Cell-Free Multi-User Massive MIMO Under Channel Non-Reciprocity
Morte Palacios, Jorge (2019)
2019
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2019-08-08
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-201908012800
https://urn.fi/URN:NBN:fi:tuni-201908012800
Tiivistelmä
In Cell-Free (CF) Massive multiple-input multiple-output (MIMO), a large number of access points (AP) are geographically distributed over the coverage area, and jointly serve a smaller number of users on the same time/frequency resources.
In this thesis, we study the impact of non-reciprocal channels (NRC) and imperfect channel state information (CSI) on Cell-Free massive MIMO systems performance. As non-reciprocity sources, we consider transceiver frequency response mismatches and mutual-coupling mismatches in uplink and downlink analogue processing chains. We study both single-antenna and multi-antenna AP configurations, and in this last case, we also include non-reciprocal mutual coupling in addition to transceiver frequency responses.
We present a novel non-reciprocal channel model based on experimental results from massive MIMO reciprocity calibration tests. Previous models consider that channel non-reciprocity characteristics are fast-varying like random variables; conversely, we consider a model where non-reciprocity values change substantially slower in time, as demonstrated in experimental results. Besides, we derive closed-form analytical expressions of capacity lower bounds for zero-forcing and conjugate beamforming schemes.
The conclusion is that non-reciprocal channels can be a limiting factor for Cell-Free systems performance; nevertheless, only AP mismatches impact on performance while UE mismatches do not affect performance. Furthermore, only phase non-reciprocity degrades MRT performance, whereas both phase and amplitude non-reciprocity degrade ZF performance. Therefore, calibration requirements may dispense with amplitude compensation when APs use MRT scheme, and prioritise phase over amplitude compensation when APs use ZF scheme. Mutual coupling considerately affects both MRT and ZF precoders, but ZF to a greater extent. Hence, calibration procedures should always try to compensate for mutual coupling non-reciprocity.
In this thesis, we study the impact of non-reciprocal channels (NRC) and imperfect channel state information (CSI) on Cell-Free massive MIMO systems performance. As non-reciprocity sources, we consider transceiver frequency response mismatches and mutual-coupling mismatches in uplink and downlink analogue processing chains. We study both single-antenna and multi-antenna AP configurations, and in this last case, we also include non-reciprocal mutual coupling in addition to transceiver frequency responses.
We present a novel non-reciprocal channel model based on experimental results from massive MIMO reciprocity calibration tests. Previous models consider that channel non-reciprocity characteristics are fast-varying like random variables; conversely, we consider a model where non-reciprocity values change substantially slower in time, as demonstrated in experimental results. Besides, we derive closed-form analytical expressions of capacity lower bounds for zero-forcing and conjugate beamforming schemes.
The conclusion is that non-reciprocal channels can be a limiting factor for Cell-Free systems performance; nevertheless, only AP mismatches impact on performance while UE mismatches do not affect performance. Furthermore, only phase non-reciprocity degrades MRT performance, whereas both phase and amplitude non-reciprocity degrade ZF performance. Therefore, calibration requirements may dispense with amplitude compensation when APs use MRT scheme, and prioritise phase over amplitude compensation when APs use ZF scheme. Mutual coupling considerately affects both MRT and ZF precoders, but ZF to a greater extent. Hence, calibration procedures should always try to compensate for mutual coupling non-reciprocity.