ICACT20210112 Slide.17
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
Thanks everyone for attention! If you have any questions, please send them via email to address rusly.eic04g@nctu.edu.tw
|
|
ICACT20210112 Slide.16
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
Finally, we would like to conclude our work. First, through extensive simulations we have shown an impact of channel polarization of the CDL channel model for arrays with co-polarized and cross-polarized antennas on MIMO transceivers is evaluated. Next, we concluded that polarization causes 2-6dB degradation for traditional co-polarized arrays, while for cross polarized arrays, this value is 1-3dB. Therefore, given cross-polarized antenna arrays invariant to slant angle, it is able to provide better and stable QoS and likely to be extensively used in 5G NR.
|
|
ICACT20210112 Slide.15
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
Finally, to show correlation pattern we simulate dependence of BER from spacing, with small step, to see the fluctuation. We do the experiment for three different SNR values. The results show, that performance is greatly affected by antenna spacing for separation distance of less than 20 wavelength and may change in range of half of the BER order.
|
|
ICACT20210112 Slide.14
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
In the second that of measurements, we want to show how different antenna spacing affects the performance for CDL channels. Here we consider Cpol-Cpol antenna array only and spacing varies from 0.5 wavelength to 20 wavelength. As the spacing increases, performance approaches those of Rayleigh channel as signal becomes less correlated.
|
|
ICACT20210112 Slide.13
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
For Alamoudi transmission the situation is slightly different. That is due to the fact that former transmission scheme is robust to a fading at single antenna. Therefore, now performance of Xpol-Xpol and Xpol-Cpol matches. Therefore when severe fading is at one antenna of Cpol array, Alamoudi decoding is able to recover symbol during the second transmission interval
|
|
ICACT20210112 Slide.12
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
For MMSE receiver performance trend is exactly the same as for ZF. Xpol-Xpol outperforms other configurations.
|
|
ICACT20210112 Slide.11
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
On the LHS we can BER for ZF in CDL-A,B,C channels (NLoS) and on the RHS - CDL- D,E (LoS). There are 6 curves on each. The reference curve (black) represents performance of the 2x2 MIMO receiver in Rayleigh channel. The rest are different antenna configurations. Since performance of Cpol-Cpol is heavily dependent on the slant angle, we consider three scenarios: 1) worst - orthogonal positioning of Tx-Rx (slant=90) 2) best - aligned positioning of Tx-Rx (slant=0) 3) random positioning (slant = 0~90). For other two configurations, only random placement is considered. On the figure we can see, that when Cpol-Cpol is at it best scenario the performance matches those of the Rayleigh channel, while for worst alignment, the performance is off by a lot. In realistic scenario, orientation is random and performance is also located between best and worst (4db off Rayleigh). Interestingly performance of Xpol-Xpol configuration for random orientation is quite close to the Rayleigh channel, which gives it advantage over Cpol-Cpol configuration. Similar situation can be observed for NLoS scenario, however here the performance of Cpol is further off due to strong signal correlation at Rx antennas.
|
|
ICACT20210112 Slide.10
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
On this slide, detailed list of parameters that were used in simulations is shown. it was chosen according to 3GPP Rel-15. All 5 profiles for CDL are considered. The channel is simulated at 28GHz and QPSK modulation was used. Next we show receive performance in terms of BER.
|
|
ICACT20210112 Slide.09
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
Given effect on received signal power dependency on slant angle, we are now interested, what effect it has on the BER performance of state-of art receivers. Here we consider ZF, MMSE, and Alamoudi coding scheme (STBC).
|
|
ICACT20210112 Slide.08
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
In first set of measurements we want to show how the signal strength changes for different slant angle (Tx-Rx orientation). On the left side, we can see 4 different figures which represent signal strength in each of the links of 2x2 MIMO system. Three different antenna configurations are considered: Cpol-Cpol, Xpol-Xpol, and Xpol-Cpol. Cpol-Xpol is not considered, because it is essentially the same as Xpol-Cpol. Let's take a look at top left figure. In there Cpol-Cpol an Xpol-Xpol have similar behavior with peaks 0 and 180 degrees (max) and 90 degrees (min). Xpol-Cpol has opposite pattern because in it two antennas are perpendicular for slant 0. For bottom right figure patterns for all three configuration match, because another element of Xpol is copolarized with both elements of Cpol. Top right and bottom left figures can be explain using same principle. Next, let's take a look at the right figure, which shows aggregated power, which is sum of squared magnitudes of all four links. Here, we can observe interesting result. For Xpol-Xpol and Xpol-Cpol antenna configurations overall magnitude is independent of slant angle, while strong variations are persistend for Cpol-Cpol configuration.
|
|
ICACT20210112 Slide.07
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
For the system model, we assume Tx and Rx equipped with multiple antennas. Also we assume OFDM transmission, so that received signal can be simply written as transmitted signal multiplied by channel matrix plus noise. In our analysis we focus on size 2x2, where polarization effects can be well described theoretically.
|
|
ICACT20210112 Slide.06
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
On this slide you can see 5 different CDL channel profiles, named A-E, where A, B, C describe NLoS scenario (energy is spread) and D, E focus on LoS scenario (energy is concentrated in first few taps).
|
|
ICACT20210112 Slide.05
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
As was mentioned in motivation section, herein we consider CDL channel model, which is a new 5G channel model for wide range of frequencies. It is is a geometry-based channel model, which is suitable for angle-dependent analysis, such as antenna array processing. This model includes wide range of parameters: delay spread, delay values, angle spread, shadow fading, and cross polarization ratio.
|
|
ICACT20210112 Slide.04
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
In our work we consider two antenna configurations that act differently in polarization environment as will be show later. On the left side we can see copolarized antenna array, two parallel antenna elements. On the right side, the crosspolarized antenna is shown, which consists of two elements that are perpendicular to each other.
|
|
ICACT20210112 Slide.03
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
So let's do a small theoretical review on antenna polarization. First of all, the polarization of an antenna refers to the orientation of the electric field (E-plane) of the radio wave with respect to the Earth's surface. Usually antennas are linearly (horizontally, vertically) or circularly polarized. Cross polarization ratio is a ratio of the copolar component of the specified polarization compared to the orthogonal cross-polar component over the sector or beamwidth angle. Simply speaking, cross polarization ratio is how much wave polarization coincides with an antenna polarization.
|
|
ICACT20210112 Slide.02
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
Let me begin with the motivation for our work. First of all, Antenna polarization further offers extra degrees of freedom for MIMO processing, for example, diversity and beamforming. The contribution of this work lies in investigating how antenna polarization influences link performance using the actual CDL channel models, which are proposed in recent Rel-15. Our simulation Simulation results reflect performance gaps between the different polarization scenarios, which can serve as a reference design for future 5G communication.
|
|
ICACT20210112 Slide.01
[Big slide for presentation]
[YouTube] |
Chrome Text-to-Speach Click!! |
 |
Hello everyone! My name is Mykola Servetnyk, I'm from Foxconn Advanced Communication academy, and today I will present a work called "Performance Evaluation of Polarization Effects on 5G Clustered Delay Channels". I've done this work in collaboration with Wei-Han Hsiao, who is a member of Foxconn Advanced Communication Academy and Alisher Mukashev, who is a Ph.D. student at National Chiao Tung University.
|
