In this paper, we investigate the fundamental trade-off between communication and sensing performance ofad hocintegrated sensing and communication (ISAC) wireless networks. Specifically, we consider thatnnodes are randomly located in an extended network with areanand transmit ISAC signals. Under the pure path loss channel gain model and the condition that the transmission power scales according to the communication distance, we fully characterize the optimal scaling law trade-off between throughput and sensing distance by proposing an achievable scheme and proving its converse. Our results can be interpreted as follows: by reducing the throughput by a factor of a function ofn, the sensing range order improves according to the same function ofn, raised to the power of the ratio between the path loss factors in communication and sensing. We prove that the same result also holds true for ISAC networks with random fading, despite the uncertainty on the connectivity and power level created by random fading. In addition, we show that the scaling law tradeoff cannot be improved by allowing the transmission power and communication distance to scale freely. To the best of our knowledge, this is the first work formally formulating and characterizing the communication and sensing performance scaling law tradeoff ofad hocISAC networks.

To realize integrated sensing and communication (ISAC), multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) ISAC systems have been widely discussed. However, current beamforming designs for multi-user MIMO-OFDM ISAC systems are mostly based on iterative optimization-based approaches, which commonly induce high runtime complexity. To resolve this issue, this paper first investigates the relevant beamforming design approach that is based on iterative weighted minimum mean square error (WMMSE) problem and successive convex approximation (SCA). Then, a deep unfolding learning-based beamforming design framework is proposed via unfolding the iterative WMMSESCA procedure into a neural network (NN)-based architecture. To efficiently train the proposed framework, a loss function for unsupervised learning is proposed. Also, approaches that can enhance the training efficiency are discussed. Simulation results demonstrate that our proposed approach can outperform existing methods in ISAC performance, while significantly reducing the computation time.

The multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM)-based integrated sensing and communication (ISAC) system has been intensively investigated in the past years, as it is a promising candidate for realizing ISAC in next-generation wireless systems. However, even though it is clear that OFDM-based systems could suffer from high peak-to-average power ratio (PAPR), the design of low-PAPR multi-user MIMI-OFDM ISAC systems has not been well-explored. To fill this gap, this paper investigates the low-PAPR multi-user MIMI-OFDM ISAC system design. Specifically, by using radar ambiguity function and symbol-level precoding concept, this paper first formulates a low-PAPR multi-user MIMI-OFDM ISAC system design problem that optimizes the sensing performance subject to communication performance and PAPR constraints. To solve the problem, the communication performance and PAPR constraints are first transformed to convex constraints, and then the successive convex approximation is used to derive an iterative solution approach. Simulation results show that our proposed design approach can provide the effective low-PAPR ISAC design that outperforms the reference scheme.

This work investigates the dynamic service caching, computing, and communication (3C) resource optimization for edge-cloud networks, where users, edge node, and cloud all having 3C capabilities. Based on the developed hierarchical network model, we propose a two-timescale problem aiming to minimize energy consumption under queue stability and latency and reliability constraints. A 3C resource optimization approach is then developed based on the formulated problem and Lyapunov optimization. Simulation results show that our approach can outperform the reference schemes in terms of average service delay and reliability, and can effectively provide tradeoff between energy consumption and service delay.

In this paper, we investigate the communication and sensing performance tradeoff of ad hoc integrated sensing and communication (ISAC) wireless networks. Specifically, we consider that$n$nodes are randomly located in an extended network with area$n$and transmit ISAC signals. Our goal is to answer the following questions: what is the tradeoff between the throughput and sensing range of an ISAC network and how does it scale with the network size or node numbers? Under the condition that the transmission power scales according to the communication distance, we fully characterize the scaling law tradeoff between throughput and sensing distance by proposing an achievable scheme and proving its converse. Interestingly, our results reveal that by reducing the throughput by a factor of a function of$n$, the sensing range order improves according to the same function of$n$, raised to the power of the ratio between the path loss factors in communication and sensing. We also show that the scaling law tradeoff cannot be improved by allowing the transmission power and communication distance to scale differently. To the best of our knowledge, this is the first work formally formulating and characterizing the communication and sensing performance scaling law tradeoff of ad hoc ISAC networks.

This paper explores the joint optimization of caching, computing, and communication (3 C) in multiple-input multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) wireless networks. It also incorporates the use of reconfigurable intelligent surfaces (RISs) to enhance communication performance. Given the different timescales for updating various resource allocations, we propose a two-timescale joint 3C optimization approach aimed at minimizing network latency. In this approach, task offloading, computing power allocation, precoding, and RIS optimization occur on the small timescale, while caching is updated on the large timescale. Our simulation results demonstrate that this approach effectively reduces network latency and outperforms the reference schemes.

Synthetic Aperture Radar (SAR) has emerged as an indispensable tool for maritime surveillance, providing reliable all-weather, day-and-night imaging capabilities. However, automated ship recognition in SAR imagery presents significant challenges, particularly due to variations in patch sizes, with small-size SAR image patches posing the greatest difficulties. To address this challenge, we propose Radiometric Resolution and Scale-aware Masked Autoencoder (RSMAE), a novel framework designed for SAR ship recognition. Our method incorporates three key innovations: (1) a scale-aware augmentation that adapts to images of varying image sizes for masked image modeling, enabling the model to learn multi-scale features and reconstruct fine-grained details lost during upscaling; (2) a foreground-background balanced masking strategy that independently handles the ship region and its surrounding region, ensuring the ship area is neither over-masked nor under-masked; and (3) a radiometric resolution-aware (RadRe-aware) reweighting mechanism that leverages SAR-specific radiometric characteristics to enhance reconstruction of challenging samples. Experimental results on the OpenSARShip dataset demonstrate that the proposed RSMAE consistently outperforms state-of-the-art methods by at least 2.07% in terms of recognition accuracy. These findings highlight the robustness and efficiency of the proposed RSMAE, making it a compelling solution for SAR ship recognition tasks.

