AUGUST 2012: This year, Bharti Airtel launched a ‘4G’ service in Kolkata and Bengaluru, making India one of the first nations to adopt the newest baby in the wireless world. 4G or IMT-Advanced is the fourth generation in telecommunications, characterised by reliable lightning-speed broadband wireless access (BWA)—a dream come true for not just the young and raging population but also businesses, media agencies, governments and many more stakeholders. It is hoped that 4G will catalyse national development, as it means wireless communications to the nooks and corners, even across rugged terrains and the poorest of villages.
With the launch of Bharti Airtel’s service, umpteen questions have arisen in the minds of consumers across the country—what really is 4G, what sets it apart as a new generation, is Bharti Airtel’s service (based on the TDD-LTE standard) really 4G, what are the competing 4G standards, how is 4G doing worldwide, what are the pros and cons of this generation, and what is next in the pecking order? Here we try to answer these questions.
The fourth generation has just begun
Some service providers in countries like South Korea and Scandinavia started terming their service as 4G as early as 2006. Others started doing so in the US a few years later. Humbug and marketing tricks!
4G is a very nascent technology and there are very few deployments worldwide that qualify as 4G today—even by a very liberal definition. If we keep to the true definition of IMT-Advanced, 4G is not even here yet.
“In the mid 2000s, as 3G systems were being deployed, there was the initial definition of 4G as providing 100Mbps bandwidth. Subsequent to 2006, many systems which exceeded 40-50 Mbps started using ‘4G’ as a marketing ploy for their networks and some providers called their advanced 3G networks as ‘4G Lite’ with ‘Lite’ written in small print. This included WiMAX and HSPA+ systems. There were even legal litigations in the US courts by operators against each other for false advertising. It is interesting to note that the early versions of Long Term Evolution (LTE) were being dubbed as 3.99G in published literature. ITU finally provided a more definitive definition of 4G,” explains Dr Suresh Borkar, a member of faculty at the Illinois Institute of Technology, Chicago. Dr Borkar has wide consulting experience in commercial and public safety systems, telecom strategy formulation, spectrum and regulatory policies, and dynamic spectrum management.
In addition to general specifications on inter-working, handovers and quality of service, the key attributes defined by the ITU included an all-IP packet core and (downlink) bandwidth of 100 Mbps for mobile applications and 1 Gbps for nomadic and quasi-stationary applications. Taking into account the estimates and techniques like carrier aggregation and wider spectrum, there is now a general consensus in the telecommunications community that 3GPP’s LTE-Advanced (Release 10) and WiMAX 2.0 (IEEE 802.16m) can be classified as 4G systems.
LTE vs WiMax—A no-brainer!
The third generation of mobile communications is characterised mainly by HSPA+ and EV-DO. Similarly, two technologies qualify—technically—to be called as 4G. These are 3GPP’s LTE-Advanced and IEEE’s 802.16m (WiMAX 2.0). However, the race seems to be a no-brainer. With WiMAX 2.0 failing to take off, it looks like LTE is going to dominate the 4G world.
“WiMAX 802.16m has not been commercially successful. Hence deployment of 4G will be based on LTE-Advanced technology. These implementations are in a nascent stage worldwide. Large-scale deployment of 4G will take place only next year,” says Dr Abhay Karandikar, professor and head-Department of Electrical Engineering, IIT Bombay. IIT Bombay has made several contributions to IEEE 802.16m 4G standards, including bandwidth reservations, quality of service (QoS), relay, etc.
Experts agree that LTE-Advanced (LTE Releases 10 and 11) is the main technology on the 4G horizon. It is being standardised to use up to 100MHz channel bandwidth to achieve up to 1Gbps downlink throughput.
“In our opinion, it is okay to consider LTE (Releases 8 and 9) also to be 4G technology as it makes a radical breakthrough by adopting multi-carrier orthogonal frequency-division multiplexing (OFDM) technology. LTE has retained some level of backward compatibility to benefit from the UMTS experience and coexist with UMTS,” says Harpinder S. Matharu, senior product marketing manager-Communications Division, Xilinx.
