Effects of Interference on CDMA Mobile Phone Receiver Testing

fly

Effects of Interference on CDMA Mobile Phone Receiver Testing [Copy link]

When testing mobile phones, it is important to understand the possible sources of interference signals and their effects on the front end of the phone. This can help determine the sensitivity of the receiver test to RF interference and find feasible methods to eliminate the interference signals. This article introduces the interference characteristics encountered when testing CDMA mobile phone receivers, analyzes them in detail, and proposes reference solutions.

A major concern in testing a cell phone is how it responds over a wide dynamic range. This is especially true for CDMA phones compared to analog or TDMA phones because they have an 80dB dynamic range over a wide bandwidth. This high dynamic range is a unique advantage of CDMA technology, and CDMA phones and base stations transmit at lower power, on average, than other technologies. Since such phones can receive signals below -104dBm, noise within the specified spectrum can affect many of the tests listed in the minimum performance standards such as TIA/EIA-98C, so it is very important to understand the impact of interference on CDMA phone receiver testing.

Noise Impact

When testing a mobile phone, you should first understand the structure of the mobile phone before considering noise. The mobile phone has both a transmitter (TX) and a receiver (RX). The transmitter's transmission power is between +30dBm and -55dBm, and the receiver's signal reception range is between -20dBm and -108dBm. The dynamic range of the transmitter and receiver of a CDMA mobile phone is very large, exceeding 80dB. The maximum power (+23dBm for a Class III mobile phone) and the receiver sensitivity (-104dBm) are the most important indicators to understand.

CDMA mobile phone testing requires multiple tests to ensure that the mobile phone meets specific standard requirements. For example, testing the mobile phone transmitter power ensures that the transmission power range is accurate within a wide dynamic range. In particular, the maximum transmission power of the mobile phone must be tested to ensure that the mobile phone radiation power is close to the minimum value specified by the maximum EIRP (effective isotropic radiated power). The transmission power range of Class III mobile phones is +23dBm to +30dBm. Compared with noise, since the power measured by the instrument is larger, noise is generally only an insignificant factor in the entire measured power. In addition, the mobile phone transmitter must also be tested to see if it meets the minimum transmission requirements. The CDMA minimum power transmission stipulates that the mobile phone transmission must be less than -50dBm. Even in this case, the impact of the transmission channel noise is usually negligible and will not affect the minimum power measurement. Problems in measuring CDMA minimum power are generally caused by the minimum noise limit of the power meter used.

The receiver sensitivity measurement will be affected by the noise on the phone's receive channel. The CDMA phone receiver sensitivity is specified as -104dBm. The phone must be able to demodulate the forward link signal transmitted at -104dBm with less than 0.5% frame error rate. Eb/Nt is a parameter that represents the phone's ability to accurately receive and demodulate the forward link signal, where Eb is the energy per bit of the communication channel and Nt is the total noise over the receive bandwidth. This is a bit like the signal-to-noise ratio (S/N) in analog circuits. As the Eb/Nt ratio increases, the receiver is better able to accurately demodulate the signal; as Eb/Nt decreases, the phone is more likely to incorrectly demodulate the forward link signal. The actual bit energy of the forward link communication channel is 15.6dB lower than the total forward link power specification of -104dBm. In other words, the actual signal received by the phone in the measurement is -119.6dBm. From this point on, we will refer to the forward link signal in the content mixed with Walsh codes. Depending on the design tolerance of the mobile phone receiver, its sensitivity to noise is also different. Usually when the performance of the mobile phone is at a marginal value, there will be noise problems on the forward link message channel. There are many factors that affect Nt in Eb/Nt.

Noise sources on the receiving channel

◆The essence of KTB noise floor Nt is thermal noise, which always exists in the environment. Thermal noise is also called KTB noise floor, where:
K = Boltzmann constant (1.38 10-23)
T = reference temperature (Kelvin)
B = receiver effective noise bandwidth

For a CDMA system operating at 1.23MHz bandwidth, the thermal noise is about -113dBm. You may ask, how does the receiver demodulate the -119dBm communication signal with a -113dBm noise floor? This is because the CDMA processing gain is nearly 21dB, which can convert 14.4kbps/9.60kbps to 1.228Mcps.

◆Component noise
The noise of the receiver front-end components (down converter and amplifier) also generates Nt, which affects the sensitivity of the phone, that is, the -104dBm specified sensitivity level achieved by the forward link power test level. In addition, all other noise factors will increase Nt and affect the success of the sensitivity test. Compared with phones on the edge of performance indicators, phones with a certain tolerance have more room to accommodate the increased noise.

◆Environmental noise
and many other noise sources can also reduce the Eb/Nt of the forward communication channel. For example, any circuit with transmitting power will generate spectral noise, which depends on whether the signal itself or the intermodulation of two interfering signals has a higher power falling on the measured bandwidth, and whether it occurs with the sensitivity test. External units can also generate interference noise, especially for AMPS systems when testing 800MHz CDMA phones. There are even reports that microwave ovens can interfere with mobile phone sensitivity tests. There are many production plants where lunch is close to the production line. If you find a lot of sensitivity problems at lunch or during breaks, you should know where to look for the cause.

