The discussion so far has focused on the implications of transceiver design resulting from the additional 3GPP requirements for co-location deployment. These requirements are easily addressed when using RF sampling data converters such as those found in the RFSoC DFE together with widely available low-cost bandpass filters. In practice, there are other considerations and requirements that need to be addressed in addition to co-location. In particular for the receiver, the radio designer must take into account the in-band selectivity and blocking [REF 1, section 7.4] requirements when making design choices on the architecture.
The ZIF receiver has been a popular architecture for 3G and 4G due to the narrowband nature of these radios and the ease of addressing the RF image filter needs of ZIF. However, this is not as applicable for 5G as its wideband nature significantly impacts the key shortcomings of the ZIF transceiver architecture.
To realize the gigabit throughput of a 5G enhanced mobile broadband (eMBB) use case, the bandwidth of the component carriers has increased from 20 MHz in 4G LTE to 100 MHz for FR1 and to 400 MHz for FR2, with the radio RF operating bandwidth often covering 400 MHz in FR1 and 1600 MHz in FR2. Non-contiguous operating bands are also becoming more common due to these very wide RF radio bands or the need to support multiple bands for carrier aggregation. Consequently, the 3GPP compliance requirements for adjacent channel selectively and in-band blocking have become more difficult as the interfering signal is now as wide as 20 MHz [REF 1, section 7.4]. The following sections examine two of the most challenging aspects of using ZIF receivers when handling wideband signals.