Yifeng Zhao

Power control with reinforcement learning in cooperative cognitive radio networks against jamming

In this paper, we study the anti-jamming power control problem of secondary users (SUs) in a large-scale cooperative cognitive radio network attacked by a smart jammer with the capability to sense the ongoing transmission power. The interactions between cooperative SUs and a jammer are investigated with game theory. We derive the Stackelberg equilibrium of the anti-jamming power control game consisting of a source node, a relay node and a jammer and compare it with the Nash equilibrium of the game. Power control strategies with reinforcement learning methods such as Q-learning and WoLF-PHC are proposed for SUs without knowing network parameters (i.e., the channel gains and transmission costs of others and so on) to achieve the optimal powers against jamming in this cooperative anti-jamming game. Simulation results indicate that the proposed power control strategies can efficiently improve the anti-jamming performance of SUs.

Secure Machine-Type Communications toward LTE Heterogeneous Networks

Machine-type communication is undergoing ubiquitous development by connecting with trillions of intelligent devices and providing a broad range of applications through all of our lives. MTC can access different networks, which is similar to a fusion of heterogeneous networks, including LTE, WiFi, ZigBee, and so on. However, security and privacy issues still bring about a great challenge for the development of MTC, particularly secure transmission under heterogeneous networks. This article first presents an overview of the MTC architecture and its evolution toward LTE, security threats, and solutions. Then, as 5G systems have been designed to support a diverse set of communications, we propose a scalable cross-layer authentication solution based on MTC devices’ hardware fingerprints toward an LTE heterogeneous network.

Active authentication with reinforcement learning based on ambient radio signals

Active authentication of mobile devices such as smartphones and ipads is promising to enhance security to access confidential data or systems. In this paper, we propose an active authentication scheme, which exploits the physical-layer properties of ambient radio signals to identify mobile devices in indoor environments. More specifically, we discriminate mobile devices in different locations by analyzing the ambient radio sources, because the received signal strength indicator set of the ambient signals measured by a smartphone is usually different from that observed by its spoofer located in another area. We formulate the interactions between the legitimate mobile device and its spoofer as an active authentication game, in which the receiver chooses its test threshold in the hypothesis test in the spoofing detection, while the spoofer chooses its attack strength. In a dynamic radio environment with unknown attack parameters, we propose a learning-based authentication algorithm based on the physical-layer properties of the ambient radio environments. Simulation results show that the proposed scheme accurately detects spoofers in typical indoor environments.

P2P-based resource allocation with coalitional game for D2D networks

Abstract Device-to-Device (D2D) communication is one of the promising technologies in 5G wireless communication system. The concept of D2D communication is to allow the direct communications between two devices reusing the cellular spectrum. Considering the large scale D2D networks, where multiple D2D users can download common files (called content below) from one or multiple other users directly by reusing the same spectrum with cellular users (CU). In this paper, we propose the peer-to-peer (P2P) based resource allocation for D2D networks in which resource includes spectrum and content. We aim at maximizing the sum download rate based on jointly effective spectrum and content allocation. However, this problem is NP-hard. Thus, we firstly develop a distance constrained coalitional game to solve the optimal resource allocation problem. We formulate the optimal spectrum and content allocation as an utility maximization problem and distribute the utility based on each player’s successful transmission probability (STP) which is derived by the tool of stochastic geometry. By theoretical analysis, we derive the proper constrained distance, prove that the proposed algorithm is Nash-stable and get the improvement of the convergence rate. By extensive simulation, we demonstrate the effectiveness of our proposed algorithm which obtains the best sum download rate comparing with other schemes and it is close to the optimal solution.

HEVC video quality assessment database toward LTE system

Though a lot of video quality assessment (VQA) models have been researched, most of them are lack of a recognized video quality assessment database, especially for High Efficiency Video Coding (HEVC) encoded video transmit on Long Term Evolution (LTE) system. On the one hand, HEVC video coding technology is rapidly growing in popularity, thus, it needs to build a new VQA model for HEVC; on the other hand, the LTE system is more popular than other communicate systems currently, as well as, different channels of mobile communication system has different effects on video, therefore, it is necessary to damage video over LTE channel. To achieve that, this paper combines the Gilbert model with LTE system to damage HEVC video, and assesses the quality of deterioration video by the method of Single Stimulus Continuous Quality Evaluation(SSCQE). Finally, building HEVC video quality assessment database over LTE systems.

