Showing papers by "Mugen Peng published in 2017"

On the Spectral Efficiency and Security Enhancements of NOMA Assisted Multicast-Unicast Streaming

Abstract: This paper considers the application of non-orthogonal multiple access (NOMA) to a multi-user network with mixed multicasting and unicasting traffic. The proposed design of beamforming and power allocation ensures that the unicasting performance is improved while maintaining the reception reliability of multicasting. Both analytical and simulation results are provided to demonstrate that the use of the NOMA assisted multicast-unicast scheme yields a significant improvement in spectral efficiency compared with orthogonal multiple access (OMA) schemes which realize multicasting and unicasting services separately. Since unicast messages are broadcast to all the users, how the use of NOMA can prevent those multicast receivers intercepting the unicasting messages is also investigated, where it is shown that the secrecy unicasting rate achieved by NOMA is always larger than or equal to that of OMA. Simulation results are provided to verify the developed analytical results and demonstrate the superior performance of the proposed NOMA scheme.

A Data Mining Approach Combining $K$ -Means Clustering With Bagging Neural Network for Short-Term Wind Power Forecasting

Abstract: Wind power forecasting (WPF) is significant to guide the dispatching of grid and the production planning of wind farm effectively. The intermittency and volatility of wind leading to the diversity of the training samples have a major impact on the forecasting accuracy. In this paper, to deal with the training samples dynamics and improve the forecasting accuracy, a data mining approach consisting of ${K}$ -means clustering and bagging neural network (NN) is proposed for short-term WPF. Based on the similarity among historical days, ${K}$ -means clustering is used to classify the samples into several categories, which contain the information of meteorological conditions and historical power data. In order to overcome the over fitting and instability problems of conventional networks, a bagging-based ensemble approach is integrated into the back propagation NN. To confirm the effectiveness, the proposed data mining approach is examined on real wind generation data traces. The simulation results show that it can obtain better forecasting accuracy than other baseline and existed short-term WPF approaches.

Resource Allocation Optimization for Delay-Sensitive Traffic in Fronthaul Constrained Cloud Radio Access Networks

Abstract: The cloud radio access network (C-RAN) provides high spectral and energy efficiency performances, low expenditures, and intelligent centralized system structures to operators, which have attracted intense interests in both academia and industry. In this paper, a hybrid coordinated multipoint transmission (H-CoMP) scheme is designed for the downlink transmission in C-RANs and fulfills the flexible tradeoff between cooperation gain and fronthaul consumption. The queue-aware power and rate allocation with constraints of average fronthaul consumption for the delay-sensitive traffic are formulated as an infinite horizon constrained partially observed Markov decision process, which takes both the urgent queue state information and the imperfect channel state information at transmitters (CSIT) into account. To deal with the curse of dimensionality involved with the equivalent Bellman equation, the linear approximation of postdecision value functions is utilized. A stochastic gradient algorithm is presented to allocate the queue-aware power and transmission rate with H-CoMP, which is robust against unpredicted traffic arrivals and uncertainties caused by the imperfect CSIT. Furthermore, to substantially reduce the computing complexity, an online learning algorithm is proposed to estimate the per-queue postdecision value functions and update the Lagrange multipliers. The simulation results demonstrate performance gains of the proposed stochastic gradient algorithms and confirm the asymptotical convergence of the proposed online learning algorithm.

A Non-Orthogonal Multiple Access-Based Multicast Scheme in Wireless Content Caching Networks

Abstract: A key problem of content caching networks is that extra radio resource blocks are consumed to push content objects, which leads to a decline of spectrum efficiency. To solve this problem, a non-orthogonal multiple access-based multicast (NOMA-MC) scheme is proposed in this paper, where pushing and multicasting content objects can be accomplished simultaneously, and thus the spectrum efficiency can be improved significantly. To evaluate the performance of the NOMA-MC scheme, an explicit expression of outage probability is derived, which shows that full diversity gains can be achieved in the single-cell scenario. Moreover, the theoretical results can be extended to the multi-cell scenario by establishing a stochastic geometry-based network model, which show that the NOMA-MC scheme can achieve better performance than the conventional orthogonal multiple access-based multicast scheme. Then, the joint design of power allocation and content matching is studied to enlarge the performance gains of the NOMA-MC scheme, and two distributed optimization algorithms are proposed by solving a hospitals/residents matching problem. Finally, simulation results are provided to verify the analytical results, and also demonstrate the performance gains of the NOMA-MC scheme.

