Design: Multimode, multiband transceiver technology delivers LTE

With the introduction of smart phones and tablets, mobile data transmission has grown dramatically over the last five years. The Long Term Evolution (LTE) standard improves spectrum utilization efficiency, increases speed, and handles greater data traffic. Success of LTE technology is tied to an ecosystem development that must take place at the same pace—or faster—than that of the infrastructure implementation.

Due to an expected explosive growth of data usage, it has become imperative for operators to make effective use of spectrum and implement LTE as quickly as possible with an ever-increasing number of bands. This is a challenge for transceivers. The 3rd Generation Partnership Project (3GPP) has responded by coming up with a unified approach to FDD and TDD technologies. Currently, the wireless communication spectrum (up to 3.8 GHz) is separated into 43 bands; bands 1 to 32 are classified as LTE-FDD and bands 33 to 43 are classified as LTE-TDD.

From a transceiver standpoint the challenges are:

1. Multiband: such a plethora of LTE bands necessitates a multiband transceiver.
2. Multimode: roaming requirements in legacy operating networks (WCDMA, EVDO, TD-SCDMA, CDMA and GSM) require transceivers to be multimode.
3. Dual technology: transceivers need to support both TDD and FDD technologies.

Transceivers for 0.7 to 2.7 GHz spectrum need to handle both FDD and TDD technologies in order to cover FDD bands 1-32 and TDD bands 33-41. Currently available state-of-the-art transceivers support all of the above bands with multiple ports available from a single RFIC (see Figure 1). This requires significant processing power. In the case of the Fujitsu RFIC shown in the figure, the processing requirement is addressed by distributing the load between the baseband and transceiver. For example, the transceivers are embedded with ARM® processors that reduce baseband processing requirements. This, in turn, reduces power consumption and improves response time for dynamic adjustments.



In addition to being multimode and multiband, today’s multi-featured RF transceivers need features such as:

• Low power consumption
• Small size (for both the transceiver and the RF front end)
• Standards-based baseband interface
• Flexible RF interfaces
• Carrier aggregation capability
• Compliance with 3 GPP standards

Low Power Consumption
Modem power consumption depends on the mode of operation (2G, 3G, 4G), technology (TDD, FDD), band use and 3GPP requirements. The transceivers, the baseband chipset and the RF front-end components must consume low power in all modes, individually and collectively. Flexibility and ease of programming for all the modem components are critical to achieve these goals.

The primary factors that govern power consumption for the transmit path are:  

• Power output
• Power amplifier (PA) efficiency in max-power and backoff modes
• Linearity, adjacent channel power leakage ratio (ACLR) and noise performance.

Factors that govern power consumption on the receive path are:

• Dynamic range of the receiver
• Power consumed by the low-noise amplifiers (LNAs)

Within the RF subsystems, PAs are the major source of power consumption. Efficiency of the PA is greatest in the high-power saturated mode of operation. Most vendors cite their efficiency data based on this mode. However, efficiency drops when power is backed off, as the DC bias to the PA is only optimal at high power, leading to inefficient operation of the PA at a normal backoff power.
Partly to address the inefficiency at backed-off power, we are working on transceivers that could take advantage of a technique called envelope tracking (ET). ET allows the DC bias to be optimized in real time for the desired power reduction, while maintaining good PA efficiency numbers. The ET feature is illustrated in Figure 2. 

                      

Typically, ET optimization realizes a 20 to 30 percent reduction in power consumption.
To comply with 3GPP requirements for different bands, the transceiver has to be optimized for linearity, ACLR and CIM3 in each band. For this optimization, the transceivers must be easily programmable with ample processing capability. This flexibility with high-level API also allows multiple bands to be driven through a single port.

On the receive path, an enhanced transceiver design that eliminates the need for LNAs on the front end is ideal. With multiple-band support, elimination of LNAs and SAW filters reduces power consumption as well as component count and footprint. 

Power consumption can vary based on the dynamic range requirements. Receivers with high dynamic ranges require high linearity, which increases power consumption. Conversely, receivers with low dynamic ranges need less linearity and therefore consume less power. Transceiver programmability provides the flexibility needed to optimize for each situation.

Programmable transceivers that enable flexible RFIC configurations allow manufacturers to customize various settings and optimize tradeoffs for customers. These transceivers can manage power requirements for various elements including operating mode (GSM, WCDMA and LTE), phase noise, linearity, and ease of calibration. Unlike non-programmable transceivers, such optimization does not require any hardware changes, which allows for scalability and re-use with only firmware changes.
Another common element that governs power consumption is the voltage-controlled oscillator (VCO). Power consumption of the VCO increases with reduced phase noise. Thus, phase noise adjustments can be made to hit noise targets that achieve 3GPP compliance. Tuned transmission lines can reduce the number of LO buffers required and further reduce power consumption.