Accurate downlink channel state information (CSI) is required to be fed back to the base station (BS) in frequency division duplexing (FDD) massive multiple-input multiple-output (MIMO) systems in order to achieve maximum antenna diversity and multiplexing. However, downlink CSI feedback overhead scales with the number of transceiver antennas, a major hurdle for practical deployment of FDD massive MIMO systems. To solve this problem, we propose a compressive sampled CSI feedback method based on deep learning (SampleDL). In SampleDL, the massive MIMO channel matrix is sampled uniformly in time/frequency dimension before being fed into neural networks (NNs), which will reduce the computational resource/time at user equipment (UE) as well as enhance the CSI recovery accuracy at the BS. Both theoretical analysis and normalized mean square errors (NMSE) results confirm the advantages of the proposed method in terms of time complexity and recovery accuracy. Besides, a suitable CSI feedback period is explored by link level simulations, which aims to further reduce the overhead of CSI feedback without degrading the communication quality.

The advantages of massive multiple-input multiple-output (MIMO) techniques depend heavily on the accuracy of channel state information (CSI). In frequency division duplexing (FDD) massive MIMO systems, the user equipment (UE) needs to feed downlink CSI back to the base station (BS) through the feedback link. The excessive feedback overheads and low reconstruction accuracy are the main obstacles for actual deployment of FDD massive MIMO systems. In recent years, deep learning (DL) has been widely used to address the above problems. In this letter, we propose a neural network by utilizing the self-attention learning and dense refine (SALDR), which improves the accuracy of CSI feedback. Furthermore, a unified decoder named SALDR-U is designed to realize different compression ratios for CSI feedback without changing any parameter. Simulation results show that the proposed SALDR and SALDR-U outperform the state-of-the-art network in terms of accuracy and overhead of CSI feedback. The source code for all the experiments is available at GitHub.The code of this letter can be downloaded from GitHub link: https://github.com/XS96/SALDR.

Unmanned aerial vehicle (UAV) assisted wireless caching networks (WCN) have been recognized as a promising way to reduce the network load and improve the energy efficiency in the sixth generation (6 G) communication systems. Aiming to improve spectrum efficiency and system capacity, we apply non-orthogonal multiple access (NOMA) in UAV-assisted WCN to serve multiple users on the same spectrum simultaneously and propose the cross-layer resource allocation strategy including the scheduling of UAVs, the grouping of users, and the allocation of power. First, the $\rho$-K-means algorithm is proposed to assign users to multiple clusters and deploy UAVs according to the distance from UAVs to the base station in the UAV deployment layer. Then, the base station broadcasts the popular files to UAVs via NOMA in the content placement layer. Based on the existing fixed power allocation strategy, we propose a statistic quality of service (QoS) based fixed (SQF) power allocation method to take the statistic QoS of the popular files into consideration and improve the energy efficiency through introducing the discount factor. On the basis of SQF, an instantaneous QoS based adaptive (IQA) strategy allocates power according to the instantaneous QoS of the popular files to reduce the file outage probability. Furthermore, we propose an improved method that is a cross-layer based optimal (CLO) power allocation strategy to maximize the system hit probability. Finally, in the content delivery layer, users in each cluster are grouped according to the channel gain from users to UAVs. In addition, each UAV serves two users on the same time-frequency resource block based on the cognitive radio inspired power allocation for the NOMA user pairs. Simulation results confirm that the proposed $\rho$-K-means algorithm and CLO strategy reduce the file outage probability and improve the hit probability.

This paper proposes a decentralized automatic modulation classification (DecentAMC) method using light network and model aggregation. Specifically, the lightweight network is designed by separable convolution neural network (S-CNN), in which the separable convolution layer is utilized to replace the standard convolution layer and most of the fully connected layers are cut off, the model aggregation is realized by a central device (CD) for edge device (ED) model weights aggregation and multiple EDs for ED model training. Simulation results show that the model complexity of S-CNN is decreased by about 94% while the average CCP is degraded by less than 1% when compared with CNN and that the proposed AMC method improves the training efficiency when compared with the centralized AMC (CentAMC) using S-CNN.

