Sorting out 4G: Are we there yet?
Janine Love, Contributing Editor - January 31, 2012
Thanks to clever marketing campaigns and branding, the general public believes that 4G has arrived. The performance of 4G/LTE mobile handsets has improved, and designers are pushing the outer limits of performance in all parts of the handset to inch closer to the rates that the IMT (International Mobile Telecommunications)-Advanced standard dictates. In many cases, the limitation is available spectrum, so designers must develop creative approaches until governments act, preferably by making contiguous spectra available.
AT A GLANCE
Riding the leading edge of the possible, engineers are turning to new technologies, materials, and techniques, and they are taxing the upper limits of their simulation and test-and-measurement tools. On the bright side, all of the engineering communities are working together to squeeze the most performance from low-power handsets and the most diverse wireless networks. With users downloading more than 1 billion apps a month and sending 8 trillion texts in 2011, the surge in mobile usage will need all the support it can get.1
The ITU-R (International Telecommunications Union-Radio Communication Sector) manages IMT-Advanced, whose features enable worldwide usability. The standard's targeted data rates of 100 Mbps for high mobility and 1 Gbps for stationary use get the most attention, however.2 For the physical-layer requirements, IMT-Advanced calls for a bandwidth exceeding 40 MHz, peak downlink spectral efficiency of better than 15 bps/Hz and uplink efficiency of 6.75 bps/Hz, and latency of less than 100 msec for users and 10 msec for control.
Although the latest commercial LTE, HSPA (high-speed-packet-access)+ and EVDO (evolution-data-optimized) Revision B services provide higher data throughputs than do traditional wireless services, they still do not provide a high enough data rate to meet IMT-Advanced requirements. "A wider bandwidth and as much as 4×4 [four antennas at the transmitter and four antennas at the receiver] MIMO [multiple input/multiple output] for the downlink are essential for achieving the IMT-Advanced requirement," says Jung-ik Suh, wireless-marketing-program lead at Agilent Technologies.
However, researchers are aiming for a 100-MHz bandwidth with as much as 8×8 MIMO for the downlink. Unfortunately, wireless-service providers will be unlikely to achieve this goal in real-world conditions.
One approach to the bandwidth problem is to use carrier aggregation, which requires the deployment of parallel receivers. "There are still many other technical hurdles to overcome to reach IMT-Advanced," says Suh. "It is hard to say if we are now at either true-4G performance or [at] LTE-Advanced, but lots of technical research and efforts will keep closing the gap to reach IMT-Advanced."
Developers of the 3GPP (third-generation partnership project) in 1998 created the global initiative to develop specifications to satisfy the latest mobile standards. To address the IMT-Advanced requirements, 3GPP has introduced LTE-A (LTE-Advanced). Initially part of Release 10 of the 3GPP specifications, LTE-A is the new target for mobile developers.3
"A main advantage of LTE-A over any other proposal is the backward compatibility with LTE Release 8," says Markus Willems, PhD, senior manager of product-marketing system-level solutions at Synopsys. "[LTE-A's] being an evolution rather than a revolution [should] result in [its] widespread adoption."
According to Chris Pearson, president of 4G Americas, a wireless-industry trade association representing the 3GPP family of technologies, some carriers will deploy LTE-A in 2013. In the meantime, he notes, HSPA+ and LTE now include tremendous technical advancements, providing subscribers with fast mobile. There are now 173 HSPA+ deployments in 86 countries, with peak speeds ranging from 21 to 42 Mbps; 39 operators in 25 countries commercially deploy LTE, which has more than 250 commitments for deployment in the coming years.
For Pearson, the major hurdles for LTE-A are not of a technical nature. He notes, however, that the mobile-broadband industry is growing at a fast rate and that customers are using large amounts of data (figures 1 and 2). "There is a major concern and opportunity for governments throughout the Americas to act soon to plan, reserve, allocate, and auction additional internationally harmonized spectra to the industry," he says. "The 3GPP standards will offer substantial gains in spectral efficiency as they evolve; however, the current rate of mobile-data growth demands additional spectrum resources globally."
Figure 2: Traffic growth has exploded as manufacturers develop new mobile terminals. In 2000 (a), traffic included 50-kbps GPRS (general packet-radio service) and RTT (radio-transmission technology. By 2010, with approximately 500 million deployed smartphones, traffic had moved to 5- to 7-Mbps HSPA+ and EVDO (b). By 2020, traffic will move at 1 Gbps (c), with LTE-A, 4 billion new smartphones, 10 billion new smart devices, millions of new apps, and video- and cloud-based services (courtesy 4G Americas).
Defining the challenges
Technical challenges remain for achieving the performance that the IMT-Advanced standard outlines. According to Madan Jagernauth, vice president of mobile broadband for Huawei Technologies USA, those challenges include the use of multiple antennas in smartphones. Consumers' desire for larger screens is driving bigger form factors, however, which eases this challenge.
However, when deploying MIMO in the device side of the link, form factor is not the only concern. According to Jagernauth, manufacturers will not deploy 4×4 transmission in smartphones because of its drain on battery life. He does, however, expect to see virtual MIMO, in which two subscribers can transmit simultaneously on the same time slot and in which the base station can decode it. This approach enables higher network usage. Huawei provides equipment on both the infrastructure and the terminal sides of the link, including a number of Android devices.
According to Synopsys' Willems, when it comes to chip design, the main hurdles are in the RF part. "Carrier aggregation as a key concept requires running multiple transceivers in parallel." He says. Balancing area, power, and performance is a key challenge." Synopsys offers the LTE-A library with a simulation setup that allows users to run all the tests in the 3GPP standard documents, thereby serving as an executable specification. The product is currently undergoing verification against equipment from Rohde & Schwarz.
