The IEEE 802.16 standard promulgated by the IEEE in December 2001 regulates the air interface physical layer and MAC layer of fixed broadband wireless access systems operating in the 10 to 66 GHz band for line-of-sight (LOS) transmission. IEEE 802.16a extends it to non-line-of-sight (NLOS) transmission and specifies Orthogonal Frequency Division Multiplexing (OFDM) and Media Access Control Layer (MAC) orthogonal frequency division multiple access for the physical layer (PHY). (OFDMA) working mode, supporting real-time services such as voice and video. IEEE 802.16d further improves system performance and simplifies deployment. The biggest difference between the IEEE 802.16e/Mobile WiMax standard and the previous ones is the support for mobility. As the technology matures, it has gained more and more attention and application. The IEEE 802.16 working group can be seen as a standard setter, while WiMax is the standard enabler. Terminal equipment is an important part of WiMax applications, and its RF front-end design is also one of the most highly regarded parts.
In general, in modern RF systems, the antenna receives a high frequency signal and has a very small channel bandwidth. If you consider directly filtering out the required channel, the Q value of the filter will be very large, and the high-frequency circuit has problems in terms of gain, accuracy, and stability. Under the current technical conditions, the signal is directly demodulated in the high frequency band. It is unrealistic. Using a mixer to down-convert high-frequency signals, channel filtering, amplification, and demodulation at an intermediate frequency can solve the above-mentioned difficulties encountered in high-frequency signal processing, but introduce another serious problem, that is, image frequency interference. : When the frequency difference between the frequency of the two signals and the local oscillator (LO) signal is symmetrically located on both sides of the local oscillator signal on the frequency axis, or their absolute values ​​are equal but the signs are opposite, then both signals are mixed after mixing Will be moved to the same IF frequency. If one of them is a useful signal and the other is a noise signal, the frequency at which the noise signal is located is called the image frequency. This mixed interference phenomenon is often called image interference. In order to suppress image frequency interference, a commonly used method is to use a filter to filter out image frequency components. However, since the filter operates in the high frequency band, the filtering effect depends on the distance between the image frequency and the signal frequency, or on the frequency of the intermediate frequency. If the IF frequency is high and the signal frequency is far from the image frequency, the image frequency component is greatly suppressed. Conversely, if the IF frequency is lower, the signal frequency is not far from the image frequency, and the filtering effect is poor. On the other hand, since the channel selection is performed in the intermediate frequency band, for the same reason, the higher intermediate frequency has higher requirements for the channel selection filter. Therefore, image frequency suppression and channel selection form a pair of contradictions, and the selection of the intermediate frequency becomes the key to balancing this contradiction. In some demanding applications, two or three conversions are often used to achieve a better compromise.
In general, filters have a high Q value due to the filtering out of a narrow channel with a very high center frequency and high interference. In the heterodyne structure, the signal band is transformed to a much lower frequency, thereby reducing the requirements for the channel selection filter. The heterodyne structure can be comprehensively considered from the aspects of image rejection and channel selection. Since the image signal reduces the sensitivity of the receiver, the selection of the intermediate frequency requires a trade-off between sensitivity and selectivity. In the RF front-end design of IEEE802.16e/WiMAX, the heterodyne transmitter has lower requirements for DAC than the direct conversion, and the mirroring problem becomes less prominent. But the number of module devices has increased, which means more power consumption. The most common way to suppress the image signal is to use an image rejection filter placed in front of the mixer. The filter is designed to have a small loss in the useful frequency band and a large attenuation in the image band. The heterodyne structure requires a mirror filter, but due to the large frequency separation, the design of the image filter is relatively simple. Also note here are the reachability and physical size of the different frequency filters.
Superheterodyne receivers have great advantages in suppressing image frequency interference, sensitivity and selectivity, and multi-level conversion has no DC offset and signal leakage, but it also has high cost, high requirements for IR filters, and low requirements. The disadvantages of noise amplifier (LNA) and mixer (Mixer) are well matched with 50W. In some cases, the image frequency suppression filter and the channel selection filter are not suitable for monolithic integration, resulting in 50 ohm impedance matching of the front stage (such as LNA), which increases the gain, stability, power consumption, etc. of the LNA. A tradeoff in performance.