An Evolutionary Game for User Access Mode Selection in Fog Radio Access Networks

Abstract: The fog radio access network (F-RAN) is a promising paradigm to provide high spectral efficiency and energy efficiency. Characterizing users to select an appropriate communication mode in F-RANs is critical for performance optimization. With evolutionary game theory, a dynamic mode selection is proposed for F-RANs, in which the competition among the groups of potential users’ space is formulated as a dynamic evolutionary game, and the game is solved by an evolutionary equilibrium. Stochastic geometry tool is used to derive the proposals’ payoff expressions for both fog access point and device-to-device users by considering node location, cache sizes, as well as the delay cost. The analytical results for the proposed game model and the corresponding solution are evaluated, which show that the evolutionary game-based access mode selection algorithm has a better payoff than the max rate-based algorithm.

A Data Mining Approach Combining K-Means Clustering with Bagging Neural Network for Short-term Wind Power Forecasting

Abstract: Wind power forecasting (WPF) is significant to guide the dispatching of grid and the production planning of wind farm effectively. The intermittency and volatility of wind leading to the diversity of the training samples have a major impact on the forecasting accuracy. In this paper, to deal with the training samples dynamics and improve the forecasting accuracy, a data mining approach consisting of K-means clustering and bagging neural network is proposed for short-term WPF. Based on the similarity among historical days, K-means clustering is used to classify the samples into several categories, which contain the information of meteorological conditions and historical power data. In order to overcome the over fitting and instability problems of conventional networks, a bagging-based ensemble approach is integrated into the back propagation neural network. To confirm the effectiveness, the proposed data mining approach is examined on real wind generation data traces. The simulation results show that it can obtain better forecasting accuracy than other baseline and existed short-term WPF approaches.

Network Slicing in Fog Radio Access Networks: Issues and Challenges

Abstract: Network slicing has been advocated by both academia and industry as a cost-efficient way to enable operators to provide networks on an as-a-service basis and meet the wide range of use cases that the fifth generation wireless network will serve. The existing works on network slicing are mainly targeted at the partition of the core network, and the prospect of network slicing in radio access networks should be jointly exploited. To solve this challenge, enhanced network slicing in F-RANs, called access slicing, is proposed. This article comprehensively presents a novel architecture and related key techniques for access slicing in F-RANs. The proposed hierarchical architecture of access slicing consists of a centralized orchestration layer and a slice instance layer, which makes access slicing adaptively implementable in a convenient way. Meanwhile, key techniques and their corresponding solutions, including radio and cache resource management as well as social-aware slicing, are presented. Open issues in terms of standardization developments and field trials are identified.

Secrecy-Optimized Resource Allocation for Device-to-Device Communication Underlaying Heterogeneous Networks

Abstract: Device-to-device (D2D) communications have recently attracted much attention for the potential capability to improve spectral efficiency (SE) underlaying the existing heterogeneous networks (HetNets). Due to no sophisticated control, D2D-worked user equipments (DUEs) themselves cannot resist eavesdropping or security attacks. It is urgent to maximize the secure capacity for both cellular users and DUEs. This paper formulates the radio resource-allocation problem to maximize the secure capacity of DUEs for D2D communication underlaying HetNets, which consist of high-power nodes (HPNs) and low-power nodes (LPNs). The optimization objective function with transmit bit rate and power constraints, which is nonconvex and hard to directly derive, is first transformed into a matrix form. Then, the equivalent convex form of the optimization problem is derived according to Perron–Frobenius theory. A heuristic iterative algorithm based on the proximal theory is proposed to solve this equivalent convex problem through evaluating the proximal operator of the Lagrange function. Numerical results show that the proposed radio resource-allocation solution significantly improves the secure capacity with a fast convergence speed.

Diffusion based molecular communication: principle, key technologies, and challenges

Abstract: Molecular communication (MC) is a kind of communication technology based on biochemical molecules for internet of bio-nano things, in which the biochemical molecule is used as the information carrier for the interconnection of nano-devices. In this paper, the basic principle of diffusion based MC and the corresponding key technologies are comprehensively surveyed. In particular, the state-ofthe- art achievements relative to the diffusion based MC are discussed and compared, including the system model, the system performance analysis with key influencing factors, the information coding and modulation techniques. Meanwhile, the multi-hop nano-network based on the diffusion MC is presented as well. Additionally, given the extensiveness of the research area, open issues and challenges are presented to spur future investigations, in which the involvement of channel model, information theory, self-organizing nano-network, and biochemical applications are put forward.