Acquisition of downlink channel state information (CSI) is an important procedure performed at the base station (BS) for high quality wireless communication in frequency division duplexing (FDD) communication system. Generally, the downlink CSI is fed back to the BS through the user equipment (UE). Compared with traditional methods, neural network (NN) can effectively compress the downlink CSI, thus greatly reducing the feedback overhead. However, the generalization of the NN is poor, hence it is necessary to train a NN from scratch whenever there is a change in the wireless channel environment. Nevertheless, training a NN this way requires huge data and time cost in 5G massive MIMO systems. In this paper, deep transfer learning (DTL) is proposed to solve the problem of high training cost of the downlink CSI feedback NN. In a new wireless environment, our proposed technique utilises relatively small number of samples to fine-tune a pre-trained model, in order to obtain a new model with low training cost. The performance of this model is shown to be comparable with that of the NN trained with large samples. Experiment results demonstrate the effectiveness and superiority of the proposed method.

In massive multiple input multiple output (MIMO) systems, the base station (BS) requires channel state information (CSI) to better utilize the available spatial diversity and multiplexing gains. However, in frequency division duplex (FDD) systems, user equipment (UE) needs to keep on feeding downlink CSI back to the BS, thereby consuming precious bandwidth resources. In this paper, we propose a deep learning (DL) based downlink CSI limited feedback scheme, called FullyConv, which is composed of all convolutional layers to compress and decompress the downlink CSI. FullyConv will improve reconstruction accuracy and robustness as well as reduce the time and space complexity, thus enhancing the system feasibility. Experimental results demonstrate that the FullyConv has a gain of nearly 5 dB compared to baseline. The performance of the FullyConv degrades slightly in the noisy uplink channel, which shows the robustness of FullyConv. Meanwhile, the complexity of the model composed of time complexity and space complexity is significantly reduced.

Deep learning (DL) is an efficient method for botnet attack detection. However, the volume of network traffic data and memory space required is usually large. It is, therefore, almost impossible to implement the DL method in memory-constrained Internet-of-Things (IoT) devices. In this article, we reduce the feature dimensionality of large-scale IoT network traffic data using the encoding phase of long short-term memory autoencoder (LAE). In order to classify network traffic samples correctly, we analyze the long-term inter-related changes in the low-dimensional feature set produced by LAE using deep bidirectional long short-term memory (BLSTM). Extensive experiments are performed with the BoT-IoT data set to validate the effectiveness of the proposed hybrid DL method. Results show that LAE significantly reduced the memory space required for large-scale network traffic data storage by 91.89%, and it outperformed state-of-the-art feature dimensionality reduction methods by 18.92–27.03%. Despite the significant reduction in feature size, the deep BLSTM model demonstrates robustness against model underfitting and overfitting. It also achieves good generalisation ability in binary and multiclass classification scenarios.

Future denser air-ground vehicle networks (AGVNs) face challenges such as resource allocation, mobility management, secure transmission, and so on. At the same time, surveillance is a must for modern air traffic management. This motivates us to find opportunities in the aerial vertical by forming a conceptual surveillance plane for aerial vehicles. In this article, we propose an enhanced software-defined network architecture where the surveillance plane can provide local and global surveillance information to macro stations, acting as a side system for the communication links. We review air- ground communications and, by summarizing challenges and opportunities, propose the enhanced architecture of side-information-assisted networks in detail. We then present how we obtain, organize, manage, and utilize the local and global side information by a so-called aviation data lake (ADL). The data lake can be easily connected with advanced machine learning schemes and, thus, provide timely, context-aware metrics and predictions.

As exploitation of low and medium airspace for air traffic management (ATM) is gaining more attention, aerial vehicles’ security issues pose a major challenge to the air–ground-integrated vehicle networks (AGIVNs). Traditional surveillance technology lacks the capacity to support the intensive ATM of the future. Therefore, an advanced automatic-dependent surveillance-broadcast (ADS-B) technique is applied to track and monitor aerial vehicles in a more effective manner. In this article, we propose a grouping-based conflict detection algorithm based on the preprocessed ADS-B data set, and analyze the experimental results and visualize the detected conflicts. Then, in order to further improve flight safety and conflict detection, the trajectories of the aerial vehicles are predicted based on machine learning-based algorithms. The results are fed into the conflict detection algorithm to execute conflict prediction. It was shown that the trajectory prediction model using long short-term memory (LSTM) can achieve better prediction performance, especially when predicting the long-term trajectory of aerial vehicles. The conflict detection results based on the trajectory prediction methods show that the proposed scheme can make it possible to detect whether there would be conflicts within seconds.

