Mobile phone RF test summary (II) - receiving sensitivity

For RF modules, the most important parameter of the receiver is sensitivity, so I have made a summary of the sensitivity tests of various modules, hoping that everyone can gain something.

　　The sensitivity of GSM is -102dBm, and the bit error rate requirement at this time is less than 2.439%. But does anyone know that the sensitivity test of GSM should include the sensitivity test of CS1-CS4 and MCS5-MCS9? CS1-CS4 is the modulation method of GMSK, while MCS5-MCS9 is the modulation method of 8PSK. For the same modulation method, the coding efficiency is also different, that is, the coding efficiency of MCS9 is higher than that of MCS5, so the sensitivity requirement of MCS9 is only more than -80 (sorry, I don’t have 3GPP on hand and can’t remember it clearly). Of course, as long as one of these sensitivity tests can pass the others, there will be no big problem, because this test proves that the RF path is fine, and the others are mostly verification of chip performance.

　　Different coding methods or different modulation methods have different information volumes, so according to the Shannong formula, the signal-to-noise ratio requirements are also different, so the sensitivity requirements are also different. Next, let's look at the requirements for sensitivity testing of other RF modules. Let's start with Bluetooth. Bluetooth has three modulation modes, so there are also three sensitivity test standards, namely GFSK, π/4-DQPSK and 8DPSK. In addition, Bluetooth is a TDD system, so there are single and multi-time slot tests (corresponding to GSM's GPRS test, which is multi-time slot. For example, the general mobile phone platform supports CLASS10, so when testing the sensitivity, we use the setting of 4 down and 1 up), such as the DH1-DH5 test. The specification requires that the bit error rate is less than 0.1% when -70dBm. From the above, we can see that the coding methods of GSM's CS1-CS4 and BT's DH1, DM1, DH3, DM3, etc., the multi-time slot test of GPRS corresponds to Bluetooth's DH1, DH3, DH5, GMSK and 8PSK correspond to Bluetooth's GFSK, π/4-DQPSK and 8DPSK. I won't correspond to the following, and everyone can see it naturally.

　　Let's continue to talk about CMMB. The modulation modes of the S band of the ground supplementary network are BPSK, QPSK, and 16QAM, and the Ku band of the satellite end only has BPSK and QPSK. Since most of our CMMB receivers receive signals from the ground supplementary network, we should use signal sources with three modulation modes for testing. I don't have the test specifications for CMMB, so I can only infer that different sensitivity requirements are required according to different modulation modes and rates. Now the resolution supported by mobile phones and PADs has reached 1080P, so the source may be from 520i-1080p, and the data rate will definitely be very different.

　　For WIFI working at 2.4G, there are several standards including 802.11b, g, and n. The modulation modes are BPSK, QPSK, 16QAM, and 64QAM, and the rates are also from 1M to hundreds of megabytes. Of course, this is the same as GSM. One modulation mode corresponds to two or more rates, not one debugging mode and one rate. For details, you have to study the source channel coding and other things by yourself. So there are many corresponding sensitivity tests. Of course, for wifi, we can test the highest rate first. If it can meet the standard, there is no need to test the others. The specification also requires sensitivity from the lowest -82dBm to -62dBm. Since there are many, I will not list them one by one.

　　For TD and WCDMA testing, it is relatively simple. Just pay attention to the following points. First, the sensitivity requirement of TD is -108dBm, because it is necessary to consider that TT, that is, the test error, is +-0.7dB, so it can be below -107.3 (for other sensitivity tests, there is this test error). In addition, the TD specification also mentions a 3.84MHz sensitivity, which is a standard that has not yet been applied. It is estimated that it is for the TDD mode corresponding to W's FDD5M communication mode. Due to its wide bandwidth, the sensitivity is specified as -105, but its bandwidth should be equivalent to W's 5M, but the noise floor is twice that of 1.6M, that is, 3dB is added, which is not clear. For WCDMA, the sensitivity of each frequency band ranges from -103.7 to -106.7. It can be said that it ranges from -114 to -117. Here, a conversion is involved, that is, the higher power is the total downlink power of the base station, which contains a lot of information, while the reference method of the lower power is the power of the user's data channel. The power ratio Ior value of this DPCH is -10.3dB. Comparing W to the sensitivity of TD, you should know that -108 is the total downlink signal power, because the ratio with the data channel is 0dB by default, so TD has only this standard. Students who want to know more can take a good look at the definition of the specification.

　　Finally, the sensitivity requirements for the RF module corresponding to multiple frequency bands are also different, which is easy to understand. For example, GSM has 4 frequency bands, a bandwidth of 200K, and a frequency interval of 200K. TD has two frequency bands, a bandwidth of 1.6M, and a frequency interval of 200K (the frequency interval of the additional channel is 100K). WCDMA has more bandwidth, and we usually use 1, 2, 5, and 8. The bandwidth is 5M, and the frequency interval is the same as TD (I have encountered many cases where I thought the channel interval of W was 5M. You can check the channel number definition in the specification, but the actual channel allocated to a cell is definitely greater than 5M, so please pay attention to the difference). BT is a 2.4G frequency band with a bandwidth of 1M and a frequency interval of 1M. Most of us still only use 2.4G for WIFI, with a bandwidth of 5M, but the bandwidth is adjusted according to its rate, such as 20M or 25M. 5.8G is rarely used. CMMB is from 474M-794M, with a frequency interval of 8M and a bandwidth of 8M.

　　The above is a summary of the sensitivity tests of various RF modules, mainly for your understanding and memory. There is nothing too deep. The sensitivity confirmation of each system is calculated based on the system parameters. I have discussed this with many friends in the forum before. If you are interested, you can look for it. I think even if you are a professional tester, you should know why it is measured this way and why the specification is defined this way. I happen to have a few days off recently, so I have time to help you organize some knowledge. If I am busy, I won’t update so quickly. I hope you can understand then!