Cost-Efficient Resource Allocation in Cloud Radio Access Networks With Heterogeneous Fronthaul Expenditures

Abstract: As an advanced paradigm, the cloud radio access network (C-RAN) promises high spectral efficiency (SE) and energy efficiency (EE). However, the capacity-constrained front-haul has become a key performance bottleneck in C-RANs. Besides SE and EE metrics, the cost related to different kinds of fronthaul is another major concern for operators. In this paper, an economical SE (ESE) metric is proposed to jointly take traditional EE and the impact of wired/wireless fronthaul cost into account. Aiming to maximize ESE, a non-convex beamformer design problem with fronthaul capacity and transmit power constraints is formulated, and an algorithm containing outer and inner loops is proposed to deal with this non-convexity. In particular, in the outer loop, the bisection search method is adopted to transform the primal problem into an equivalent subproblem; while in the inner loop, owing to the equivalence between the subproblem and the weighted sum rate, the subproblem is solved by the weighted minimum mean square error approach. Simulation results demonstrate that the proposed ESE is more reasonable than SE and EE for the fronthaul-constrained C-RAN. Furthermore, the proposed optimization solution can significantly improve ESE, in which the impact of fronthaul cost on ESE is evaluated.

Resource Allocation in Cloud Radio Access Networks With Device-to-Device Communications

Abstract: To alleviate the burdens on the fronthaul and reduce the transmit latency, the device-to-device (D2D) communication is presented in cloud radio access networks (C-RANs). Considering dynamic traffic arrivals and time-varying channel conditions, the resource allocation in C-RANs with D2D is formulated into a stochastic optimization problem, which is aimed at maximizing the overall throughput, subject to network stability, interference, and fronthaul capacity constraints. Leveraging on the Lyapunov optimization technique, the stochastic optimization problem is transformed into a delay-aware optimization problem, which is a mixed-integer nonlinear programming problem and can be decomposed into three subproblems: mode selection, uplink beamforming design, and power control. An optimization solution that consists of a modified branch and bound method as well as a weighted minimum mean square error approach has been developed to obtain the close-to-optimal solution. Simulation results validate that D2D can improve throughput, decrease latency, and alleviate the burdens of the constrained fronthaul in C-RANs. Furthermore, an average throughput-delay tradeoff can be achieved by the proposed solution.

Diffusion Based Molecular Communication: Principle, Key Technologies, and Challenges

Abstract: Molecular communication (MC) is a kind of communication technology based on biochemical molecules for internet of bio-nano things, in which the biochemical molecule is used as the information carrier for the interconnection of nano-devices. In this paper, the basic principle of diffusion based MC and the corresponding key technologies are comprehensively surveyed. In particular, the state-of-the-art achievements relative to the diffusion based MC are discussed and compared, including the system model, the system performance analysis with key influencing factors, the information coding and modulation techniques. Meanwhile, the multi-hop nano-network based on the diffusion MC is presented as well. Additionally, given the extensiveness of the research area, open issues and challenges are presented to spur future investigations, in which the involvement of channel model, information theory, self-organizing nano-network, and biochemical applications are put forward.

Wireless communication networking method and device based on information perception

Abstract: The embodiment of the invention provides a wireless communication networking method and device based on information perception and relates to the technical field of wireless communication. The method includes the steps that parameters of each protocol layer in a network are obtained periodically by the wireless communication networking device based on information perception, wherein each protocol layer sequentially comprises a physical layer, a media access control layer, a network layer, a network slicing and arranging control layer and a business layer from bottom to top; the parameters are subjected to data mining to achieve information perception, and information perception results are obtained; according to the information perception results, the varieties of network slices are determined, and a communication networking configuration scheme of each variety of network slices is determined. Whether or not each variety of network slices needs to be adjusted by periodically obtaining the information perception results, and if yes, the configuration schemes of the network slices are adjusted. Compared with existing wireless communication networking methods, the method can meet various business requirements, improve the user service perception capacity of the network and improve the utilization of network resources.

Fully Exploiting Cloud Computing to Achieve a Green and Flexible C-RAN

Abstract: By merging cloud computing into the RAN, C-RAN has been foreseen as a prospective 5G wireless systems architecture. Due to the innovative move of migrating the baseband processing functionalities to the centralized cloud baseband unit pool, C-RAN is anticipated to reduce energy consumption significantly to be a green RAN. Moreover, with the cloud-based architecture, lots of new functionalities and RAN designs are ready to be incorporated, which redefines the RAN as a flexible RAN. In this article, we review the recent advances of exploiting cloud computing to form a green and flexible C-RAN from two cloud-based properties: centralized processing and the software-defined environment. For the centralized processing property, we include coordinated multipoint and limited fronthaul capacity, multicasting, and CSI issues in C-RAN. For the software-defined environment property, we summarize elastic service scaling, functionality splitting, and functionality extension. We also include some of our recent research results and discuss several open challenges.

Economical Energy Efficiency: An Advanced Performance Metric for 5G Systems

Abstract: The performance of 5G wireless communication systems will be significantly affected by edge cache and transport network. These emerging components bring substantial costs of placement and utilization, and the evaluation of the cost impact is beyond the capability of traditional performance metrics, including spectral efficiency and energy efficiency. In this article, economical energy efficiency (E3) is proposed, whose core idea is to take SE/EE and cost into account to evaluate comprehensive gains when different kinds of advanced technologies are used in 5G systems. The E3 results are shown when the transport network and edge cache are separately or jointly used. Open issues in terms of modeling the cost, E3 optimization-based radio resource allocation, and E3 optimization for the Internet of Things are identified as well.

Joint resource allocation and caching placement for network slicing in fog radio access networks

Abstract: To satisfy diverse use cases and business models in fifth generation (5G) wireless communication, network slicing in fog radio access networks (F-RANs) is proposed, which provides a cost efficient networking in a convenient way. However, the resource management for network slices is challenging, especially when the edge caching is utilized to alleviate the fronthaul burden and reduce the delay. In this paper, we investigate the joint sub-carrier allocation and caching placement for two network slices, and an average delay optimization problem for one slice with user data rate guarantee for the other slice is formulated, which can be solved by a two-step iterative algorithm. The core idea is to optimize decoupled variables iteratively via Hungarian method, linear integer programming, and geometric programming(GP) with the help of decomposition. Simulation results reveal a fast convergence speed and a near-optimal performance of the proposed algorithm.

Delay-Aware Uplink Fronthaul Allocation in Cloud Radio Access Networks

Abstract: In cloud radio access networks (C-RANs), the baseband units and radio units of base stations are separated, which requires high-capacity fronthaul links connecting both parts. In this paper, we consider the delay-aware fronthaul allocation problem for C-RANs. The stochastic optimization problem is formulated as an infinite horizon average cost Markov decision process. To deal with the curse of dimensionality, we derive a closed-form approximate priority function and the associated error bound using perturbation analysis. Based on the closed-form approximate priority function, we propose a low-complexity delay-aware fronthaul allocation algorithm solving the per-stage optimization problem. The proposed solution is further shown to be asymptotically optimal for sufficiently small residual interference. Finally, the proposed fronthaul allocation algorithm is compared with various baselines through simulations, and it is shown that significant performance gain can be achieved.

Performance analysis and optimal signal detection for amplify-and-forward relay in diffusion-based molecular communication systems

Abstract: Molecular communication (MC) is a promising paradigm which utilizes molecules to implement communication among nano-machines. Due to short transmit range, amplify-and-forward (AF) relay is used to extend the communication distance and achieve a reliable remote communications in diffusion-based MC systems. In this paper, an AF relay strategy in diffusion-based MC systems is researched, in which the capture probability function is formulated by searching for the essence of AF, and the distribution of the number of molecules received at the receiver is modeled as an approximate normal distribution. The impact of AF relay location and the signal detection threshold on the system performance are exploited, and the optimal signal detection threshold is derived by using the maximum a posteriori probability. The simulation results show that the proposed AF relay strategy can effectively improve the system performance, and the optimal relay location can achieve the best performance.

Performance analysis for cooperative relaying in diffusion-based molecular communication

Abstract: Molecular communications (MC) is a promising paradigm which utilizes molecules to implement communication among nano-machines. Cooperative relaying strategies can be used to extend communication distance and achieve reliable remote communications. In this paper, a two-hop diffusion-based molecular communication with three nano-machine nodes is considered. To mitigate the drawback of traditional cooperative relay strategies in diffusion-based MC, such as decode-and-forward and amplify-and-forward, an opportunistic cooperative relaying strategy is proposed. Closed-form expressions of bit error probability for these three cooperative relaying strategies and two molecular type schemes are derived. Analytical and simulation results show that opportunistic decode-amplify-forward relaying strategy with different type of molecules modulation achieve the best gain to implement reliable remote communications.