Although manufacturers are on the verge of deploying LTE-based products, a lot of work still remains, according to Manfred Schlett, vice president of marketing at RMC (Renesas Mobile Corp), a division that Renesas formed after it acquired Nokia's wireless-modem business. Despite the availability of chip sets from multiple vendors, both power consumption and costs are high.
Schlett sees the chip sets, definitions, and band allocations in the market as immature. Band fragmentation, which makes it challenging to define chip sets, further complicates these issues. RMC is now launching a dual-mode HSPA+/LTE platform, and it plans to launch LTE-focused parts this spring (Figure 3).
Figure 3: Renesas Mobile Corp has released the SP2531 integrated triple-mode LTE slim modem.
Power efficiency is crucial for the modem because of the need for more operations due to MIMO schemes. Selecting smart algorithms for the receiver part directly affects power efficiency, according to Synopsys' Willems. "There is a dedicated need to explore algorithm alternatives that allow the system to stay close to the optimum maximum in throughput but that are less computationally intensive," he says. According to Willems, using software-defined approaches with multiple DSP cores is becoming the norm because it allows the reuse of functional units across a variety of standards.
"One of the key [remaining technical hurdles] is achieving capacity gains to users that are farther from the cell tower," says YJ Kim, general manager of the infrastructure-processor group at Cavium Networks. "A heterogeneous network consisting of macrocells and small cells somewhat solves this [problem] at a higher level. Once you peel the onion, however, there is the issue of interference. The base stations need to be capable of achieving error-free high throughput to all users that demand it."
He says that sophisticated interference-cancellation algorithms, which demand high processing power, can achieve that throughput. This throughput is important in both macrocells and small cell base stations. Effectively built multicore-base-station SOCs enable these approaches without consuming much power, he says. Cavium is addressing the needs of 4G with its multicore Octeon processors, which target power management.
Meeting the challenges
Systems implementing 4G involve a multitude of design challenges, including dealing with multiple channel bandwidths, transmission schemes for uplink and downlink, frequency- and time-domain transmission modes, battery-life management, and backward compatibility with 2 and 3G technologies. To address the limited-bandwidth issues, designers have developed carrier aggregation and MIMO techniques, but these technologies, in turn, bring their own challenges.
With the possibility of the emergence of 20 LTE bands worldwide, the issue of coexistence also arises. Coexistence gives rise to a number of issues, including board area, antenna design, sensitivity, configurability, and cost, according to RMC's Schlett. In response, RMC has designed its RF platform to support seven LTE, five HSPA, and four GSM bands. These products underscore the need for design-simulation tools and robust virtual prototyping for analyzing the real-time interaction between the protocol stack and the modem.
Carrier aggregation requires developers to pay more attention to control channels, cell edge, power management, spur management, and self-blocking, whereas MIMO design presents interference issues, especially on the terminal side, according to Agilent's Suh. On the user-equipment side of the link, designers are considering multistandard radios, which can help reduce the number of RF components by supporting multiple wireless technologies on one chip.
However, unlike with previous generations, 4G designs have many complex considerations, so designers have more issues to address before moving to a prototype, according to Suh. He points to tools such as Agilent's SystemVue, which can help to overcome the challenges during 4G user-equipment design before the specifications are defined. The tool also offers functional tests, such as a battery-drain test that can help designers solve battery issues using the terminal's status of sleep, standby, or call.
Power-consumption issues also lead to heat-dissipation problems, and the issue for 4G stems primarily from the need for multiple CPU cores in the RF section and the need to thermally manage the baseband chip. Some designers are addressing the power issue by using new process technologies. For instance, process technologies are now at 28 nm, with 20 nm on the horizon. As they design new architectures, developers are finding that traditional spreadsheet-based analysis is no longer enough because it lacks sufficient support for dynamic-use case scenarios, according to Willems. Designers need simulation-based analysis of what-if scenarios.
Willems says that hardware- and software-development engineers are moving to prototyping as early as possible, and virtual prototyping, which gained favor with 3G design, has become the mainstream approach as systems advance. High-performance simulation is also critical for algorithm design in modems, which allows designers to explore alternatives and maintain standards compliance. Synopsys offers simulation and prototyping tools from IP (intellectual property) and architecture to chip implementation (Figure 4).
Figure 4: The Synopsys LTE-A physical-layer-simulation library allows users to run tests that the 3GPP standard documents specify.
Although much work remains, the path to true 4G performance seems achievable, and the popularity of smartphones and tablets is creating a huge market demand for improvements. In the meantime, the wireless industry continues to take defined steps toward the goals of IMT-Advanced, leaning on the improvements from HSPA+ and evolving LTE to LTE-A.
Although a disconnection remains between the theoretical, such as 8×8 MIMO, and the achievable, today's wireless networks have significantly higher performance than their predecessors. Look for dramatic improvements over the next two years. To address the spectrum crunch, you are likely to continue to see interest in heterogeneous networks combining macrocell and microcell deployments. Until blocks of contiguous spectra become available, however, you can expect to see more creative design approaches to boosting data rates and maintaining battery life in next-generation mobile phones.
1. Love, Janine, "2011: the year mobile took over the world," EE Times RF & Microwave Designline.
2 "ITU global standard for international mobile telecommunications 'IMT-Advanced'."
3 Rumney, Moray, "Introducing LTE-Advanced," EE Times RF and Microwave DesignLine, Feb 5, 2011.
Janine Love is senior editor, Test & Measurement World, and managing editor, Test & Measurement Designline and RF & Microwave Designline. She can be reached at email@example.com.
Share your thoughts.
EDN.com Communications Network Design Center Player
Datasheets.com Parts Search
185 million searchable parts
(please enter a part number or hit search to begin)