This work examines the impact of recommendation on both the shaping of user preference and the caching decisions at wireless edge devices. While most studies focus on the optimization from a network perspective, we investigate the joint caching and recommendation optimization for wireless device-to-device (D2D) networks from both network and user perspectives and identify their key differences. To achieve this goal, optimization problems for network offloading and user offloading probabilities as well as their tradeoff are first formulated. Two types of preference-shaping models are considered. The first type assumes that the user preference can be arbitrarily shaped whereas the second type only selectively enhances the users’ original preference. The proposed optimization problems are solved using alternating optimization where the users’ caching variables and the recommendation variables are optimized in turn until convergence. The corresponding convergence and complexity analyses are also provided. Extensive simulations are conducted to validate the efficacy of the solution approaches. Results show that designs obtained from different perspectives can lead to very different performance behaviors, and the difference is especially large when there is little similarity among users’ preferences.

To address spectrum scarcity and achieve the perceptive network, integrated sensing and communication (ISAC) systems have been extensively studied. However, research on multi-user ISAC systems utilizing multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) architectures remains incomplete. This paper aims to bridge this gap by exploring digital and hybrid beamforming designs for multi-user MIMO-OFDM ISAC systems. Specifically, the paper investigates the impact of beamforming on the range-Doppler domain of radar sensing and system capacity, and formulates a corresponding design problem. The paper then addresses the capacity maximization problem by converting it into a weighted minimum mean square error problem and employs successive convex approximation (SCA) to develop an iterative approach for digital beamforming design. Similarly, the paper formulates a hybrid beamforming design problem and proposes a design approach based on the alternating direction method of multipliers (ADMM) and SCA. The complexity of the proposed approaches is analyzed. Simulation results demonstrate that our approaches are effective and can outperform existing approaches in the literature.

The investigation of the joint radar and communication (RadCom) systems has drawn attention in recent years. However, the relevant investigation of the multiple-input multiple-output (MIMO)-orthogonal frequency division multiplexing (OFDM) based joint RadCom system is incomplete, especially for the robust beamforming design. To fill this gap, by first formulating a robust design problem that jointly considers the estimation of ranges, Doppler velocities, and angles of targets under uncertainty along with the system capacity, this paper studies the robust beamforming design for the MIMO-OFDM joint RadCom system. Then, since the design problem can be reformulated as a semidefinite programming with rank-1 constraint, we propose using the semidefinite relaxation to solve the problem. Simulation results show that our proposed design can outperform the reference methods in the literature.

To resolve the spectrum scarcity and realize the perceptive network, integrated sensing and communication (ISAC) systems have been widely studied. However, the study of the multi-user ISAC system based on the multiple-input multiple-output (MIMO)-orthogonal frequency division multiplexing (OFDM) architecture is far from complete. Thus, this paper aims to fill this gap by investigating the beamforming design for multi-user MIMO-OFDM ISAC systems. By studying the beamforming effect on the range-Doppler domain of the radar sensing and the system capacity, a design problem is formulated. Then, by converting the capacity maximization to the weighted minimum mean square error minimization along with the use of successive convex approximation, an iterative approach is proposed to solve the design problem. Simulation results show that our design approach can outperform the reference methods in the literature.

To facilitate massive multiple-input multiple-output systems and millimeter-wave communications, reconfigurable intelligent surface (RIS) has drawn significant attention. On the other hand, to serve computation- and data-intensive applications, the optimization jointly considering caching, computing, and communication (3C) resources has also been widely discussed. To fill the gap that the optimization considering 3C has not been well-explored in RIS-aided networks, this paper investigates the joint optimization of computation and communication (2C) resources and RIS design in multi-cell multi-user wireless caching networks. Based on the formulated latency minimization problem, we propose a joint 2C and RIS optimization approach. Our approach solves the latency minimization problem by first decomposing it into several subproblems, and then iteratively solving the subproblems until convergence. Computer simulations show that our approach can effectively improve the network performance and outperform the reference schemes.

To satisfy the demands of emerging mobile applications, the design and optimization for wireless edge networks jointly considering the caching, computing, and communication (3C) functionalities has drawn attention in the past years. However, robust approaches to address network uncertainty have been underexplored. To bridge this gap, this paper investigates the robust optimization technique for computing and communication resources in wireless caching networks, considering channel uncertainty, task workload uncertainty, and resource uncertainty. We formulate a robust latency minimization problem aimed at jointly optimizing offloading decisions, bandwidth allocation, transmit power allocation, and computing power allocation. Subsequently, we propose an iterative algorithm to solve this problem. Simulation results show that our proposed approach is more robust than the reference schemes.

