| Presentation | 2012-06-21 Cell Range Expansion Using Distributed Q-Learning in Heterogeneous Networks Toshihito KUDO, TOMOAKI Ohtsuki, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Heterogeneous networks (HetNets) that put pico base stations (PBSs) in the macro cells are necessary to improve the network throughput. PBSs are usually put around hot spots to serve UEs within that area. However, pico cell range is too small to cover the whole area, because the hot spot's location and traffic amount change dynamically. To this end, cell range expansion (CRE), which expands pico cell range by adding a bias value to pico received power during hand-over, instead of increasing pico BS power, is considered. Many studies have focused on inter-cell interference coordination (ICIC) in CRE, because macro strong transmit power harms the expanded region (ER) UEs that select PBSs by added bias value. However, to the best of our knowledge, there has been no studies on the optimal bias value that minimizes the number of outage UEs. This value depends on several factors such as the dividing ratio of radio resource between macro base stations (MBSs) and PBSs, and it is determined only by trial-and-error method. Moreover, most papers use the common bias value among all UEs, however the optimal value varys from UE to another. Thus, in this report we propose a scheme using Q-Learning that UEs learn bias values that minimize the number of outage UEs from past experience. Simulation results show that, compared to the fixed bias value, the proposed scheme reduces the number of outage UEs and improves network throughput. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Heterogeneous network / Cell Range Expansion / picocell / Reinforcement learning (RL) / Q-learning / multi-agent system |
| Paper # | RCS2012-51 |
| Date of Issue |
| Conference Information | |
| Committee | RCS |
|---|---|
| Conference Date | 2012/6/14(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Radio Communication Systems (RCS) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Cell Range Expansion Using Distributed Q-Learning in Heterogeneous Networks |
| Sub Title (in English) | |
| Keyword(1) | Heterogeneous network |
| Keyword(2) | Cell Range Expansion |
| Keyword(3) | picocell |
| Keyword(4) | Reinforcement learning (RL) |
| Keyword(5) | Q-learning |
| Keyword(6) | multi-agent system |
| 1st Author's Name | Toshihito KUDO |
| 1st Author's Affiliation | Keio University() |
| 2nd Author's Name | TOMOAKI Ohtsuki |
| 2nd Author's Affiliation | Keio University |
| Date | 2012-06-21 |
| Paper # | RCS2012-51 |
| Volume (vol) | vol.112 |
| Number (no) | 89 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |