Please use this identifier to cite or link to this item: http://localhost/handle/Hannan/205068
Title: Caching Placement with Recommendation Systems for Cache-Enabled Mobile Social Networks
Authors: Yanfeng Wang;Mingyang Ding;Zhiyong Chen;Ling Luo
Year: 2017
Publisher: IEEE
Abstract: Caching popular contents at user devices and sharing the cached content among users is one promising solution to alleviate the heavy base station burden in mobile social networks. In this letter, we investigate social relationships and physical coupling among users, and, then, choose one important user (IU) as a helper to cache target contents and other users can get contents from IUs devices. In particular, because users tend to pay more attention to contents they interested in, a recommendation system with caching placement is proposed to maximize the offloading probability for mobile social networks. We then improve the system to combine three operations: pre-filtering, collaborative filtering algorithm, and latent factor algorithm. Finally, simulation results show the proposed recommendation system-based caching placement scheme achieves a great performance gain over existing approaches, and indicate that important user devices should cache files that recommended most to users.
URI: http://localhost/handle/Hannan/205068
volume: 21
issue: 10
More Information: 2266,
2269
Appears in Collections:2017

Files in This Item:
File SizeFormat 
7931552.pdf460.78 kBAdobe PDF
Title: Caching Placement with Recommendation Systems for Cache-Enabled Mobile Social Networks
Authors: Yanfeng Wang;Mingyang Ding;Zhiyong Chen;Ling Luo
Year: 2017
Publisher: IEEE
Abstract: Caching popular contents at user devices and sharing the cached content among users is one promising solution to alleviate the heavy base station burden in mobile social networks. In this letter, we investigate social relationships and physical coupling among users, and, then, choose one important user (IU) as a helper to cache target contents and other users can get contents from IUs devices. In particular, because users tend to pay more attention to contents they interested in, a recommendation system with caching placement is proposed to maximize the offloading probability for mobile social networks. We then improve the system to combine three operations: pre-filtering, collaborative filtering algorithm, and latent factor algorithm. Finally, simulation results show the proposed recommendation system-based caching placement scheme achieves a great performance gain over existing approaches, and indicate that important user devices should cache files that recommended most to users.
URI: http://localhost/handle/Hannan/205068
volume: 21
issue: 10
More Information: 2266,
2269
Appears in Collections:2017

Files in This Item:
File SizeFormat 
7931552.pdf460.78 kBAdobe PDF
Title: Caching Placement with Recommendation Systems for Cache-Enabled Mobile Social Networks
Authors: Yanfeng Wang;Mingyang Ding;Zhiyong Chen;Ling Luo
Year: 2017
Publisher: IEEE
Abstract: Caching popular contents at user devices and sharing the cached content among users is one promising solution to alleviate the heavy base station burden in mobile social networks. In this letter, we investigate social relationships and physical coupling among users, and, then, choose one important user (IU) as a helper to cache target contents and other users can get contents from IUs devices. In particular, because users tend to pay more attention to contents they interested in, a recommendation system with caching placement is proposed to maximize the offloading probability for mobile social networks. We then improve the system to combine three operations: pre-filtering, collaborative filtering algorithm, and latent factor algorithm. Finally, simulation results show the proposed recommendation system-based caching placement scheme achieves a great performance gain over existing approaches, and indicate that important user devices should cache files that recommended most to users.
URI: http://localhost/handle/Hannan/205068
volume: 21
issue: 10
More Information: 2266,
2269
Appears in Collections:2017

Files in This Item:
File SizeFormat 
7931552.pdf460.78 kBAdobe PDF