HY
Hongxi Yin
Author with expertise in Content-Centric Networking for Information Delivery
Achievements
Cited Author
Key Stats
Upvotes received:
0
Publications:
4
(0% Open Access)
Cited by:
1,140
h-index:
23
/
i10-index:
47
Reputation
Biology
< 1%
Chemistry
< 1%
Economics
< 1%
Show more
How is this calculated?
Publications
0

Deep-Reinforcement-Learning-Based Optimization for Cache-Enabled Opportunistic Interference Alignment Wireless Networks

Ying He et al.Sep 13, 2017
Both caching and interference alignment (IA) are promising techniques for next-generation wireless networks. Nevertheless, most of the existing works on cache-enabled IA wireless networks assume that the channel is invariant, which is unrealistic considering the time-varying nature of practical wireless environments. In this paper, we consider realistic time-varying channels. Specifically, the channel is formulated as a finite-state Markov channel (FSMC). The complexity of the system is very high when we consider realistic FSMC models. Therefore, in this paper, we propose a novel deep reinforcement learning approach, which is an advanced reinforcement learning algorithm that uses a deep Q network to approximate the Q value-action function. We use Google TensorFlow to implement deep reinforcement learning in this paper to obtain the optimal IA user selection policy in cache-enabled opportunistic IA wireless networks. Simulation results are presented to show that the performance of cache-enabled opportunistic IA networks in terms of the network's sum rate and energy efficiency can be significantly improved by using the proposed approach.