Secrecy-Optimized Resource Allocation for Device-to-Device Communication Underlaying Heterogeneous Networks

Abstract: Device-to-device (D2D) communications recently have attracted much attention for its potential capability to improve spectral efficiency underlaying the existing heterogeneous networks (HetNets). Due to no sophisticated control, D2D user equipments (DUEs) themselves cannot resist eavesdropping or security attacks. It is urgent to maximize the secure capacity for both cellular users and DUEs. This paper formulates the radio resource allocation problem to maximize the secure capacity of DUEs for the D2D communication underlaying HetNets which consist of high power nodes and low power nodes. The optimization objective function with transmit bit rate and power constraints, which is non-convex and hard to be directly derived, is firstly transformed into matrix form. Then the equivalent convex form of the optimization problem is derived according to the Perron-Frobenius theory. A heuristic iterative algorithm based on the proximal theory is proposed to solve this equivalent convex problem through evaluating the proximal operator of Lagrange function. Numerical results show that the proposed radio resource allocation solution significantly improves the secure capacity with a fast convergence speed.

Recent Advances in Fog Radio Access Networks: Performance Analysis and Radio Resource Allocation

Abstract: As a promising paradigm for the fifth generation wireless communication (5G) system, the fog radio access network (F-RAN) has been proposed as an advanced socially-aware mobile networking architecture to provide high spectral efficiency (SE) while maintaining high energy efficiency (EE) and low latency. Recent advents are advocated to the performance analysis and radio resource allocation, both of which are fundamental issues to make F-RANs successfully rollout. This article comprehensively summarizes the recent advances of the performance analysis and radio resource allocation in F-RANs. Particularly, the advanced edge cache and adaptive model selection schemes are presented to improve SE and EE under maintaining a low latency level. The radio resource allocation strategies to optimize SE and EE in F-RANs are respectively proposed. A few open issues in terms of the F-RAN based 5G architecture and the social-awareness technique are identified as well.

Energy-efficient resource allocation in delay-aware wireless virtualized networks

Abstract: To satisfy various quality-of-service (QoS) requirements of different services in future fifth generation wireless networks, wireless virtualized network (WVN) architectures have been proposed to concurrently fulfill diversified service demands via network slicing technologies. Considering the dynamic characteristics of wireless channels and user traffic, the resource allocation problem in a WVN with two network slices is formulated, aiming to optimize energy efficiency (EE) while constrained by the power consumption, queue stability and QoS requirements. The formulated nonconvex problem is transformed based on the Lyapunov optimization approach and solved via the Lagrange dual decomposition method and a weighted minimum mean square error approach, with an arbitrarily near-optimal solution obtained. Simulation results demonstrate that there is an EE-delay tradeoff with V being the control parameter and a balance between the EE and the queue delay can be achieved on demand by tuning V flexibly.

Collaborative UFH-based anti-jamming broadcast with learning

Abstract: Uncoordinated frequency hopping (UFH) technique has been developed to address smart jamming attacks in wireless networks, in which the receiver does not need to know the pre-shared physical-layer secret keys such as the frequency hopping pattern with the transmitter and thus cannot be efficiently blocked by smart jammers that eavesdrop the public control channel of the network. However, collaborative UFH-based broadcast (CUB) that exploits both the spatial and frequency diversity still suffers from a low communication efficiency, because the probability that a receiver happens to use the same channel with a transmitter with a random channel selection is very low. In this paper, we propose a collaborative UFH-based broadcast scheme based on reinforcement learning to further improve the communication efficiency against smart jamming. More specifically, by applying Q-learning algorithm, a radio node can achieve the optimal transmit strategy via trials in the repeated game without being aware of the jamming model and the network model. Simulation results show that the Q-learning based anti-jamming broadcast can significantly decrease the broadcast delay and reduce the total energy cost, compared with the CUB scheme.