The simplicity of the zero-IF (zero difference, direct downconversion) structure has two important advantages over the heterodyne structure. First, the mirroring problem is overcome, so no mirror filter is needed, so the LNA does not need to drive a 50 ohm load. Second, the SAW filter and subsequent downconversion stages can be replaced with low-pass filters and baseband amplifiers that are suitable for monolithic integration. However, zero-IF architecture suppresses out-of-channel interference through active low-pass filters during channel selection. It is more difficult than using passive filters, and produces DC offset, IQ mismatch, even-order distortion, flicker noise, and LO leakage. And other issues.
The zero-IF architecture has looser requirements for RF filters, and in the IF section, baseband filters are generally easier to implement than bandpass filters. In such a structure, MIMO technology is also easy to implement. In addition, in general, the zero-IF architecture is also superior in terms of power consumption. However, it is necessary to pay special attention to the IQ equalization problem, the high SNDR DAC design, the DC offset cancellation, etc., especially the DC offset cancellation problem at the receiving end, which needs to be treated with great care and pay attention to the channel equalization. Filtering out-of-band noise requires a high-order filter.
Compared to the zero-IF architecture, digital processing can avoid IQ mismatch problems. Moreover, the digital intermediate frequency structure has various advantages. The IQ equalization problem in the baseband-intermediate frequency, the DC offset problem is easy to solve; the lower out-of-band shaping leakage requirements and the adjustable amplitude; and the signal has a lower requirement for the bandpass filter after the baseband-midband. It is easy to reach the indicator; the load capacity and amplitude requirements of the mixer are not high enough for the attenuator. However, digital IF receivers have high requirements for analog-to-digital converters (ADCs), such as ADCs with high enough dynamic range, low quantization noise and thermal noise, good linearity, and large enough dynamic range. In some low-rate applications, such as IEEE802.15.4, the bandpass Σ-Δ ADC is more suitable, but the bandpass Σ-Δ ADC has a larger design difficulty. It also means higher DSP performance requirements for the baseband portion, such as window filtering. Moreover, the DAC requirements are correspondingly increased, usually requiring 10 to 12 bits of resolution and a higher rate. The performance of the image rejection filter is also severe, and RF filter compensation is required even in certain frequency regions. So in WiMax applications, digital IF architecture has great potential, but there is a trade-off between design capabilities.
The simple comparison of the number of modules of the three receiving structures corresponding to the WiMax terminal transceiver system and its digital baseband processing section is shown in the following table. Some minor modules and some modules that are not transceiver paths are omitted here.
Here, the blank bar does not fully indicate that the module is not needed, but is determined according to specific design indicators. In addition, the digital intermediate frequency structure has a slight advantage over the other two in comparison with the error vector amplitude performance difference caused by phase imbalance in QAM64 and QAM16 modulation. In the QPSK modulation mode, the digital intermediate frequency structure only slightly increases when the gain imbalance is large. Disadvantages. In general, under the three modulation modes of WiMax, the error vector amplitude performance difference caused by phase imbalance in the three receiving structures is extremely small.
The performance parameters of several WiMax chips are given. It can be seen that among these chips, the zero-IF structure is more common.
Here, the blank bar does not fully indicate that the module is not needed, but is determined according to specific design indicators. In addition, the digital intermediate frequency structure has a slight advantage over the other two in comparison with the error vector amplitude performance difference caused by phase imbalance in QAM64 and QAM16 modulation. In the QPSK modulation mode, the digital intermediate frequency structure only slightly increases when the gain imbalance is large. Disadvantages. In general, under the three modulation modes of WiMax, the error vector amplitude performance difference caused by phase imbalance in the three receiving structures is extremely small [1].
HuiZhou Superpower Technology Co.,Ltd. , https://www.spchargers.com