| Presentation | 2021-07-16 Joint Transmit Power and Beamforming Control based on Unsupervised Machine Learning for MIMO Wireless Communication Networks Naoto Tamada, Yuyuan Chang, Kazuhiko Fukawa, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In mobile communications, densely deployed cell systems are expected to improve the system capacity drastically. However, many overlapping cells cause inter-cell interference (ICI), which can damage the improvement of the system capacity. As one of ICI coordination (ICIC) to compensate for the damage, base stations (BSs) control both transmit power and transmit beamforming. Since this kind of ICIC can be regarded as an optimization problem, a conventional scheme conducts exhaustive search (ES) in order to choose the optimal combination of transmit power levels and precoding matrices from a pre-defined codebook. However, ES requires a prohibitive amount of computational complexity that grows exponentially with the number of BSs, and thus can not be applied to a large scale system. To reduce the complexity, this report applies a convolutional neural network (CNN) into the ICIC. The reason for adopting CNN is that CNN requires a small amount of computational complexity for predicting optimal values, although its training process needs a large amount of complexity. Machine learning for CNN can be mainly classified into supervised and unsupervised learning. Since the supervised learning needs results of ES as the training sequence, it is very difficult to adopt the supervised learning in case of a large scale system. Therefore, this report proposes CNN employing unsupervised learning for the ICIC. Computer simulations under MIMO communications with 3 cells having 3-sector antennas clarify that the proposed scheme can improve the system capacity drastically while requiring a less amount of complexity. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | MIMO / inter-cell interference coordination / transmit power control / beamforming control / convolutional neural network / unsupervised learning |
| Paper # | CS2021-29 |
| Date of Issue | 2021-07-08 (CS) |
| Conference Information | |
| Committee | CS |
|---|---|
| Conference Date | 2021/7/15(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Next Generation Networks, Access Networks, Broadband Access, Power Line Communications, Wireless Communication Systems, Coding Systems, etc. |
| Chair | Jun Terada(NTT) |
| Vice Chair | Daisuke Umehara(Kyoto Inst. of Tech.) |
| Secretary | Daisuke Umehara(NICT) |
| Assistant | Takahiro Yamaura(Toshiba) / Yuta Ida(Yamaguchi Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Communication Systems |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Joint Transmit Power and Beamforming Control based on Unsupervised Machine Learning for MIMO Wireless Communication Networks |
| Sub Title (in English) | |
| Keyword(1) | MIMO |
| Keyword(2) | inter-cell interference coordination |
| Keyword(3) | transmit power control |
| Keyword(4) | beamforming control |
| Keyword(5) | convolutional neural network |
| Keyword(6) | unsupervised learning |
| 1st Author's Name | Naoto Tamada |
| 1st Author's Affiliation | Tokyo Institute of Technology(Tokyo Tech) |
| 2nd Author's Name | Yuyuan Chang |
| 2nd Author's Affiliation | Tokyo Institute of Technology(Tokyo Tech) |
| 3rd Author's Name | Kazuhiko Fukawa |
| 3rd Author's Affiliation | Tokyo Institute of Technology(Tokyo Tech) |
| Date | 2021-07-16 |
| Paper # | CS2021-29 |
| Volume (vol) | vol.121 |
| Number (no) | CS-113 |
| Page | pp.pp.63-68(CS), |
| #Pages | 6 |
| Date of Issue | 2021-07-08 (CS) |