In recent years, the integration of sensing and communication has gained significant attention. In this context, the adoption of the multiple-input multiple-output (MIMO)orthogonal frequency division multiplexing (OFDM) dual-functional radar and communication (DFRC) system is promising. However, comprehensive exploration of the MIMO-OFDM DFRC system remains incomplete, particularly concerning beamforming design across various radar transmission modes. To address this gap, this paper investigates the joint design of beamforming and subcarrier allocation for MIMO-OFDM DFRC systems. We formulate an ambiguity function-based radar and communication joint design problem. Subsequently, we propose a solution approach that solves the problem by iteratively solving the beamforming and subcarrier allocation subproblems. Simulations are conducted to assess the efficacy of the proposed design approach. The results demonstrate that our approach can effectively adapt to different radar modes and outperform reference methods.

The use of mobile integrated access and backhaul (mIAB) is promising due to its high cost-effectiveness and fast deployment. However, the study of the deployment optimization for mIAB nodes is still incomplete, especially for air-ground integrated networks. To fill the gap, this paper aims to develop non-model-based deployment approaches for both unconstrained and constrained deployment problems. To this end, approaches with the zeroth order optimization techniques are proposed for solving deployment problems along with the use of deep neural network (DNN) surrogate models to help predicting network performance. In addition, effective training and design approaches for DNN surrogate models are provided by using data augmentation and active learning techniques. Simulation results show that our proposed approaches are effective and can outperform reference schemes.

The problem of computing batches of matrix multiplications in distributed computing systems with stragglers is studied. Unlike existing works in the literature, the matrices in a batch are assumed to be sparse, and the sparsity levels for matrices in different batches can be different. A novel coding scheme, called generalized sparse code (GSC), is proposed, in which the matrices are partitioned into smaller chunks that are re- grouped and encoded by respective sparse codes. The expected runtime of the proposed GSC scheme is analyzed, based on which a task assignment problem associated with the proposed GSC is formulated and solved. The solution follows the reverse water-filling principle, by which an efficient worker assignment algorithm whose worst-case time complexity equal to the total number of workers can be developed. Simulation results validate the advantage of the proposed GSC over four existing schemes, including entangled polynomial codes (EP), generalized cross-subspace alignment (GCSA), Lagrange coded computing (LCC) codes and factored Luby transform (FLT) codes at all sparsity levels. As a potential application of the proposed GSC, the problem of computing a batch of matrix multiplications with similarity is discussed.

Fault diagnosis for wireless networks is commonly conducted by human experts. However, such manual diagnosis becomes much less feasible due to the growing complexity of wireless networks. To resolve this issue, automatic fault diagnosis has been studied in self-organizing networks (SONs). However, existing works mostly consider that only a single network fault could occur at a time, which might not be true in practice. Therefore, we in this paper consider that multiple faults with different levels of severity can occur simultaneously and investigate the multi-fault and severity diagnosis for SONs. We first consider using supervised learning techniques to conduct the diagnosis and propose the corresponding diagnosis neural networks. Then, since the characteristics of different network scenarios could be different and it is costly to collect labeled data for all network scenarios, we further propose an unsupervised transfer learning approach that can effectively transfer the diagnosis system from the source domain with labeled data to the target domain with unlabeled data. We conduct extensive simulations to validate our approaches. Results show that our approach can outperform all the reference approaches. Furthermore, results also show that the performance of our transfer learning-based diagnosis is close to that of the supervised learning-based diagnosis.

As appropriate deployment of unmanned aerial vehicles (UAVs) in UAV-assisted wireless networks is critical for the next-generation wireless networks, we in this paper propose centralized and decentralized UAV deployment approaches that can be applied to any UAV-assisted wireless networks for any performance metrics. The proposed centralized deployment combines the deep neural network (DNN)-based surrogate model with the zeroth-order optimization (ZOO) such that the deployment can be optimized via using the predicted network performance of the surrogate model. Since the accurate prediction of the DNN surrogate model is critical, we discuss its design and update approaches. To let UAVs update their locations for better network performance by exchanging local information with neighboring UAVs, the proposed decentralized deployment approaches combine distributed optimization frameworks with ZOO and DNN surrogate model. We conduct realistic simulations in different network scenarios with different performance metrics to evaluate our proposed approaches. Results show that our proposed can outperform all the reference schemes in all scenarios considering different performance metrics.

This work examines a novel wireless knowledge caching framework where machine learning models (i.e., knowledge) are cached at local small cell base-stations (SBSs) to facilitate both federated training and access of the models by users. We first consider a single-SBS scenario, where the caching decision, user selection, and wireless resource allocation are jointly determined by minimizing a training error bound subject to constraints on the cache capacity, the communication and computation latency, and the energy consumption. The solution is obtained by first computing the minimum achievable training loss for each model, followed by the optimization of the binary caching variables, which reduces to a 0-1 knapsack problem. The proposed framework is then extended to the multiple-SBS scenario where the user association among SBSs is further examined. We adopt a dual-ascent method where Lagrange multipliers are introduced and updated in each iteration to regularize the dependence among user selection and association. Given the Lagrange multipliers, the caching decision, user selection, resource allocation and user association variables are optimized in turn using a block coordinate descent algorithm. Simulation results show that the proposed scheme can achieve a training error bound that is lower than preference-only and random caching policies in both scenarios.

