Challenges Facing Front Ends for 3G and 4G Multimode Handsets
As consumers, we are used to the continual productivity enhancements derived from having ever-increasing, computer-like handsets at our disposal. For those of us lucky enough to have in-depth RF industry knowledge, we know full well the complexity of a system that can deliver this near desk-like performance in the palm of your hand. Still, it helps to step back now and again to consider how such an intricate system actually works in order to appreciate the arduous nature of developing these state-of-the-art mobile devices. Much like the awe we experience watching a jumbo jet take off— we may know how the basic interaction of forces occur, enabling the jet to break the sound barrier or cruise at an altitude of 30,000 feet, yet we’re still fascinated with how well it all works and how seamless designers have made it appear.
Similar to the jet analogy, engineers may know how these complex handsets, and the highly integrated systems they contain, function, but that knowledge doesn’t preclude us from feeling a sense of awe when watching someone use their handset to check e-mail, download a video, or take a picture. There are also many parallels to the technical achievements and boundaries being pushed in the crop of current third-generation (3G) and future fourth-generation (4G) multimode phones under development. This article examines some of the drivers that are pushing designers to seek new and groundbreaking solutions for a key system in these mobile devices—the RF front end.
Figure 1 Front end block diagram.
“RF front end” has become the industry-standard term for the radio frequency components functionally located between the wireless transceiver and the antenna, required to both transmit and receive the radio signal. More specifically, the entire cellular front end system is comprised of power and low noise amplifiers; filtering, such as receive filters and duplexers; RF switches, including antenna, mode and bypass; the antenna, and the interface to the transceiver and baseband (see Figure 1).
Market Forces Propelling Multimode Handset Development
Communication has evolved quickly in the last two to three decades. Consumers expect and demand ubiquitous “communication” throughout their mobile world. The industry has developed multiple solutions to satisfy this need for high-quality communications. No longer is simple “on-the-go” voice communication, which is delivered mainly through 2G handsets, acceptable. Consumers expect frequent access to a variety of data-intensive information, such as corporate databases, e-mail, multimedia messaging, real-time financial information and social networking, regardless of where they may be or what time of the day it is. You can see this theme of “a world where everyone is connected to each other, to information and to entertainment, in all places, at all times” reflected in company advertising, taglines and mission statements. This captures the true essence of the wireless industry’s main challenge over the next few years—to deliver a compelling ecosystem capable of a ubiquitous voice and data connection anywhere in the world. And, rightly so, this is the promise of the new 3G and future 4G networks—to deliver a better consumer experience with more accessibility and usability of the data content, which is both being generated and delivered.
Consumer Expectations Impacting Front End Development
In order for original equipment manufacturers (OEM) to satisfy these strong consumer requirements they must implement multimode systems that contribute toward a holistic handset solution. To gain a better sense of how comprehensively marketplace demands have affected 3G/4G multimode front end systems, one has to consider several key factors, including:
• Carrier spectrum necessitating multi-band requirements
• Handset complexity driving development resource requirements and other costs
• Technical issues specific to front end development
• Impact of system architecture choices on the handset battery life
How Carrier Spectrum is Shaping Multi-band Requirements
Next-generation multimode phones are designed to support up to 16 different frequency bands and more than 20 different band combinations. As the number of bands and band combinations grow, frequency flexibility at the platform level has increased in importance as a critical parameter for 3G handset development. This is a far cry from the single global frequency originally envisioned for 3G.
Frequency Spectrum and Regulation Issues
The motivation and benefits of a single, harmonized, global communication spectrum was one of the primary goals of the International Telecommunications Union (ITU) as they headed into the UN-sponsored global telecommunication meetings in 1992. The World Administrative Radio Conference 1992 ( WARC-92 band 1920 to 1980 MHz, 2110 to 2170 MHz UL/DL) was intended to offer a single band plan that could be used both in GSM-based 3G technologies (3GPP WCDMA) and CDMA-based 3G technologies (3GPP2 CDMA-2000). It was hoped that a single worldwide frequency would accelerate the adoption of new technology as well as reduce implementation and deployment costs. Services such as global roaming and the inherent economies of scale were to prevail.
Competing technologies and a lack of clear spectrum, due to previous frequency allocation in some countries, caused some significant changes in spectrum policy. The resulting compromises allowed for as much commonality in frequencies as could be agreed upon.
International Mobile Telecommunications-2000 (IMT-2000) was the global standard born from years of collaborative work between ITU and the mobile industry. The first frequency bands for IMT-2000 were identified at the World Administrative Radio Conference in 1992, with additional bands identified at the 2000 World Radiocommunication Conference (WRC-2000).
The harmonization in effect after the World Radio Telecommunication meeting in 2000 resulted in the new 3G spectrum in 700 to 900 MHz and 2600 MHz as well as modifications to the existing UMTS spectrum. This led to some regional specialization of band combinations, whereby phones would need to be customized to the frequency spectrum allocated for specific geographical regions, typically by covering one or two of these frequency bands. Current WCDMA deployments can be found in 2100, 1900, 1800, 1700, 900 and 850 MHz bands throughout the world.
As one can see in Table 1, there are multiple geographies calling for many different combinations of spectrum or radio frequency coverage. If we think about the increasingly desirable characteristic of global roaming, it is evident that the complexity required from the front end to be able to deal with all these frequencies and bands is accelerating at a fast clip. Mobile data networks must satisfy both local regulatory and physical transmission requirements.
