All-digital RF transmitter in 28nm CMOS with programmable RX-band noise shaping

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Researchers

Research units

  • Huawei Technologies

Abstract

The crowded radio spectrum allocated for 3G/4G mobile communication, together with the growing demand for higher data-rates, has led to the situation where transceivers need to support FDD operation in multiple frequency bands with different TX-RX duplex spacing. In order to reduce costs and enable SAW-less operation, many recent transmitter implementations have thus targeted stringent out-of-band (OOB) emission levels. Analog-intensive TX architectures achieve low OOB noise at the price of large area consumption, as complex reconstruction filters are used to suppress DAC quantization noise and image replicas [1,2]. On the other hand, due to the lack of analog filtering, digital-intensive TX architectures need 12-14b DAC resolution for low OOB noise, which complicates DAC design and typically requires DPD or calibration [3-5]. This work presents an RF transmitter implementing a fully digital solution to the aforementioned challenge. Instead of using bulky analog filtering or high resolution DAC, the disclosed TX employs digital ΔΣ modulation and mismatch shaping to attenuate the DAC noise at a programmable duplex distance. This solution enables -160dBc/Hz noise in the RX-band, by using only a 10b DAC without DPD, calibration or analog filtering.

Details

Original languageEnglish
Title of host publication2017 IEEE International Solid-State Circuits Conference, ISSCC 2017
Publication statusPublished - 2 Mar 2017
MoE publication typeA4 Article in a conference publication
EventIEEE International Solid-State Circuits Conference - San Francisco, United States
Duration: 5 Feb 20179 Feb 2017
Conference number: 64

Conference

ConferenceIEEE International Solid-State Circuits Conference
Abbreviated titleISSCC
CountryUnited States
CitySan Francisco
Period05/02/201709/02/2017

Download statistics

No data available

ID: 13574280