To improve unmanned aerial vehicles (UAVs)-assisted wireless networks, the appropriate deployment of UAVs is critical. However, as existing deployment approaches are commonly based on specific models that cannot be easily generalized, we in this paper propose a generalized deployment approach for UAV-assisted wireless networks by combining the deep neural network (DNN) based surrogate model with a zeroth-order optimization (ZOO). The design of the ZOO is presented. Furthermore, since the accuracy of the surrogate model is critical, we discuss its design and update approaches. We conduct practical simulations to evaluate our proposed approach. Results show that our proposed approach can significantly outperform all the reference schemes.

Since many new applications require a combination of caching, computing, and communications (3C), there is an increasing interest in the optimization of networks that jointly perform 3C. Although a number of studies for particular designs of such networks have been published, the characterization of fundamental performance limits has been rarely explored. Aiming to fill this gap, this paper provides an asymptotic outage analysis for downlink interference-limited wireless edge networks with joint 3C. In particular, we derive analytical closed-form expressions as functions of delay and 3C parameters for outage probability and optimal caching policy, respectively. Based on these, we then derive insightful outage probability expressions for the optimal caching policy and two important reference caching policies using asymptotic analysis. We provide insights and interpretations based on the derived expressions. Computer simulations validate our analytical results and insights.

To enable the intelligent transportation systems (ITSs), using radars to monitor the road environments has recently drawn attention. However, the optimization of deployment and parameters for radars in road environments has not been well-explored. To fill this gap, we in this paper investigate the joint radar deployment and parameter optimization approach for radar networks in road environments. Specifically, considering the radar FoVs, signal and interference powers, static blockage, and the impact of radar configuration, we formulate a coverage reward maximization problem that helps optimize the configuration parameters, pan angles, transmit powers, frequency band allocation, and locations of radars. Then, based on the problem, we develop an optimization approach by first decomposing the problem into subproblems, and then conducting optimization by iteratively solving the subproblems. Simulation results show that our proposed approach can outperform the reference schemes.

To further improve the performance of millimeter wave (mmWave) communications, the reconfigurable intelligent surface (RIS)-aided hybrid precoding has been studied in recent years. Following this direction, we in this paper study the joint hybrid precoder and RIS design for RIS-aided multiple-input multiple-output (MIMO)-orthogonal frequency division multiplexing (OFDM) systems. We consider two types of the phase shifters which have infinite or finite resolution, and formulate the joint design problem. Designs that can jointly optimize the hybrid precoder and RIS for two types of phase shifters are then respectively proposed. Simulation results show that our proposed designs can outperform the reference design and the design in the literature.

Performance assessment and optimization for networks jointly performing caching, computing, and communication (3C) has recently drawn significant attention because many emerging applications require 3C functionality. However, studies in the literature mostly focus on the particular algorithms and setups of such networks, while their theoretical understanding and characterization has been less explored. To fill this gap, this paper conducts the asymptotic (scaling-law) analysis for the delay-outage tradeoff of noise-limited wireless edge networks with joint 3C. In particular, assuming the user requests for different tasks following a Zipf distribution, we derive the analytical expression for the optimal caching policy. Based on this, we next derive the closed-form expression for the optimum outage probability as a function of delay and other network parameters for the case that the Zipf parameter is smaller than 1. Then, for the case that the Zipf parameter is larger than 1, we derive the closed-form expressions for upper and lower bounds of the optimum outage probability. We provide insights and interpretations based on the derived expressions. Computer simulations validate our analytical results and insights.

To realize the gain guaranteed by the massive multiple-input multiple-output (MIMO) technology with a balanced tradeoff between the performance improvement and the hardware complexity, systems based on dynamic-connected hybrid analog and digital architectures are considered. To provide a more practical design, we in this paper investigate the hybrid precoder design for MIMO-OFDM systems with dynamic-connected architecture and low-resolution phase shifters and propose a design that is suitable for the systems with and without the subarray structure. Simulation results show that our proposed design can have the performance that is almost identical to the state-of-the art designs with the infinite-resolution phase shifters.

To satisfy the increasing demands of wireless traffic in vehicular networks, how to significantly improve vehicular networks becomes a critical issue. Motivated by the potential benefits of jointly using edge-computing and edge-caching in vehicular networks, this paper considers investigating the quality-aware caching, computing and communication (3C) optimization for video delivery in vehicular networks. By incorporating the quality-awareness with 3C models, we formulate a quality-aware joint 3C optimization problem. Then, by considering the practice that the caches might be pre-determined along with the formulated 3C problem, we obtain the quality-aware joint 2C optimization problem with caching. We propose effective approaches to solve these two problems. Simulation results show that our proposed approaches can outperform the reference schemes significantly.

Performance assessment and optimization for networks jointly performing caching, computing, and communication (3C) has recently drawn significant attention because many emerging applications require 3C functionality. However, studies in the literature mostly focus on the particular algorithms and setups of such networks, while the theoretical understanding and characterization of such networks has been less explored. To fill this gap, this paper conducts the asymptotic (scaling-law) analysis for the delay-outage tradeoff of noise-limited wireless edge networks with joint 3C. In particular, we derive closed-form expressions for the optimum outage probability as function of delay and other network parameters via first obtaining the outage probability expression and then deriving the optimal caching policy. We provide insights and interpretations based on the derived expressions. Computer simulations validate our analytical results and insights.