A novel algorithm of virtual resource allocation in heterogeneous radio access networks

The existing radio access network system is static and rigid which could not satisfy the communication requirements of the modern society - flexibility, mobility and intelligence. This paper studies the problems of virtualized resource allocation with multi-objective mapping in heterogeneous radio access network. First, a mathematical model of the multi-objective mapping for virtualized resources is established for heterogeneous radio access network. Next, the dynamic differential evolutionary algorithm is used to solve the multi-objective model. During the process, the weight values of objective function are adjusted by machines learning algorithm, so as to realize the result that the unilaterally value of multi-objective optimization is close to the value of solo-objective optimization. Finally, simulated experiment shows that the proposed algorithm has good fast convergence in different network load scenarios.

Interference alignment based on antenna selection for massive MIMO system

This paper presents an anti-interference method of massive multi-input multi-output (Massive MIMO) based on antenna selection. Assume that both the base station (BS) and mobile station (MS) can acquire real-time channel state information (CSI), the transmitter and receiver select a set of antennas to transmit signals according to the optimization criterion of maximum channel capacity using an improved distributed antenna selection algorithm with low complexity. After antenna selection, the interference from different spatial data flow is then aligned in one direction or into one subspace, leaving most of the spatial degrees of freedom for useful signals. Simulation results show that the proposed scheme achieves considerably better performance than the zero-forcing (ZF) precoding solution, this advantage is especially obvious in a high SNR regime.

Optimized adaptive algorithm of digital self-interference cancellation based on improved variable step

Full-duplex wireless communication system transmits and receives signals in a same frequency band and at a same time, which solves the problem of the low using efficiency of limited radio spectrum resource, and doubles the channel capacity. However, it also brings a new problem that the full-duplex node will generate a strong self-interference signal. And to deal with the self-interference signal, using existing negative self-interference cancellation and analog self-interference cancellation technology are not enough, effective digital self-interference cancellation is needed for further cancellation. This paper proposes an optimized adaptive algorithm of digital self-interference cancellation based on improved variable step. By using a iteration threshold, we establish a new nonlinear relationship between the step factor and the error signal, which overcomes the slow change problem when the error signal is getting closed to zero, and accelerates the convergence rate.

Capacity analysis for multiple-input multiple-output relay system in a low-rank line-of-sight environment

This article focuses on channel model and capacity of multiple-input multiple-output (MIMO) relay system in a low-rank line-of-sight environment. According to the channel characteristics of MIMO and the correlations among antenna arrays, the expression for system capacity is deduced. Various factors influencing MIMO relay system are analysed based on this expression, and the numerical simulation is performed for verification. The simulation results show that system capacity is greatly increased by adding relay nodes. However, the antenna array spacing and angle at the transmitting and receiving ends, relay node position, Rice K -factor and other factors can all produce certain effects on the capacity of MIMO relay system.

Real-time detection algorithm of abnormal behavior in crowds based on Gaussian mixture model

Recently, abnormal evens detection in crowds has received considerable attention in the field of public safety. Most existing studies do not account for the processing time and the continuity of abnormal behavior characteristics. In this paper, we present a new motion feature descriptor, called the sensitive movement point (SMP). Gaussian Mixture Model (GMM) is used for modeling the abnormal crowd behavior with full consideration of the characteristics of crowd abnormal behavior. First, we analyze the video with GMM, to extract sensitive movement point in certain speed by setting update threshold value of GMM. Then, analyze the sensitive movement point of video frame with temporal and spatial modeling. Identify abnormal behavior through the analysis of mutation duration occurs in temporal and spatial model, and the density, distribution and mutative acceleration of sensitive movement point in blocks. The algorithm can be implemented with automatic adapt to environmental change and online learning, without tracking individuals of crowd and large scale training in detection process. Experiments involving the UMN datasets and the videos taken by us show that the proposed algorithm can real-time effectively identify various types of anomalies and that the recognition results and processing time are better than existing algorithms.