Yan Xu

Yan Xu

Auto-generated profile, based on publicly available information

Yan Xu

School of Computer Science and Cyber Engineering, Guangzhou University, Guangzhou 510006, China

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 August 2019 in IEEE Access
Verified
Journal Article
Open access
GLGuangqun Liu
Guangqun Liu
YXYan Xu
Yan Xu
ZHZongjiang He
Zongjiang He
YRYanyi Rao
Yanyi Rao
JXJunjuan Xia
Junjuan Xia
LFLiseng Fan
Liseng Fan
With the development of intelligent connected vehicles (ICVs), there emerge many new services and applications which involve intensive computation. To support the intensive computation in vehicle-to-everything (V2X) communication system, the framework of edge computing networks has been proposed, which exploits the computation ability of edge nodes at the cost of wireless transmission. Hence, it is of vital importance to predict the wireless channel parameters, which can help schedule the system resource management and optimize the system performance in advance. To fulfil this challenge, this paper proposes a novel prediction model based on long short-term memory (LSTM) network, which is powerful in capturing valuable information in the sequence and hence is good at analyzing the spatio-temporal correlation in the channel parameters. To validate the proposed model, we conduct extensive simulations to show that the proposed model is quite effective in the channel prediction. In particular, the proposed model can outperform the conventional ones substantially.

11 April 2019 in IEEE Access
Verified
Journal Article
Open access
YXYan Xu
Yan Xu
JXJunjuan Xia
Junjuan Xia
HWHuijun Wu
Huijun Wu
LFLiseng Fan
Liseng Fan
In this paper, we consider a Q-learning-based power allocation strategy for a secure physical-layer system under dynamic radio environments. In such a system, the transmitter sends the information to the receiver threatened by ${M (M \geq 2)}$ intelligent attackers which have several attack modes and will bring out the severe issue of information security. To safeguard the system security, we formulate the insecure problem as a stochastic game which consists of ${M+1}$ players: the transmitter which can flexibly choose its transmit power, and ${M}$ smart attackers that can determine their attack types. Then, the Nash equilibria (NEs) of the physical-layer secure game are derived, and their existence conditions are taken into account. The simulation results show that the proposed power allocation strategy in the stochastic game can efficiently suppress the attack rate of smart attackers even if there exist multiple smart attackers.

2 October 2019 in IEEE Access
Verified
Journal Article
Open access
JXJunjuan Xia
Junjuan Xia
YXYan Xu
Yan Xu
DDDan Deng
Dan Deng
QZQingfeng Zhou
Qingfeng Zhou
LFLiseng Fan
Liseng Fan
In this paper, we investigate the power control strategy of intelligent secure communication with statistic channel state information (CSI) for Internet of Things (IoT) networks, where a transceiver and an attacker with several attack types, including silent, eavesdrop, jamming and spoofing, are considered. In order to solve the security problem that the transmitter only knows the statistical CSI of attacker, we propose a power control strategy based on Q-learning. In particular, Alice and Eve can choose their actions flexibly to maximize their reward under different system state and learn their best strategy according to the proposed strategy. In addition, the interactions between Alice and Eve are formulated as a zero-sum game, the Nash equilibrium and its existence conditions are deduced. Simulation results show that the impact of statistical CSI of attacker on system security performance can be reflected by the cost of attacker to launch attack and the average channel gain parameters. More importantly, the obtained results also show that the proposed power control strategy based on statistical CSI of attacker is worse than the scheme based on instantaneous CSI for statistical CSI leads a performance loss in terms of security.