◆ Testing in close contact
In a production environment, there are many test benches that test phones at various stages according to the test plan. This means that a phone being tested at one stage will interfere with another phone being tested at a different stage. Generally speaking, the phone being interfered with is testing sensitivity. Remember that the forward link is set at -104dBm. The main interference may be a neighboring phone that is receiving a large forward link signal, and a phone is being tested for sensitivity with its forward link signal set at -104dBm or below. Tests of phones that set the forward link at a higher level include dynamic range, minimum transmit power, and open loop range. Generally speaking, the forward link for these tests is -25dBm.

◆ Receiver and transmitter coupling
Another noise source that affects the size of Nt is the cross coupling of the mobile phone transmitter and receiver. This is a mobile phone design problem caused by the lack of proper isolation or matching between the receive and transmit message channels. Because it is a design problem, the solution is very costly and difficult. The forward and reverse links are separated by 45 or 80MHz, so the interference is highly isolated between the links for the unit band and PCS band respectively. Crosstalk from one link to another will be found on the front end of the mobile phone. This is a mobile phone front end design problem.

Receiver Message Channel Sensitivity

Now that the main sources of spectrum noise have been listed above, let's take a look at how these spectrum noises enter the mobile phone's receiving channel.

There are two common methods for testing mobile phones. The most common method is to test via a physical RF connection, which is often called a 'through-power' connection. The less common method of testing is directly through the mobile phone antenna. We assume that all tests discussed here are performed via a physical 'through-power' connector, and whether the antenna is connected depends on the mobile phone production process. In order to find the worst case, we assume that the antenna is already assembled in the mobile phone.

In most cases, mobile phones are susceptible to noise due to poorly shielded RF cables or unconnected antenna ports. Generally speaking, the antenna port is the main source of noise, especially if the antenna is connected during testing. Figure 3 shows the impact that a mobile phone undergoing sensitivity testing may have on another mobile phone with a forward link power of -25dBm. Mobile phone 1 is the mobile phone undergoing testing and has a higher forward link power of -25dBm. Since the transmit signal channel power of mobile phone 1 is -25dBm, assuming the attenuation of the unconnected switch is 20dB, the signal leakage through the unconnected antenna port may be as high as -45dBm.

The attenuation value varies depending on the design. Since the signal leaked through the antenna is propagated through the air, its propagation attenuation can be calculated using the Friis conversion equation:

Where:
Pr = received power
Pt = transmitted power
Gr = antenna transmission gain
Gt = antenna receiving gain
λ = wavelength
d = distance between Tx and Rx

Assuming the worst case scenario, with phone 1 21 meters away from the other phone (assuming only far-field effects, near-field factors are difficult to consider), and the two phones are aligned parallel and there is no attenuation material between the phones, the antenna of phone 2 will receive a signal as high as -83.92dBm. This will result in a -103.92dBm interference signal at the receiver of phone 2, because the attenuation from the antenna to the power switch is assumed to be 20dB. This example illustrates the general reason why one phone generates noise on the receiving channel of another phone. Based on the assumptions made in the example, there are many other situations in actual implementation that will cause different results, which also provides a basis for understanding what kind of noise will enter the message channel of one phone from another. Aspect, distance, switch attenuation, shielding, antenna design and implementation methods, etc., all play a role in the overall conversion of imported noise.

A few points to note

?If testing by 'through-power' connection, it is better to have the antenna disconnected than connected. Since mobile phones are designed around transmitting and receiving signals effectively in the required frequency band, if tested by physical through-power connection, the antenna can still receive external signals, added to the front end of the transmitter and receiver. After connecting the antenna, it is also possible to transmit a large signal to the outside at a higher power when measuring the transmitter of the mobile phone. After eliminating the antenna port matching (removing the antenna from the antenna connector) at the desired test frequency, the mobile phone can receive and transmit signals at a higher attenuation level.

?Since external interference can enter the connection between the RF cable under test and the mobile phone, it is important to use properly shielded cable, preferably N-type connectors with triple-shielded cable.

Shielding the phone during testing will significantly attenuate external interference from other phones or unknown spectrum noise sources. The previous example of phone-to-phone interference is useful in determining shielding attenuation when specific testing needs it. Shielding the phone from external interference is not the only way to eliminate noise. Careful processing of all tested phone frequencies can also effectively avoid interference, which requires more complex control software to initialize the number of message channels and frequencies to eliminate frequency conflicts.

? Paying close attention to the distance between two phones using the same frequency can also enable frequency reuse. The example of phone-to-phone interference is very useful in determining safe reuse distances.

Conclusion

When dealing with receiver interference problems, especially in a production environment, it is not easy to find the reason for the loss of yield due to sensitivity test failures. The location of the factory, working hours, test methods and other seemingly unrelated factors can cause random sensitivity failures. It will be easier to solve the problem of low sensitivity test throughput by taking a step back and understanding the main sources of interference noise and how these noise sources affect the receiver front end.

By: Ryan Hendrickson
CDMA Handset Test Division
Agilent Technologies