Energy efficiency (EE) is an important performance metric in communication systems. However, to the best of our knowledge, the energy-efficient resource allocation (RA) problem in non-orthogonal multiple access enabled backscatter communication networks (NOMA-BackComNet) comprehensively considering the user’s quality of service (QoS) has not been investigated. In this letter, we present the first attempt to solve the EE-based RA problem for NOMA-BackComNet with QoS guarantee. The objective is to maximize the EE of users subject to the QoS requirements of users, the decoding order of successive interference cancellation and the reflection coefficient (RC) constraint, where the transmit power of the base station and the RC of the backscatter device are jointly optimized. To solve this non-convex problem, we develop a novel iteration algorithm by using Dinkelbach’s method and the quadratic transformation approach. Simulation results verify the effectiveness of the proposed scheme in improving the EE by comparing it with the other schemes.

In beyond fifth-generation (B5G) era, massive multiple-input multiple-output (M-MIMO) will be a key technology to offer higher network capacities. Due to the different frequency of uplink and downlink channels in FDD systems, the channel state information (CSI) feedback from user terminal to the base station is necessary, but this reduces the spectrum efficiency. This letter proposes a deep learning based solution to predict the downlink CSI in frequency division duplex (FDD) systems, which is termed as complex-valued three dimensional convolutional neural network (CV-3DCNN). The proposed network uses a complex-valued neural network in complex domain to deal with the complex CSI matrices, and adopts three-dimensional convolution operations for feature extraction. The proposed scheme aims to make full use of the hidden information of the complex matrices of the CSI data, and to minimize information loss caused by data processing. The experimental results demonstrate that the proposed architecture can improve accuracy of the downlink CSI prediction by approximately 6 dB.

In this paper, the tradeoff between spectrum efficiency (SE) and energy efficiency (EE) is investigated in terms of interference management and power allocation for heterogeneous networks (HetNets) with non-orthogonal multiple access (NOMA). The EE and SE tradeoff is modeled as a multi-objective problem (MOP) under the maximum power and quality of service (QoS) constraints, which is non-convex. The MOP is relaxed into a convex single objective problem (SOP) by adopting a weighted sum strategy with the hypograph transformation. The SOP is solved in two steps. In the first step, we propose a power allocation technique based on non-cooperative (NC) game for EE and SE in NOMA HetNets. In the proposed NC game, the macro base station (MBS) and the small BSs (SBSs) compete with an equal priority in order to optimize their transmit powers towards maximizing the weighted sum of SE and EE. In the second step, a closed-form formula is proposed to control the power allocated to users taking into account both QoS constraint and successive interference cancellation (SIC) condition. From simulations, the proposed technique can, in some dedicated settings, considerably improve the tradeoff between EE and SE over conventional techniques.

In this paper, the tradeoff between spectrum efficiency (SE) and energy efficiency (EE) is investigated in terms of interference management and power allocation for heterogeneous networks (HetNets) with non-orthogonal multiple access (NOMA). The EE and SE tradeoff is modeled as a multi-objective problem (MOP) under the maximum power and quality of service (QoS) constraints, which is non-convex. The MOP is relaxed into a convex single objective problem (SOP) by adopting a weighted sum strategy with the hypograph transformation. The SOP is solved in two steps. In the first step, we propose a power allocation technique based on non-cooperative (NC) game for EE and SE in NOMA HetNets. In the proposed NC game, the macro base station (MBS) and the small BSs (SBSs) compete with an equal priority in order to optimize their transmit powers towards maximizing the weighted sum of SE and EE. In the second step, a closed-form formula is proposed to control the power allocated to users taking into account both QoS constraint and successive interference cancellation (SIC) condition. From simulations, the proposed technique can, in some dedicated settings, considerably improve the tradeoff between EE and SE over conventional techniques.