In this paper, a machine learning-based object detection and tracking approach in radar system is proposed via using the range-angle map as the input. Specifically, by using the You Only Look Once (YOLO) for object detection and Deep Simple Online and Realtime Tracking (D-SORT) for tracking, the proposed approach can improve the detection and tracking performance, reducing the parameters needed to be manually selected, and providing more relevant information, such as the shape, size, and category of the object. We conduct the realistic simulations to evaluate the proposed approach. Results show that our proposed approach can outperform the conventional radar processing approach in terms of detection and tracking performance. Furthermore, results indicate that the object categorization of the proposed approach is accurate.

Motivated by the emerging internet of things (IoT) applications, wireless systems encounter the challenges of providing accurate indoor localization to massive IoT devices. Although the received signal strength indicator (RSSI)-based fingerprinting can provide accurate localization with low system requirements, it still suffers from multipath and fading effects. To resolve this, we propose a beam domain-based fingerprinting localization that can leverage the spatial feature with multiple antenna systems to improve the localization. Specifically, we consider using the beam domain receive power map (BDRPM), which is an RSSI-based map that captures important features of spatial fingerprints of the environment, for localization. To learn the environmental fingerprints via using BDRPMs and to conduct the localization, we propose a deep-learning approach based on the 2D convolutional neural network and auto-encoder structure. We conduct practical simulations to evaluate our proposed localization approach. The results show that our approach can provide very accurate localization, be resistant to environmental changes, and outperform the reference schemes in the literature.

Although the frequency-modulated continuous-wave (FMCW)-based scheme has been considered a candidate for realizing integrated radar-communication (RadCom) systems, it is limited by low spectral efficiency. Thus, this study proposes a frequency hopping-aided FMCW-based RadCom system along with the corresponding radar detection and communication signal demodulation approaches. The proposed system can significantly improve the transmission rate with slight radar performance degradation. In addition, both the radar and communication subsystems of the proposed RadCom system are analyzed. Consequently, a symbol interleaving approach that can mitigate radar performance degradation is proposed. Furthermore, the interference is shown to not degrade the radar performance significantly because of frequency hopping. However, to address the issue of the interference significantly degrading communication performance, interference avoidance and mitigation approaches using scheduling and interference rejection combining, respectively, are proposed. Simulations are conducted to evaluate the proposed RadCom system. The results validate the analysis and demonstrate that the proposed RadCom system can provide radar detection performance that is almost identical to conventional FMCW radar with a high transmission rate. Furthermore, the results confirm the effectiveness of the proposed interleaving and interference management methods.

Throughput-Outage scaling laws for single-hop cache-aided device-to-device (D2D) communications have been extensively investigated under the assumption of the protocol model. However, the corresponding performance under physical models has not been explored; in particular it remains unclear whether link-level power control and scheduling can improve the asymptotic performance. This paper thus investigates the throughput-outage scaling laws of cache-aided single-hop D2D networks considering a general physical channel model. By considering the networks with and without the equal-throughput assumption, we analyze the corresponding outer bounds and provide the achievable performance analysis. Results show that when the equal-throughput assumption is considered, using link-level power control and scheduling cannot improve the scaling laws. On the other hand, when the equal-throughput assumption is not considered, we show that the proposed double time-slot framework with appropriate link-level power control and scheduling can significantly improve the throughput-outage scaling laws, where the fundamental concept is to first distinguish links according to their communication distances, and then enhance the throughput for links with small communication distances.

This work examines a novel wireless content distribution problem where machine learning models (e.g., deep neural networks) are cached at local small cell base-stations to facilitate access by users within their coverage. The models are trained by federated learning procedures which allow local users to collaboratively train the models using their locally stored data. Upon the completion of training, the model can also be accessed by all other users depending on their application demand. Different from conventional wireless caching problems, the placement of machine learning models should depend not only on the users' preferences but also on the data available at the users and their channel conditions. In this work, we propose to jointly optimize the caching decision, user selection, and wireless resource allocation, including transmit powers and bandwidth of the selected users, to minimize a training error bound. The problem is reduced to minimizing a weighted sum of local dataset sizes subject to constraints on the cache storage capacity, the communication and computation latency, and the total energy consumption. We first derive the minimum loss achievable for each cached model, and, then, determine the optimal models to cache by solving an equivalent 0–1 Knapsack problem that minimizes the total average loss. Simulations show that the proposed scheme can achieve lower extremity error bounds compared to preference-only and random caching policies.

Having self-organizing ability is regarded as one of the vital features for modern wireless communication networks. Such self-organizing networks (SONs) thus draw significant attention in past years. As fault diagnosis is one of the essential functionalities for SONs, in this paper, we investigate the multi-fault and fault severity level diagnosis. Specifically, we propose deep learning-based approaches that can determine the faults and their corresponding levels by utilizing the network key performance indicators (KPIs). Furthermore, to enhance recall, we propose a loss function design that can effectively trade false alarm rate against recall. We conduct simulations adopting a practical setup to evaluate the performance. Results show that our proposed approaches can accurately diagnose multiple faults and determine their severity levels.

As user preferences can be influenced by the recommendation system, it has been shown that the joint recommendation and caching policy design can significantly improve the caching networks where caching is at the BSs. However, whether and how the joint recommendation and caching policy design can provide benefits to cache-aided device-to-device (D2D) networks have not been well-understood. This paper thus contributes in this direction by modeling the offloading probability of the cache-aided D2D network and proposing a social-aware joint recommendation caching policy design. Specifically, considering the preferences and social relationship of users as well as the caching and recommendation policies of the network, we formulate an offloading probability optimization problem which is non-convex. Then, an iterative algorithm with monotonicity and convergence property is proposed to solve the problem. By simulations, we show that the proposed joint recommendation and caching policy design can significantly outperform designs that only optimize the caching policy and other reference designs.

