pollution.
#3 Satellite Based Backhaul: The technology and international products have been available for a few years now. We place this under innovations since this is not widely used for backhaul solutions in the Indian context. Satellite links do not have many of the problems associated with microwave such as line of sight restrictions or interference related problems or problems due to inclement weather. Since satellite bandwidth can be leased, it is an attractive option for seasonal variations in demand.
#4 10 Gbps Microwave: A recent claim from the CSIRO (Commonwealth Scientific and Industrial Research Organization), Australia's National Science Agency, this would be the world's fastest point-to-point microwave technology with a bandwidth of 10 Gbps in each direction and a range of 50 kms. If reports translate into actual field performance, this will be the best alternative to fiber links by providing high-speed backhaul connectivity with much lower costs, complexity, and higher flexibility.
#5 MPLS-TP: In simple terms, this is MPLS with the option of provisioning services, a toolset to monitor services end-to-end, and a standardized method of laying out and falling back on resilient paths. Doing away with the requirement for signaling and associated IP/routing, the simplification, and cost benefits are promising since it brings with it the scalability necessary to deliver a plethora of services with different QoS and service levels.
#6 Carrier Ethernet 2.0: With 3 powerful new features (multiple classes of service, interconnect, and manageability), CE 2.0 offers capabilities allowing operators to collaborate by interconnecting services, endorsing ethernet wholesale services, and mobile backhaul models. It lays down a set of attributes for backhaul connections, and helps telcos and equipment manufacturers to deliver the right set of features and services.
This offers an interesting array of choices, eg, free space optics or millimeter wave in densely inhabited metropolitan cities, existing microwave infrastructure for smaller towns and rural villages, satellite for sparsely populated island states, hilly areas and tourist locations, and high capacity microwave across desert regions. Making the right choices using the latest technology innovation allows an operator to manage cost judiciously to deliver expected connectivity and performance at the right price point.
Rajesh Kumar Sundararajan
The author is assistant vice president,
product management, Aricent Group
vadmail@cybermedia.co.in