A context-aware user-driven strategy to exploit Offloading and sharing in ultra-dense deployments
| Autor: | Faouzi Bouali, Michael Fitch, Klaus Moessner |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Exploit
business.industry Computer science Quality of service 020302 automobile design & engineering 020206 networking & telecommunications Context (language use) 02 engineering and technology Spectrum management 0203 mechanical engineering Radio access technology Wireless lan 0202 electrical engineering electronic engineering information engineering Mobile telephony business Computer network |
| Zdroj: | ICC 2017 IEEE International Conference on Communications (ICC) |
| DOI: | 10.1109/icc.2017.7996703 |
| Popis: | This paper proposes a novel context-aware user- driven strategy to efficiently exploit all available bands and licensing regimes in ultra-dense deployments without prior knowledge about each combination. It relies first on fuzzy logic to estimate the suitability of each radio access technology (RAT) to support the requirements of various applications. Then, a fuzzy multiple attribute decision making (MADM) approach is developed to combine these estimates with the heterogeneous context components to assess the in-context suitability. Based on this metric, a spectrum management strategy is proposed to support interactive video sessions for a set of Bronze and Gold subscriptions. The results reveal that the proposed approach always assigns Gold users to the well-regulated licensed band, while switches Bronze users between licensed and unlicensed bands depending on the operating conditions. This results in a significant improvement of the quality-of-experience (QoE) compared to a baseline that exploits only licensed bands. Then, a comparative study is conducted between the available options to exploit unlicensed bands, namely Offloading and Sharing. The results show that the best option strongly depends on the existing load on WLAN. Therefore, a combined approach is proposed to efficiently switch between both options, which achieves the best QoE for all considered loads. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|