Throughput-Outage scaling laws for single-hop cache-aided device-to-device (D2D) communications have been extensively investigated under the assumption of the protocol model. However, the corresponding performance under physical models has not been explored; in particular it remains unclear whether link-level power control and scheduling can improve the asymptotic performance. This paper thus investigates the asymptotic throughput-outage tradeoff and derives its outer bound for cache-aided D2D networks under two common physical models. The results show that the asymptotic performance of the network under physical models is identical to that under the protocol model when requests are served with equal quality. This indicates that the throughput-outage performance cannot be improved asymptotically by using link-level power control and scheduling.

In this paper, we consider the automotive orthogonal frequency division modulation-radar in millimeter wave band. To avoid interference between different radar systems, resources need to be split and then used by different radar systems. This thus degrades the radar performance as compared to the radar system having full resources (FRs). To mitigate this issue, we develop a deep learning-based range-Doppler (R-D) map reconstruction approach along with a time-frequency resource allocation scheme. In the reconstruction approach, we propose a deep learning-based convolutional neural network to reconstruct the R-D map such that the reconstructed R-D map can be close to the R-D map under FRs. In the resource allocation scheme, we propose a block-wise interleaved method that can facilitate the proposed reconstruction approach. Simulation results show that our proposed approach can effectively mitigate the performance degradation of radar systems when resources are shared among users.

To realize the intelligent transportation, environmental awareness of roadside units (RSUs) is of paramount importance. One of the approaches to enable the environmental awareness of RSUs is to equip RSUs with radar systems. However, as more and more radar systems are installed, manually monitoring whether these radar systems work in their normal states becomes impossible. To resolve this issue, a radar system state self-diagnosis method is proposed in this paper by using the radar sensing information with deep learning techniques. Specifically, by using the proposed feature extraction approach, we first effectively convert the huge amount of radar sensing data into useful features. Then, by using the proposed deep neural network to interpret the extracted features, the radar systems can self-diagnose whether there exist faults on the systems. We verify our proposed method via real-world experiments. Results show that our proposed method can accurately diagnose the radar system and report the faults.

Cache-aided wireless device-to-device (D2D) networks have demonstrated more promising performance improvement for video distribution than conventional distribution methods; thus, understanding the fundamental scaling behavior of such networks is highly important. However, the existing scaling laws for multi-hop networks are not optimal even in the case of Zipf popularity distributions (gaps between upper and lower bounds are not constants); furthermore, there are no scaling law results for such networks for the more practical case of a Mandelbrot-Zipf (MZipf) popularity distribution. We thus in this work investigate the throughput-outage performance for cache-aided wireless D2D networks adopting multi-hop communications, with the MZipf popularity distribution for file requests and users distributed according to Poisson point process. We propose an achievable content caching and delivery scheme, and then analyze its performance. We obtain the optimal scaling law by showing that the achievable performance is tight to the proposed outer bound. Since the Zipf distribution is a special case of the MZipf distribution, the optimal scaling law for the networks considering the Zipf popularity distribution is also obtained, which closes the gap in literature.

Cache-aided wireless device-to-device (D2D) networks have demonstrated promising performance improvement for video distribution compared to conventional distribution methods. Understanding the fundamental scaling behavior of such networks is thus importance. Recently, based on real-world data, it has been observed that the popularity distribution should be modeled by a Mandelbrot-Zipf (MZipf) distribution, instead of the common Zipf distribution. We thus in this work investigate the throughput-outage performance for cache-aided wireless D2D network adopting multi-hop communications, with the MZipf popularity distribution for file requests and Poisson point process for user distribution. Considering the case that Zipf factor is larger than one, we first propose an achievable content caching and delivery scheme and analyze its performance. Then, by showing that the achievable performance is tight to the proposed outer bound, we show that an optimal scaling law for cache-aided wireless multi-hop D2D networks is obtained.

Recent investigations showed that cache-aided device-to-device (D2D) networks can be improved by properly exploiting the individual preferences of users. Since in practice it might be difficult to make centralized decisions about the caching distributions, this paper investigates the individual preference aware caching policy that can be implemented distributedly by users without coordination. The proposed policy is based on categorizing different users into different reference groups associated with different caching policies according to their preferences. To construct reference groups, learning-based approaches are used. To design caching policies that maximize throughput and hit-rate, optimization problems are formulated and solved. Numerical results based on measured individual preferences show that our design is effective and exploiting individual preferences is beneficial.

Designing beamforming techniques for the downlink (DL) of frequency division duplex (FDD) massive MIMO is known to be a challenging problem due to the difficulty of obtaining channel state information (CSI). Indeed, since the uplink-downlink bands are disjoint, the system cannot rely on channel reciprocity to estimate the channel from uplink (UL) pilots as in time division duplexing (TDD) system. Still, in this paper, we propose original designs for robust beamformers that do not require any feedback from the users and only rely on the transmission of UL pilots. The price to pay is that the beamformer is non-coherent in the sense that it does not leverage full knowledge of the phase of each multipath component. A large variety of novel designs are proposed under different criterion and partial phase knowledge.

