While the market ultimately will decide which standards constitute the so-called “fourth generation” or 4G, presently IEEE’s WiMAX and 3GPP’s Long Term Evolution (LTE) are under scrutiny and development. Both are vying for the honor of being the mobile, broadband communications vehicle that will imbue handheld devices with an impressive range of features and services, such as high-speed Internet access, GPS, and telephony.

WiMAX comes to the market from the PC world, championed by Intel and a host of companies most associated with PCs and Wi-Fi. LTE, in contrast, evolved from the 2.5G and 3G efforts of the cell-phone carrier crowd. Anyone who has taken a serious look at the physical layer (PHY) of these candidate standards cannot help but come away feeling that they are converging toward a common-denominator set of technical features. In fact, it will probably be easier to build a common set of chipset solutions that encompass both WiMAX and LTE than one that encompasses, say, Wi-Fi and WiMAX.

Most experts say WiMAX is the most developed standard of the group. Standards 802.16d and 802.16e have been accepted, and basestation and client-system ICs, reference designs, and complete systems have been built and piloted. The missing ingredient, infrastructure, appeared to be the sticking point that hampered WiMAX’s progress in 2007 and 2008 in North America, but recent developments have brought renewed optimism.

Meanwhile, LTE has been chugging along. The latest standard version offers enough specifics to fuel basestation and handheld-device design. Along with those specifics, we’ve seen a marriage of instruments and specialized software to enable developers to make reference designs that meet LTE’s exacting requirements. For example, through a collaborative effort with LitePoint Corp., Tektronix now offers real-time spectrum analysis tools coupled with LitePoint control and analysis software that constitutes the first, purpose-built, LTE development test solutions.

No Time To Be Flying Blind

LTE is every bit as complex a standard as WiMAX. To achieve high data rates without exceeding 20-MHz bandwidths, both standards take advantage of orthogonal frequency-division multiplexing (OFDM) for downloading data from basestations to client systems and a large number of subcarriers. Frequency stability becomes more stringent as data rates increase and modulation changes from binary phase-shift keying (BPSK) to 64 quadrature amplitude modulation (64QAM). Both standards also impose strict limitations on spectral power density, which is necessary to avoid interference and allow multiple clients to share a basestation without excessive error rates.

When dealing with orthogonal subcarriers and a variety of modulation schemes, the number of individual tests can become overwhelming. Error-vector magnitude (EVM) measurement enables one substantive measure of modulation quality. It will be important to be able to apply EVM analysis to specific channels, for instance. The key is to gather sufficient insight into how well a design’s resulting signals conform to a 4G standard and to do so efficiently to enable comprehensive development testing without excessive delay due to inefficient testing methods.

During the development phase, it is critical that designers are able to identify signal failures and to understand the root cause, such as frequency instability, I and Q mismatch, and the like. Thus, designers need tools that help them identify and isolate signal problems.

In terms of measurement bandwidths, if the channel being analyzed is, say, 20 MHz, the testing tools must have sufficient measurement bandwidth to examine third- and fifth-order distortion products. If not, what looks like a reasonably “clean” signal may, in fact, have distortion products that pose interference problems. For the aforementioned 20-MHz channel, designers need up to 100 MHz of real-time capture plus sufficient dynamic range to characterize those fifth-order terms.

The dynamic generation of RF waveforms through digital signal processing (DSP) and the integration of digital and RF circuits, often on the same IC, create issues not seen in traditional RF transceiver designs. Simply passing conformance testing does not ensure a device will work properly. System behavior needs to be carefully and thoroughly observed since software is continually changing the system parameters. These software-controlled changes commonly cause glitches, intermittent interference, pulse aberrations, digital to RF couplings, software-dependent phase errors, and the like.

LTE system designers must fully analyze and characterize their systems. As parameters change over time, these designers must perform frequency-selective triggering to pinpoint the instant a transient event occurs. In addition, they must be able to perform multiple-domain time-correlated analysis to determine the specific causes of problems. Furthermore, by capturing the entire event seamlessly into memory, designers can perform subsequent in-depth analysis. One thing is certain, though. When transients occur, so does spectral leakage, and the result is frequency “splatter.”