Get wireless design headline news and products - Sign up now!
SAWLESS RF Transceiver Technology Meets the Multiband, Multimode Requirements for 4G/3G/2G
By Vivek Bhan, Fujitsu Semiconductor
By the end of this year, enhanced products that utilize the Long-Term Evolution (LTE) standard will become available, providing broadband speeds to the handset and enabling the use of Internet services outdoors that are equivalent to indoor broadband services. LTE will offer the ability to deliver Internet services more quickly and efficiently than ever before. For network operators, LTE is expected to lower operating costs and increase spectral efficiency. LTE will be adopted on a global basis. ABI Research predicts that, by 2013, more than two-thirds of handsets will be capable of 3G, 3G+ or 4G communications. Of these, as many as 32 million will be provided to LTE subscribers.
Figure 1. The MB86L10A transceiver chip designed for multimode, multiband LTE, UMTS and EDGE mobile handsets.
|
The transition from current 2G and 3G communications standards will be gradual, and developing RF transceiver technology that is cost- and power-efficient is a complex challenge. Many of the first handsets supporting LTE will also need to support existing 2G and 3G capabilities. This means that the silicon must support multiple standards, air interfaces and radio technologies for backward compatibility, spanning a range between 450MHz (700MHz) and 4GHz. This requirement to support multiple bands and legacy standards contributes significantly to the design complexity.
Challenges of the LTE Specification
The LTE specification itself represents a complex challenge because it is more than a cell-phone standard. LTE has been designed to replace wireline connections in applications like broadband access and gaming; category-3 LTE offers 100 Mb data rates. In most situations, speed can only come from a 2x multiple-input, multiple-output (MIMO). Intrinsic to LTE’s high-throughput capability are MIMO radios requiring multiple equal-quality air-interface paths. LTE demands 20 MHz bandwidth in the receive chain (compared with less than 4 MHz for 3G standards) and requires a greater dynamic range than 3G. So a transceiver device has to have a second, independent receiver chain. This requires development of new data converters with a new modified topology. Also, with definition of new bands with tight band spacing, the emissions restrictions for LTE are significantly tighter.
Then there are the cost, compact size, and low power requirements typical of all applications and markets today. Mobile phone developers want their handsets to be smaller and lighter, and to consume as little power as possible. The fundamental requirement can be summarized as a highly integrated transceiver that is compact, efficient, and reduces the total component count, board space and bill of materials. Additionally, power dissipation expectations continue to be in the range of 15uA for deep sleep mode.
click to enlarge
Figure 2. The MB86L10A LTE-compliant RF transceiver developed by Fujitsu Semiconductor. |
There is another challenge as well. Emissions restrictions for LTE are tight. The transmit signal spectrum must comply with stringent spectrum emission mask requirements because of the low guard-band size for high spectral efficiency. So it is necessary to modify the transceiver device’s DSP algorithms and hardware. There are significant increases in bandwidth to get to LTE – from 1.4 MHz to 20 MHz – and the dynamic range must be increased. As a result, the digital signal processing becomes more complex.
The Fujitsu team modified data-converter speeds and topologies and reanalyzed additional bandwidth cases, redefining some of the DSP and filtering techniques. The DSP engine was modified because there is a larger signal coming in to meet the required dynamic range. Additional calibration schemes have been added and modified to take specific LTE requirements into account for special requirements such as public safety networks.
The Fujitsu team met all these requirements using a thorough and careful design process. It began with a link budget and gain analysis developed on a spreadsheet, then moved to a floating-point algorithm in Matlab85 in which the group looked specifically for noise, distortions and second-order effects in floating-point model representation. Then the process moved to a fixed-point representation in Signal-Processing Worksystem (SPW) model. Finally, with those steps completed, it was appropriate to recode the DSP engine. Throughout the entire development process, Fujitsu’s close working relationship with leading carriers provided the company with a constant flow of important input and requirements.
Meeting Cost, Space, and Legacy Requirements
The resulting transceiver, the Fujitsu MB86L10A, meets the legacy, cost, space and price requirements of the next-generation mobile phones. Designed for multimode, multiband LTE, UMTS and EDGE mobile handsets, the transceiver eliminates external LNAs and inter-stage SAW filters from the TX and RX paths of 3G and LTE lineups. The receiver incorporates anti-aliasing filters, digital channel filters, digital gain control and high-dynamic-range ADCs. A high-level programming model developed by the design team can control the radio using the 3G and 4G DigRF/MIPI interfaces, supporting a broad range of industry basebands.
The MB86L10A is equipped with DigiRF3G (which is the prior MIPI specification) along with DigiRF4G (which is needed to support LTE). The transceiver can be paired with one or two baseband processor ICs, and supports LTE bands like 1,4,7,13, and 17 and GSM bands GSM850, EGSM900, DCS1800, and PCS1900. There also is support for EGPRS Class 34 operation; WCDMA bands I, II, III, IV, V, VI, VIII, IX, X, and XI; WCDMA FDD HSDPA category 10; and WCDMA FDD HSUPA with 4 E-DPDCH category 6.
The device has RF sections and extensive analog and DSP sections, and is packaged in a flip chip LGA. The team focused intently on isolation to make sure there are no spurious issues. The high level of integration, coupled with more bands and modes, led to a device characterized by lower cost and size.
Just as critical for these advanced applications is power consumption. Current drain is an important issue. The device incorporates Fujitsu IP that is designed specifically to “pull back” when the system is not operating at maximum usage, to reduce current drain. Also with the Fujitsu software, the product can work with a variety of front ends, regardless of whether they are cost-driven or performance-driven.
The MB86L10A builds on the short-cycle RF programming method that was developed for the Fujitsu MB86L01A RF transceiver. The programming capability enables the MB86L10A to speed RF subsystem implementations with simplified layer-one programming and embedded intelligence. The transceiver uses an ARM processor as a CPU and includes dedicated DSP. There are nine primary receive inputs distributed between high and low bands along with five secondary receive inputs and eight transmitter outputs. The power amplifiers are external. An MCU core unit simplifies timing and control. SPI, MIPI RFFE and/or GPOs control PAs, switching regulators and the antenna switch.
As mobile systems based on the LTE standard move into the mainstream, the technology developed by Fujitsu Semiconductor will deliver the performance needed, with the features carriers require and a leadership product in the industry.
Vivek Bhan is Executive Vice President of Fujitsu Semiconductor.
Fujitsu Microelectronics America, Inc.
http://www.fujitsu.com/us/services/edevices/microelectronics/
©
2012
Advantage Business Media
|