Cache-aided wireless device-to-device (D2D) networks allow significant throughput increase, depending on the concentration of the popularity distribution of files. Many studies assume that all users have the same preference distribution; however, this may not be true in practice. This work investigates whether and how the information about individual preferences can benefit cache-aided D2D networks. We examine a clustered network and derive a network utility that considers both the user distribution and channel fading effects into the analysis. We also formulate a utility maximization problem for designing caching policies. This maximization problem can be applied to optimize several important quantities, including throughput, energy efficiency (EE), cost, and hit-rate, and to solve different tradeoff problems. We provide a general approach that can solve the proposed problem under the assumption that users coordinate, then prove that the proposed approach can obtain the stationary point under a mild assumption. Using simulations of practical setups, we show that performance can improve significantly with proper exploitation of individual preferences. We also show that different types of tradeoffs exist between different performance metrics and that they can be managed through caching policy and cooperation distance designs.

On-demand video accounts for the majority of wireless data traffic. Video distribution schemes based on caching combined with device-to-device (D2D) communications promise order-of-magnitude greater spectral efficiency for video delivery, but hinge on the principle of concentrated demand distributions. This paper presents, for the first time, the analysis and evaluations of the throughput-outage tradeoff of such schemes based on measured cellular demand distributions. In particular, we use a dataset with more than 100 million requests from the BBC iPlayer, a popular video streaming service in the U.K., as the foundation of the analysis and evaluations. We present an achievable scaling law based on the practical popularity distribution, and show that such scaling law is identical to those reported in the literature. We find that also for the numerical evaluations based on a realistic setup, order-of-magnitude improvements can be achieved. Our results indicate that the benefits promised by the caching-based D2D in the literature could be retained for cellular networks in practice.

Due to the concentrated popularity distribution of video files, caching of popular files on devices, and distributing them via device-to-device (D2D) communications allows a dramatic increase in the throughput of wireless video networks. However, since the popularity distribution is not static and the caching policy might be outdated, there is a need for replacement of cache content. In this work, by exploiting the broadcasting of the base station (BS), we model the caching content replacement in BS assisted wireless D2D caching networks and propose a practically realizable replacement procedure. Subsequently, by introducing a queuing system, the replacement problem is formulated as a sequential decision making problem, in which the long term average service rate is optimized under average cost constraint and queue stability. We propose a replacement design using Lyapunov optimization, which effectively solves the problem and makes decisions. Using simulations, we evaluate the proposed design. The results clearly indicate that, when dynamics exist, the systems exploiting replacement can significantly outperform the systems using merely the static policy.

Caching of video files at the wireless edge, i.e., at the base stations or on user devices, is a key method for improving wireless video delivery. While global popularity distributions of video content have been investigated in the past and used in a variety of caching algorithms, this paper investigates the statistical modeling of the individual user preferences. With individual preferences being represented by probabilities, we identify their critical features and parameters and propose a novel modeling framework by using a genre-based hierarchical structure as well as a parameterization of the framework based on an extensive real-world data set. Besides, the correlation analysis between parameters and critical statistics of the framework is conducted. With the framework, an implementation recipe for generating practical individual preference probabilities is proposed. By comparing with the underlying real data, we show that the proposed models and generation approach can effectively characterize the individual preferences of users for video content.

This work investigates the caching policy and cooperative distance designs for throughput and energy efficiency (EE) in base station (BS) assisted device-to-device (D2D) caching networks. To conduct the investigation in joint consideration of BS-, D2D-, and self-caching and the impact of cooperation distance, we configure a clustering network with specifically designed power control and resource reuse policies. After analyzing the throughput and EE of the network, their design problems and solving approaches are provided. By simulations, we validate our analyses and evaluate the proposed designs. Moreover, we show that the throughput and EE designs can significantly conflict with each other, and a trade-off design that provides a compromise between them is thus necessary for improving the system.

This paper investigates the optimal caching policy and cooperation distance design from both throughput and energy efficiency (EE) perspectives in base station (BS)-assisted wireless device-to-device (D2D) caching networks. By jointly considering the effects of the BS transmission, D2D-caching, and self-caching, and the impact of the cooperation distance, a clustering approach is proposed with specifically designed power control and resource reuse policies. The throughput and EE of two network structures are comprehensively analyzed and designs aiming to optimize the throughout and EE, respectively, are proposed. We also characterize the trade-off between the throughput and EE and provide corresponding designs. Simulations considering practical parameters are conducted to verify the analyses and evaluate the proposed designs; they demonstrate superior performance compared with state-of-the art.

Video caching at the wireless edge is a promising approach to improve throughput and reliability of video delivery. While most existing literature considers homogeneous user preference modeling when designing caching policies, this paper investigates how to exploit statistical information on individual popularity preferences in the design of caching policies in base station assisted wireless device-to-device networks. Specifically, we formulate the caching policy design problem aiming to minimize average energy consumption and propose two approaches that solve the problem under different conditions: users can or cannot coordinatedly design their policies. By simulating systems using practical individual preference models, we show that the proposed designs are effective, and the performance gain brought by exploiting individual preferences can be clearly observed.

Caching of video files at the wireless edge, i.e., at the base stations or on user devices, is a key method for improving wireless video delivery. While global popularity distributions of video content have been investigated in the past, and used in a variety of caching algorithms, this paper investigates the statistical modeling of the individual user preferences. With individual preferences being represented by probabilities, we identify their critical features and parameters and propose a novel modeling framework as well as a parameterization of the framework based on an extensive real-world data set. Besides, an implementation recipe for generating practical individual preference probabilities is proposed. By comparing with the underlying real data, we show that the proposed models and generation approach can effectively characterize individual preferences of users for video content.

