Creating and sharing knowledge for telecommunications

Project: Strongly Nonlinear Massive MIMO Techniques

Acronym: SNMIMO
Main Objective:
Although massive mIMO schemes allow huge capacity gains, they cannot be regarded as a scaled up version of MIMO schemes, since the implementation and signal processing complexity would be prohibitively high. For this reason, low complexity implementations are required in massive MIMO. This means using simplified signal processing techniques to separate different data streams such as ZF, EGC or MRC techniques and/or hybrid analog/digital schemes , together with partially connected implementations. Additionally, quantizers at the ADC and DAC should have low resolution (ideally 1- or 2-bit quantizers) and one should employ low complexity, power efficient strongly nonlinear power amplifiers. Finally there are additional hardware impairments inherent to low complexity hardware implementations such as I-Q imbalances.
The combined effects of all these restrictions and implementations means strong nonlinear effects throughout the transmission chain. The main goal of this project is to study analytically the impact of these strong nonlinear effects on the overall performance of massive MIMO schemes. This includes the spectral characterization of the transmitted signals and the evaluation of distortion levels on the transmitted signals and at the detection level. We also plan to develop transmission and detection techniques that are robust to these nonlinear distortion effects and to study the optimum massive MIMO performance under these conditions.
The basic approach for the analytical characterization nonlinear effects on massive MIMO schemes is to take advantage of the quasi-Gaussian nature of those signals and to employ known results on the impact of non-linear devices in Gaussian signals such as the Busgang theorem and classical IMP (Inter-Modulation Products) analysis.
To minimize the performance degradation due to nonlinear distortion effects we will consider the following alternatives:
- Use of multiple antennas to minimize the distortion at the detection level
- Receiver design taking into account the nonlinear distortion characteristics
- Receivers that estimate and cancel nonlinear distortion
- Quasi-optimum receivers that take advantage of the information on the transmitted signal that is inherent to the nonlinear distortion
Reference: IT
Funding: IT
Start Date: 01-09-2017
End Date: 01-08-2020
Team: Adão Paulo Soares da Silva, Sara Helena Marques Teodoro, Rui Miguel Henriques Dias Morgado Dinis, Atilio Manuel da Silva Gameiro
Groups: Mobile Networks – Av, Radio Systems – Av
Local Coordinator: Adão Paulo Soares da Silva
Associated Publications