Industry observers have noted the recent popularity of government-sanctioned auctions of frequency spectrum, usually for stipulated usage. Recent examples are AWS and vacated TV bands in the United States and Europe. US spectrum was auctioned off in 2007 for $19.6 B. Similar auctions are set to start in Europe for the 790 to 882 MHz bands, albeit with a delay in implementation, within a year’s time. Spectrum auctions have delivered billions of dollars back to government agencies and created strong motivation by global carriers to see a return on their investment.
Another important event indicating which future band expansion will be allowed was at the World Radiocommunication Conference 2007 (WRC ‘07). That conference recommended that the expansion of IMT (IMS) services be made available with the following band allocations:
• 136 MHz allocated globally for IMT (450 to 470, 790 to 806, 2300 to 2400 MHz)
• Expansion of vacated TV bands for mobile systems (698 to 790, 790 to 806, 806 to 862 MHz)
• Addition of 3400 to 3600 MHz for regions 1&3, 3400 to 3500 MHz for region 2
Drivers for Additional, Lower Frequency Bands
One trend that is also affecting the radio front end is the growing desire by carriers to re-use (and re-farm) currently available GSM frequencies, so that it can be allowed to propagate UMTS signals. One of the rationales behind the movement is the ability to optimize connectivity depending upon the population density and physical layout of the coverage area of interest. In highly congested, urban scenarios, higher frequencies could become more popular as they can be more spectrally efficient for high data rate applications. Alternatively, in less dense environments, operators may prefer to take advantage of the benefits of lower frequencies (700 to 900 MHz) that, in this case, enable a far greater coverage area for each base station as compared with its higher frequency compliment. In fact, this is the rationale for the interest in re-farming existing GSM-allocated frequencies (850 and 900 MHz). If operators are allowed to convert from GSM to WCDMA service in these lower frequency bands, there are estimates showing up to 60 percent cost reductions while increasing coverage by two-to-four times that for 2100 MHz.1
Figure 2 RFMD forecast of the number of 3G bands.
One can infer from the complex nature of this background on carrier spectrum that there will be many different varieties of cellular front ends required for the complete cellular system to work in a handset, whereby information sent on different frequencies are required to be transmitted and received from a single radio source. In fact, RFMD® internal projections, based on deployment rates and industry conversations, indicate a likely scenario where the number of WCDMA bands in handsets will be as many as five by 2010, with the rate of expansion for band combinations depicted in Figure 2.
Figure 3 Projected 3G multi-mode band combinations.
The dichotomy of the 3G is that the diverse frequency allocations worldwide have been a significant departure from the original intent of the standard. Hopefully you can gain an appreciation for the technical challenge laid out when we look at requirements coming in to support carrier smartphone deployments across the globe. This leads to the first of many challenges for front end suppliers-developing filtering and switching solutions for each of the bands. To date, spectrum homogenization has not been realized in either 3G or 4G front ends. To add further perspective to this issue, Figure 3 demonstrates the number of band combinations forecasted for the next five years. Since several of these bands will be used in combination for even the simplest 3G handset or data card, solutions must be developed in such a way as to allow suppliers to scale their offerings to accommodate the multiple bands.
Figure 4 Projected complexity of future phones.
To further complicate next-generation handset development, the evolution of the 3G standard has resulted in a relatively complex analog RF front end for these multimode, multi-band handsets. Depicted in Figure 4, as networks evolve, so must the devices that consumers use to access them. Handsets, therefore, have quickly moved from their humble voice-only beginnings, with a relatively straightforward radio design, to a device enabling real time, high data-rate communications through the use of multiple, complex transmit and receive paths. Technological advances, such as the use of MIMO and receive diversity in mobile devices, will enable higher mobile productivity. Handsets will act much more like multi-radio media devices by offering a continual linkage not only to cellular networks, but all IP-based networks through the use of data portals such as Wi-Fi and WiMAX. As these cellular standards evolve even further with the addition of HSPA+ and Long Term Evolution (LTE), consumers will continue to see additional data-rate improvements in 3G networks. This, the ultimate goal of 4G, will enable true mobile broadband access through a portable internet connection.
Next-generation Handset Platforms: More Resources, More Costs
With the addition of these extra bands, modes and associated verification, the resource requirements for next-generation platforms is much more extensive than it was for 2G cellular handsets. Multimode phones are designed to work with EDGE and WCDMA air interface standards as well as LTE for high-speed data and media transport applications and, perhaps, WiMAX mobile applications as well. Additionally, mobile operators require customized handsets to meet various consumer roaming needs. Hence, the handset OEMs, who must configure 3G handsets with multiple frequency bands and operating modes (WCDMA r99, HSDPA, HSDPA+, EDGE), are left with the daunting issue of rapid customization in order to meet both market timing and mobile operator requirements. All the while handset OEMs are seeking to integrate these advanced features and multiple band configurations into a size-reduced, platform-capable form factor. Lest we not forget the continual cost-reduction pressures these consumer devices are under, which further aggravates the dilemma of increasing content in a cost-sensitive marketplace.
In conclusion, to be able to support the different standards, applications, band combinations and mobile operator requirements, handset designs must now utilize very specific power amplifiers, filters, switches, duplexers and other RF components to implement these highly specialized multimode front end systems. Given the number and specificity of RF components required to complete the multimode front end system, in combination with the growing list of handset features, it should come as no surprise that engineers are requiring longer amounts of time to complete front end system designs. Furthermore, with this highly specialized nature of the front end system, testing and calibrating during the various phases of the handset development process have also lengthened considerably. All of these additional resources— whether they take the form of additional specialized components, additional testing, or increased engineering time—raise costs.
1. Motorola Whitepaper Deploying UMTS in 900 MHz Band, http://www.motorola.com/mot/doc/6/6796_MotDoc.pdf.