The Future of LTE in India

Broadband access to the Internet has the potential to substantially improve a country's GDP and to enable its population to connect with each other even when they are away from each other physically either temporarily or in the longer term. Wireline penetration has been historically low in India and shows no prospects of improving in the near term. In the world of voice telephony, the broad penetration of mobile phones has obviated the need for wireline phone connections. Similarly, mobile broadband access is highly likely to be the predominant path to the Internet for hundreds of millions of Indians. There exists a pent-up demand for mobile broadband access in the country, illustrated by the fast uptake of EV-DO dongles, more commonly known as Tata Photon Plus and Reliance Netconnect Plus.

After the 3G and BWA auctions and with the almost complete launches in the 3G 2.1GHz spectrum, and the impending LTE launches in the 2.3GHz BWA spectrum, the Indian population is at the cusp of having multiple mobile broadband access options. This is the beginning of a great time for Internet users in India. Quick growth in mobile broadband uptake may also present serious challenges to network capacity, AT&T's network woes being the most prominent example. As it turns out, successive auctions of 3G and BWA spectrum enabling HSPA and LTE based mobile broadband networks in India may sufficiently support that growth for a few years to come. This is subject to backhaul and backbone capacity of course, but that is a different topic altogether.

3G and LTE Evolution

3G EV-DO networks in India are in the 850 MHz band of spectrum and 3G HSPA networks are in the 2.1GHz band of spectrum, whereas LTE networks will be in the 2.3GHz band of spectrum. Radio propagation characteristics imply that the 3G networks in India would provide broad coverage while LTE networks in central business districts and other population centers would address capacity needs. Next, for voice and SMS services, circuit-switched technologies, specifically CDMA2000 1X, GSM and UMTS R99 would be the most reliable, efficient and affordable, for a long while to come. We are looking at multiple technology evolution paths owing to legacy deployments, technologies that are most appropriate for available bands and bandwidths of spectrum, and global synergies contributing to availability of affordable devices.

Click on the image to enlarge

Why Talk about 3G, When the Topic is LTE?

Now it is fair to ask, “Why must we talk about the old when we want to look towards the new?” As it turns out, so far as wireless technology life cycles go, 3G technologies are very much in their early stages. In India, most 3G networks are less than a few weeks old. Even elsewhere in the world, where they have been around for a few years, they are still in the early stages of their life. Next, in narrow bands of spectrum, 1.23 MHz carriers in case of EV-DO and 5 MHz carriers in case of WCDMA or HSPA, the currently deployed technologies in India to be precise-EV-DO and HSPA-are the most efficient technologies. LTE begins to become more efficient at 10 MHz FDD carriers and 20 MHz TDD carriers, the latter applying to the BWA spectrum in India. Next, EV-DO in 850 MHz and HSPA in 2.1 GHz spectrum are the most efficient for coverage (3G coverage in the figure), whereas LTE in 2.3 GHz is best suited for higher capacity needs in densely populated areas or of high bandwidth applications. In summary, 3G and LTE networks working in concert with each other best meet India's mobile broadband needs.

The key aspect here is that these technologies need to work in concert with each other. Specifically, as users move away from LTE networks deployed in areas of high bandwidth usage, data sessions on their mobile devices need to automatically move to 3G HSPA or EV-DO networks; likewise, when they move into high bandwidth usage areas with a multimode mobile device (eg, smartphone, tablet, dongle connected to a laptop, supporting 2G, 3G, and LTE technologies, FDD and TDD modes and multiple bands) their HSPA or EV-DO sessions may seamlessly and automatically (may be configurable by the operator or even the user) move to the higher bandwidth and lower latency LTE network. To summarize, multimode (multi-technology) devices operating in 3G+LTE networks get us close to mobile broadband nirvana, as early as this year in India. For the first time in its telecom history, India will be second to none.

What does LTE Bring to the Table?

LTE offers higher bandwidths and lower latencies enabling such diverse applications as telemedicine and multi-player gaming. Note that HSPA+ enhancements come pretty close to the type of bandwidths LTE may support, when operating in the same amount of spectrum. There has also been significant progress toward better latencies in HSPA+ and HSPA+ Advanced. LTE networks that we expect to see in 2011-12 in India will have less than a third of the latency of the 3G networks in India.

The BWA spectrum in India is a single block of 20 MHz of spectrum, which implies the same spectrum shall be used for sending or receiving data; in other words it is TDD spectrum. In contrast, 2.1 GHz FDD 3G spectrum assignments comprise a 5 MHz component for sending and a separate 5 MHz component for receiving wireless transmissions. TDD spectrum has certain advantages and some shortcomings too. Operators may choose to use the spectrum in a downlink (users receiving data) or uplink-intensive mode, depending on the nature of popular applications in their networks. For example, if most users download a lot of video clips frequently but upload only occasionally, operators may use the spectrum in a configuration that favors downlink transmissions. FDD spectrum allows no such flexibility; in our example, the uplink portion of the spectrum is effectively wasted. There is a cost however in TDD mode wherein there needs to be a guard period between downlink and uplink transmissions.

For several years to come, CDMA2000 1X, GSM and UMTS R99 voice would be most economical and reliable and hence LTE devices would support the capability to fall back to circuit switched voice (widely known as Circuit Switched Fall Back or CSFB). In essence, a device sending and receiving data via an LTE network would switch to a 3G network, for instance, HSPA for data and UMTS R99 for voice when the user makes or receives a voice call or SMS data. For a vast majority of applications this switch will not be noticeable to the user. In due time, operators may choose to deploy IMS voice services allowing users simultaneously maintain LTE data sessions and make and receive voice calls or utilize various IMS services.

Enhanced multicast broadcast and multicast services (eMBMS) is another area where Indian operators may have set their sights to offer a wider variety of video delivery services to large groups of users simultaneously. When many users in close proximity want to subscribe to the same or a limited set of content streams, it is efficient to deliver those streams via eMBMS. If there is a critical mass need for eMBMS in India, we may see it in the next couple of years. For end users, this should be seamless. However, the key here is even when a large number of users are active in an area consuming these content streams, other users may access other services without noticing any service degradation.

LTE Advanced, an IMT-advanced technology, and therefore, 4G in the true meaning of that term is close to completion at 3GPP. It supports lower latency than LTE and more importantly offers the ability to aggregate multiple 20 MHz carriers, all the way up to 100 MHz to provide speeds up to 1 Gbps. That kind of spectrum may not be available to a single operator in India, so we may not see such speeds realistically. However LTE-advanced supports other enhancements including higher order MIMO (multiple antenna technology; 8x8 DL and 4x4 UL), range expansion, heterogenous networks, with advanced interference management, effectively improving capacity in larger coverage areas. Data networks may have different cellular coverage areas experiencing different load levels at different times of day; active load balancing between cells improves data throughputs for all users in those cell coverage areas.

It is safe to say that 3G EV-DO and HSPA networks in concert with LTE, with their parallel evolution paths will more than meet India's mobile broadband needs from a technology perspective. Multimode devices will offer seamless user experience with data session continuity between 3G and LTE, and circuit switched fallback for voice.

Lakshminath Dondeti
The author is
director, engineering,
Qualcomm India
vadmail@cybermedia.co.in