A hierarchical approach for terminal awareness in fog radio access networks

Abstract: Fog radio access networks have been proposed as a promising architecture to support diverse scenarios, which provide edge computing capabilities. Meanwhile, terminal awareness can enable wireless networks respond proactively and intelligently. However, due to the acquisition of raw data and the execution of computation tasks, centralized approaches for terminal awareness can put heavy burdens on fronthaul and cloud. To alleviate these burdens and infer the terminal type precisely, a hierarchical approach integrating multiple non-linear Support Vector Machine (SVM) classifiers is proposed. The core idea is that the raw data processing and the training of classifiers are shifted to fog access points (F-APs), while the server in the cloud is only responsible for training task allocation and training data dissemination. Further, the training task allocation is formulated as an integer problem to minimize the total training time, which is solved by a branch and bound based method and a low complexity heuristic method. Simulation result shows that the heuristic method can achieve optimal performance, and the impacts of the heterogeneity in F-AP computing capabilities on the training time and the task allocation are demonstrated and analyzed.

Economical-energy-efficient resource allocation for cache-enabled cloud radio access networks

Abstract: Taking advantage of both edge cache (EC) and coordinated processing, cache-enabled cloud radio access networks (C-RANs) have shown good potential to become a solution for future wireless networks. With the demand that the network needs to be fast, green and affordable, it has become an issue both practical and challenging to design a resource allocation strategy that leads to a proper balance among throughput, energy consumption and the cost of the cache-enabled C-RAN system. In this article, the economical energy efficiency (E3) metric is adopted as a feasible way to comprehensively evaluate the impacts in these aspects. Based on the constraints of both remote radio heads (RRHs) association and fronthaul capacity, an E3 optimization problem with the resource allocation for cache-enabled C-RAN is formulated as a nonconvex objective function. An algorithm is devised to achieve the sub-optimal solution for the problem afterwards. Simulation results confirm that the proposed method can significantly enhance the E3 performance of the cache-enabled C-RAN system.

Channel Matrix Sparsity With Imperfect Channel State Information in Cloud Radio Access Networks

Abstract: Channel matrix sparsification is considered as a promising approach to reduce the progressing complexity in large-scale cloud radio access networks based on ideal channel condition assumption. In this paper, the research of channel sparsification is extend to practical scenarios in which the perfect channel state information (CSI) is not available. First, a tractable lower bound of signal-to-interference-plus-noise ratio (SINR) fidelity, which is defined as a ratio of SINRs with and without channel sparsification, is derived to evaluate the impact of channel estimation error. Based on the theoretical results, a Dinkelbach-based algorithm is proposed to achieve the global optimal performance of channel matrix sparsification based on the criterion of distance. Finally, all these results are extended to a more challenging scenario with pilot contamination. Finally, simulation results are shown to evaluate the performance of channel matrix sparsification with imperfect CSIs and verify our analytical results.

Network Slicing in Fog Radio Access Networks: Issues and Challenges

Abstract: Network slicing has been advocated by both academia and industry as a cost-efficient way to enable operators to provide networks on an as-a-service basis and meet the wide range of use cases that the fifth generation wireless network will serve. The existing works on network slicing are mainly targeted at the partition of the core network, and the prospect of network slicing in radio access networks should be jointly exploited. To solve this challenge, an enhanced network slicing in fog radio access networks (F-RANs), termed as access slicing, is proposed. This article comprehensively presents a novel architecture and related key techniques for access slicing in F-RANs. The proposed hierarchical architecture of access slicing consists of centralized orchestration layer and slice instance layer, which makes the access slicing adaptively implement in an convenient way. Meanwhile, key techniques and their corresponding solutions, including the radio and cache resource management, as well as the social-aware slicing, are presented. Open issues in terms of standardization developments and field trials are identified.

An Evolutionary Game for User Access Mode Selection in Fog Radio Access Networks

Abstract: The fog radio access network (F-RAN) is a promising paradigm for the fifth generation wireless communication systems to provide high spectral efficiency and energy efficiency. Characterizing users to select an appropriate communication mode among fog access point (F-AP), and device-to-device (D2D) in F-RANs is critical for performance optimization. Using evolutionary game theory, we investigate the dynamics of user access mode selection in F-RANs. Specifically, the competition among groups of potential users space is formulated as a dynamic evolutionary game, and the evolutionary equilibrium is the solution to this game. Stochastic geometry tool is used to derive the proposals' payoff expressions for both F-AP and D2D users by taking into account the different nodes locations, cache sizes as well as the delay cost. The analytical results obtained from the game model are evaluated via simulations, which show that the evolutionary game based access mode selection algorithm can reach a much higher payoff than the max rate based algorithm.