| Project/Area Number |
16K18109
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Communication/Network engineering
|
|---|
| Research Institution | Waseda University |
|---|
Principal Investigator |
ZHANG CHENG 早稲田大学, 理工学術院, 助教 (40755089)
|
|---|
| Project Period (FY) |
2016-04-01 – 2018-03-31
|
| Project Status |
Completed (Fiscal Year 2017)
|
| Budget Amount *help |
¥2,470,000 (Direct Cost: ¥1,900,000、Indirect Cost: ¥570,000)
Fiscal Year 2017: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2016: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
|
|---|
| Keywords | mobile offloading / MDP / operators / reinforcement learning / mobile users / mobile data offloading / wireless LAN / deep Q-network, DQN / Markov decision process / データオフロード / ネットワーク経済性 / ゲーム理論 / 情報通信工学 / スマートフォン |
|---|
| Outline of Final Research Achievements |
Cellular network capacity cannot catch up with the pace of mobile traffic increase. Mobile data offloading is utilized to solve this capacity shortage problem because it enables inexpensive and fast deployment. We firstly consider mobile data offloading problem from network side perspective. A two-stage non-cooperative game is proposed to study the oligopolistic mobile network operators’ mobile data offloading market. Then, we study Wi-Fi offloading problem from mobile user’s perspective. Firstly, it is assumed that mobile users’ mobility pattern is known in advance and we formulated the Wi-Fi offloading problem as a finite-horizon discrete-time Markov decision process (MDP). We proposed a dynamic programming algorithm to solve the MDP problem. Then, it is assumed that mobile users’ mobility pattern is unknown. We proposed a deep reinforcement learning based offloading algorithm to solve the wireless LAN offloading problem to minimize the mobile users’ monetary and energy cost.
|