LTE technology is a completely packet-based protocol and radio access technology. Its core network, leveraging flexible bandwidth, achieves much better spectral efficiency (bps/Hz) and delivers a throughput of greater than 100 Mbps (for 20MHz FDD system using 2×2 spatial multiplexing) by using higher-order modulation and multiple-input multiple-output (downlink and uplink).
Matharu adds, “In light of the above argument, niche deployments of 20MHz WiMAX cellular systems should also be considered 4G. However, LTE framing structure is better for achieving a higher throughput and lower latency, besides the benefits of a more power-efficient user terminal uplink technology. We expect LTE to replace majority of the WiMAX deployments in the coming years.”
Dr Borkar concurs, “The initial LTE-Advanced systems are expected to deploy by the year-end. Whereas WiMAX may have its niche applications like backhaul, from access viewpoint, LTE will become the norm of the future. Even the current 3GPP2 US standard based 3G operators like Verizon have abandoned the CDMA2000 evolutionary path and embraced the 3GPP LTE 4G standard.”
The flavours of LTE—TDD and FDD
“Globally, there is one harmonised standard for LTE that encompasses two modes—TDD and FDD,” remarks Dr Lakshminath Dondeti, director, engineering-technical standards, Qualcomm India.
LTE involves duplexing of uplink and downlink radio signals. The duplexing scheme is responsible for managing simultaneous transmission and reception without interference between the two. To handle this, signals for transmission and reception can be separated using either time or frequency—the former is called time-division duplexing (TDD), while the latter is frequency-division duplexing (FDD).
FDD uses a paired set of frequency blocks—one for downlink transmission and the other for uplink transmission. TDD works on only one frequency block, but uses short alternating bursts of transmission and reception.
FDD and TDD are different technologies. FDD is better suited for voice, while TDD is better suited for data. In data, there is usually more of reception or download than upload, and the system can be tweaked to allow greater download time and shorter upload time.
“Most deployments of cellular systems including LTE use FDD operation because of its simplicity of transceiver designs. Some countries like India (Bharti Airtel) and China (Chine Mobile) have selected TDD for the 4G LTE network because of its capabilities of dynamic adaptation of downlink and uplink bandwidths. WiMAX deployments from their early version have embraced TDD operation,” explains Dr Borkar.
“In India, LTE TDD has been deployed initially in the 2.3GHz spectrum band. The Global TD-LTE Initiative (GTI) has been formed to support closer network and device integration of LTE TDD and FDD. It is driven by several major wireless operators from around the world, working closely with GSMA and NGMN. Qualcomm has multimode chipsets that integrate both LTE TDD and LTE FDD, and inter-work with 3G HSPA and EV-DO,” says Dr Dondeti.
IP-based architecture—pros and cons
“Circuit-switched networking is inefficient for data traffic. With Internet data constituting a significant fraction of mobile traffic, it is more efficient to deploy packet-switched IP technology. IP also enables seamless network convergence. There are really no disadvantages of IP or packet switch, except perhaps that network security issues need to be more tightly managed in an IP network,” says Dr Karandikar, who is emphatically in favour of packet-switching.
Higher variation in the instantaneous power (due to OFDM-based architecture) that impacts power amplifier efficiency
Incompatibility with previous generations
Over 40 frequency bands make global roaming difficult
Poor voice support
High cost of embracing LTE, including high auction costs for acquiring spectrum, lack of economy of scale to make available reasonably-priced user devices, and putting together the backhaul infrastructure
Need to set up newer operations and management systems, including provisioning, performance optimisation, revenue realisation, fault handling and maintenance, security management, etc
An all-IP network can very effectively handle multiple applications and services, e.g., voice, video and data. The power of a unified IP multimedia system architecture can be exploited to efficiently support the QoS requirements of such varied applications and services. Packet transport provides much better resource utilisation than circuit-switched network.
However, all-IP systems have a flipside too. “The primary disadvantages for all-IP systems are how to recover the huge investments in legacy 2G and 3G systems, not stranding the vast number of existing user devices in the marketplace, and inability to provide the high bar of voice reliability and quality. The VoIP alternative is primarily statistical in nature and lacks the advantages of traditional circuit-switched based voice, e.g., quality, deterministic behaviour in terms of latency and routing, and advanced supplementary services,” says Dr Borkar.
