How WCDMA changes mobile phone testing
(Editor: Achim Grolman, Product Marketing Communications Manager for Willtek Communications)
Following a long development and starting phase, Wideband CDMA is a reality now: Each month, new wireless devices for voice and data communications via WCDMA enter the market; the subscriber figures for this technology are increasing slowly but steadily. Manufacturing sites, service centres and large outlets are facing new challenges as the new technology requires new measurements, and testing is a necessary evil: There is no ideal device overcoming any downthrow and large temperature changes without any damage. Customers also increasingly return phones because the data transfer with modern handsets often fails due to a wrong configuration in the menus, or due to software bugs.
WCDMA introduces access and modulation methods that differ from those of GSM and TDMA. Different channels are distinguished not by frequency and/or time slot because they are sent on the same carrier frequency and at the same time. Instead, the channels use different spreading codes to transform low-bandwidth information into a broadband spectrum, making the signal less susceptible to narrowband interferers and multipath fading. The bits of the data stream are referred to as chips after spreading. By knowing and applying the right spreading code during demodulation and decoding, the receiver can filter out the wanted channel from the multitude of signals.
The new methods for channel access and modulation require changed methods of measurement in production and repair centres. However, the well-known separation into receiver and transmitter measurements is a constant that applies to any bidirectional radio communication system.
Transmitter okay?
The transmit power is an important measurement parameter and a decisive factor for the radio coverage, in WCDMA as much as in any other radio system. In GSM the base station commands the mobile phone to set an absolute power level; this feature can be used for measurements at the handset. WCDMA base stations, however, only specify a power level change relative to the actual level. Therefore, two different power measurements are required at the wireless device: absolute measurements of the power level at the upper and lower limit (min-max power measurements), and relative measurements of the accuracy during power changes, i.e. the step width (inner loop power control, see Figure 1). The WCDMA wireless device must be able to reduce its absolute power level to –50 dBm or below. The maximum level is determined by the power class; most of the phones currently available belong to power class 4 with a nominal maximum power of 21 dBm.
Figure 1: Testing the relative power steps in the Inner Loop Power Control test
The modulation method applied by the WCDMA device is Dual-BPSK (binary phase shift keying), which means that the traffic channel is modulated onto the in-phase signal and the signalling channel is modulated onto the quadrature signal. The vector error in the constellation diagram is used to evaluate the modulation quality. This error is calculated for each symbol as the difference between the ideal and the measured signal vector; the root mean square value over all the symbols of a time slot relative to the transmit power gives the Error Vector Magnitude (EVM), which must not exceed a threshold value of 17.5%. Figure 2 shows how the error vector is composed and Figure 3 displays a typical measurement.
Another important parameter of radio communication systems is the frequency error; it should be in the range ±10-6 relative to the frequency of the base station
or test set, i.e. the transmit frequency must be accurate by 190 Hertz.
The frequency spectrum of WCDMA signals is flat over the transmit range, with steep edges towards the neighbouring channels. Spectrum measurements like Occupied Bandwidth (OBW) and Adjacent Channel Leakage Ratio (ACLR) help to verify that a transmitter does not emit too much signal power in the neighbouring frequency channels. The Third Generation Partnership Project 3GPP specifies limit values to prevent spectral components resulting in adjacent channel interference and subsequent impairment of competing networks.
Figure 2: The error vector in the I/Q diagram
Figure 3: Typical measurement display of constellation diagram and numerical results
Clear reception
As shown above, there is a lot of parameters to check the transmitter. Communication usually works in two directions; there are three parameters for the evaluation of the receiver quality: the power level and quality as received and measured by the wireless device at its antenna, and the receiver sensitivity.
GSM handsets measure RXLEV, the received channel power. The equivalent in WCDMA is the CPICH RSCP parameter. CPICH is the Common Pilot Channel that the wireless device synchronises to while looking for a suitable base station, and RSCP stands for the Received Signal Code Power, that is the power level received on the CPICH. This level must be at least –95 dBm while scanning for a new base station; during an ongoing voice call or data transmission, the wireless device also periodically checks this channel, measures the received signal strength and transmits this value to the base station (or the test set). Based on these measurement values, the base station can then initiate a handover to a different cell when appropriate.
Not only the signal strength but also the reception quality is checked during the ongoing call. In GSM, the quality is expressed by the RXQUAL parameter; in WCDMA the received energy per chip relative to the compound signal on the frequency carrier, including noise, is used for the evaluation: CPICH Ec/N0. This does not express the quality of the dedicated voice or data channel but that of the radio channel, which is the same for CPICH and traffic channel. This parameter is also being reported periodically to the base station.
The reception quality measurement on the assigned channel is performed by the test instrument. It sends pseudo- random data on a packet channel. Based on a check sum, the wireless device under test monitors if the received data packets are successfully received and decoded; if not then this is reported to the base station to request re-transmission. The test set uses this feedback to calculate the Block Error Rate (BER). It can also perform a bit error rate measurement, usually at a power level of –106.7 dBm. The maximum allowable bit error rate is 0.1%.
Quick check
Fast and precise measurements during regression tests and in production are usually performed with highly integrated communication testers. One such tester is Willtek’s 4405 Mobile Phone Tester, and even in service there are suitable instruments available like the Willtek 3100 Mobile Fault Finder. R&D engineers are now working on the next technology, as the combination of HSDPA (High Speed Downlink Packet Access) and HSUPA (High Speed Uplink Packet Access) has been announced for commercial launches in Europe and America in spring 2006 and is to fulfill the promises made for UMTS: low-latency transmission at data rates in the order of 2 Mbit/s.
|