Since 4G LTE has an all-IP core (enhanced packet core), it cannot directly support voice applications until voice-over-IP (VoIP) becomes the norm. Moreover, all operators have already invested heavily in 2G and 3G systems which use circuit-switching for voice support.
2. Countries like India, China, Japan and Brazil have deployed LTE (Releases 8 and 9) in recent months. Many players in the telecom industry tend to accept this as 4G as it uses multi-carrier OFDM technology and can be considered a major jump in technology and user/operator experience.
The general approach that current LTE operators are using is circuit switch fallback to a 2G or 3G network via the memory management entity in the enhanced packet core. This is transparent to the user device. In addition to the circuit switch fallback, the 4G user devices are designed to be multi-mode so that they can connect to a 2G or 3G system in case the LTE system is not present or the LTE signals are too weak.
“The 4G technology evolved and standardised by 3GPP has multiple options of supporting voice. One is circuit-switched voice. LTE supports circuit switch fallback to 3G/2G network for voice support. Advantages are definitely there for multi-service operators like Airtel, as it can offer varied services and bundle service plans to meet the requirements of the customers,” says Rajiv Rajgopal, CEO-broadband/data, Bharti Airtel.
Looking into the future, David Maidment, mobile segment manager, ARM, comments: “Moving forward, operators will want to break their dependency on circuit-switched services and move to an all-LTE deployment. As such, they are working to standardise and deploy voice-over-LTE (VoLTE), which allows voice to be carried over the packet network. VoLTE handsets are already available, and in the next one or two years we will begin to see the growth of these services.”
Availability of 4G devices
True 4G equipment are not widely available today. LTE Release 8 systems have been deployed since 2010, but these are not 4G—merely 3.5G. As is expected in a family of standards like LTE, device and infrastructure manufacturers will provide migration and compatibility between different releases. Current LTE smartphones and other devices will operate in future LTE networks but may not be able to exploit the advanced and new features like carrier aggregation. Real 4G devices compatible with LTE-Advanced specifications may be available by the year-end.
“With the current technologies, many advanced features can be effected in the devices via provisioning and software updates,” points out Dr Borkar.
“Initially, 4G user equipment targeted notebooks and tablets for data services. With the availability of chipsets that support both 3G and 4G technologies, the market is beginning to see 4G LTE smartphones,” adds Matharu of Xilinx—a company that provides silicon devices, intellectual property solutions and design services for wireless infrastructure.
Rajgopal notes: “Bharti Airtel through its GTI, along with other partners like CMCC and Softbank, is taking a lead in ensuring that the device ecosystem rapidly matures to the operators’ and consumers’ requirements. Currently, we have an array of compatible devices like dongles and CPE devices that support TDD-based LTE. Smartphone and tablet users can access 4G LTE services over Wi-Fi networks, from CPE devices. We are working with the OEMs in ramping up the device ecosystem in the country.”
That said, there is also the issue of devices made for one region not working in others. To this, Rajgopal responds, “At present, 4G LTE devices available in the open market or USA are not compatible in India. We are working with OEMs and chipset vendors to ensure that this will not be an impediment. As a matter of fact, Bharti Airtel is one of the founding members of the GTI that is working on standards for TD LTE.”
Frequency bands and duplex operation are the two primary reasons why equipment used in one continent are not compatible in others.
“LTE as a standard can be deployed in over 40 different frequency bands. This is a benefit as it makes the standard flexible, but it’s also a weakness from a fragmentation and roaming point of view. We have already seen issues around band support in LTE with current-generation devices that are unable to roam outside their home networks. Multi-band support is nothing new in cellular applications. From 2G onwards equipment vendors have been producing multi-band phones. The issue with LTE has been the sheer number of bands as well as the fact that it supports two duplexing methods—FDD and TDD,” explains Maidment.
“From a signal processing point of view it is already possible to implement a ‘world phone’ that can span FDD and TDD. The issue is in the radio hardware, which needs to be band-specific in order to receive and transmit at the required frequency. We expect to see innovation in this space as manufacturers work to solve this issue and produce programmable band selection in the hardware that will, in turn, allow devices to be produced that cover all the major bands worldwide,” he adds.
“During the early stages of LTE growth, 3G HSPA or EV-DO will be the primary global roaming technologies given the global harmonisation of 3G spectrum. Over time, some common LTE bands across the world that facilitate international roaming will start to emerge. In fact, 3G and LTE will inter-work and coexist for a long time,” says Dr Dondeti.
Facilitating international roaming is not a great problem, according to Dr Karandikar. “ITU has defined band plans for various technologies and standards. Various countries do harmonise their spectrum allocation accordingly. ITU holds World Radio Communications Conference every three to four years where frequency harmonisation may be achieved,” he informs.
Global roaming issues have been solved in the past using multi-standard chipsets and multi-band radios. A similar solution might be helpful here too. Qualcomm has developed the world’s first multi-mode (HSPA, EV-DO, FDD/TDD LTE) chipsets that integrate both TDD and FDD. The commercial LTE TDD network launched by Bharti Airtel uses multi-mode dongles based on Qualcomm’s MDM 9×00.
“Huawei, ZTE, BandRich and Quanta have announced LTE TDD multi-mode devices based on the Qualcomm MDM9x00 chipset in August 2011, indicating commercial availability this year,” adds Dr Dondeti.
When will 4G phase out 3G?
“Standards are quick to arrive and slow to leave. We have lived with 2G services for around 20 years now, for example. The question of when 4G will phase out 3G should instead be phrased “when will 4G phase out 2G”,” quips Maidment.
Today, in most territories, 2G and 3G services run side by side with 2G often covering less-densely populated regions and 3G services deployed in towns and cities. From a network perspective, we are seeing a growing trend towards heterogeneous network architectures which combine multiple carriers with varying access technologies—such as small cells to deliver capacity and coverage where it is needed most, and larger, more traditional macro cells providing wide-area coverage in rural areas.
As LTE is deployed both in the wide area through macro cells as well as in small cells, we expect to see operators looking to free up their existing 2G spectrum to move across to LTE services. This is a slow process which is highly dependent upon regulatory approval and will differ from region to region.
While some operators might consider leapfrogging directly from 2G to 4G, those who have deployed 3G are likely to continue enhancing and using those networks for at least five to ten years to get justifiable returns on their investments.
We can expect to see the continued adoption and rapid uptake of LTE services around the world. As a key part of that, we will see new low-power, multi-band capable devices that deliver the economies of scale required to make LTE commercially viable. From a standards perspective, device vendors are working to implement the first-generation LTE-Advanced terminals, which we would probably see a year or two down the line. These devices will be capable of supporting data rates up to 300 Mbps enabled by carrier aggregation and multi-layer multiple-input multiple-output.
“In addition to LTE-Advanced, there are plans within 3GPP to look at how LTE can be employed to enable machine-type communications. The rise of machine-to-machine communications is anticipated to be a key growth area and it is essential that we have in place the right standards to support that need. Being an all-IP based technology, LTE is well positioned to serve this emerging space during the next 20+ years,” says Maidment.
While 4G, according to its true definition, is not even here yet, we are already talking of what is yonder, beyond the big blue mountain.
Dr Borkar provides some futuristic food for thought: “Consistent with the general historical trend of a new technology standard every ten years, it is expected that ‘5G’ specifications will likely be in place in the 2018-20 timeframe.”
The framework for 5G includes higher-efficiency operation with lower battery consumption, higher system reliability, more uniform, high data rates across the coverage area, low infrastructure deployment costs, and higher spectral efficiency and capacity.
Some of the features being considered include advanced multi-cell coordination for enhanced system performance and end-user service quality, efficient infrastructure for machine-to-machine communications to support the vast number of connected devices, direct user device-to-device communications especially for performance and network outage situations, and new ways of using and adapting to the available spectrum based on the use of cognitive radios.
The author is a technically-qualified freelance writer, editor and hands-on mom based in Chennai