In this paper, we propose a novel transmission approach, namely, analog precoding-aided virtual space modulation (APAVSM), and its corresponding multimode hybrid precoder designs in the multiple-input multiple-output system. We consider the system equipped with a single radio frequency chain and a phase shift network. Our proposed designs can attain benefits of both beamforming and spatial modulation, and therefore enable the significant signal-to-noise power ratio enhancement with efficient degrees of freedom utilization and low cost. We elaborate APAVSM and formulate the multimode hybrid precoder design problem. To solve the problem, we impose restrictions on the number of candidate modes for selection, and propose the hybrid precoder design approach in which two non-convex max-min problems are solved sequentially by iterative approaches. To trade off between error rate and complexity, we propose alternative designs by reducing the dimensionality and/or relaxing constraints. Moreover, by the concept of beamspace, heuristic algorithms are proposed with extremely low complexities. We offer complexity analyses for proposed design approaches. Simulations in Ralyigh fading and millimeter-wave (mmWave) channels are used to evaluate the proposed designs. Results show that our designs can outperform the existing SM-MIMO systems, and provide the feasible operation in mmWave channels with severe path loss. Moreover, the superiority of proposed designs to the conventional precoding-aided MIMO systems is numerically demonstrated.

The distinct ability of spatial modulation (SM) to effectively utilizing spatial degrees of freedom with only a single radio frequency chain renders it promising for millimeter wave (mmWave) communications. In this work we investigate the transmitter design employing the SM concept in mmWave multiple-input multiple-output (MIMO) systems. To introduce beamforming gain while maintaining the advantages of SM, we apply the virtual antenna concept to SM and construct spatial signatures with the aid of analog beamforming. For a given configuration of the space-signal constellation, we formulate an optimization problem for designing the analog beamforming matrix and propose two approaches via exploiting characteristics of mmWave channel and transmitter architecture to acquire effective solutions with low complexity. To further improve the performance, we optimize the configuration of the spacesignal constellation by selecting the optimal sizes of spatial and signal constellations while guaranteeing the transmission rate requirement. The overhead and design complexity of proposed design approaches are analyzed. Besides, we exploit simulations to evaluate the proposed designs, and briefly compare between analog beamforming aided SM-MIMO and conventional MIMO to show advantages of our designs.

In this paper, we investigate the bit error rate (BER) for multicast multiple-input multiple-output (MIMO) systems, employing spatial modulation (SM) and its variants, called multicast SM-type MIMO systems, in Rayleigh fading channels. The system BER, here, is derived by first attaining the BER of the worst receiver of each channel realization set, and then averaging over all possible sets. We first consider the uncorrelated channels. By exploiting the system statistics, a tight BER upper bound is proposed and the diversity is discussed for the systems. We then perform the asymptotic analysis, and show that the BER of the multicast SM-type MIMO system can be alternatively analyzed by analyzing the simple point-to-point SM-type MIMO system with Weibull fading channels. Through this property, a closed-form asymptotic BER upper bound is provided and the impact of the receiver number on BER is analyzed. Subsequently, our investigation is extended to correlated channels where all the receivers share the same correlation statistics. The BER analysis is performed again through the framework similar to the uncorrelated case. In the analysis, the tight upper bound is derived and the effect of correlations is analyzed. Moreover, we provide an explicit expression for the SNR degradation caused by the receive correlation through the analysis. Finally, simulations are exploited to evaluate the BER of the multicast SM-type MIMO systems and validate the analyses.

In this paper, we investigate the configuration (i.e., both mode and antenna) selection and precoder design to improve the multicast multiple-input multiple-output systems employing spatial modulation. We first elaborate the advantages of configuration selection and then propose two selection schemes to maximize minimum Euclidean distance with low complexity. By configuring the selection problem as a tree search problem and adopting tree pruning technique, the first selection scheme attains low complexity while obtaining the optimal solution. In the second scheme, by pruning the less dominant nodes and approximating the signal constellation, the Euclidean distance computation is converted to a simple table look-up operation. Combining this with the first scheme, the second scheme attains even lower complexity without optimality guarantee. Besides configuration selection, the precoder design is investigated, and a precoder design approach is proposed to improve the system. Finally, simulation results demonstrate the efficacy of the proposed approaches in the bit error rate improvement and complexity reduction.

In this paper, we propose two generalized precoder designs to enhance the bit error rates for the general category of spatial modulation (SM) in multiple-input multiple-output (MIMO) systems with channel state information at the transmitter (CSIT). We investigate typical SM-MIMO systems and propose two optimization formulations for designing precoders. Our design rationale for the first formulation is to maximize the minimum Euclidean distance among codewords; for the second formulation, it is to minimize the total signal power for the lower-bounded Euclidean distances among codewords. Since both formulations are non-convex and their optimal solutions are generally intractable, we propose an algorithm that acquires effective solutions by iteratively solving the alternative convex problem linearized and approximated from the original non-convex problem. Discussions on complexity analysis, performance comparisons, design challenges, and robustness in imperfect CSIT are then provided. By generalizing the design formulations, the proposed precoder designs can be extended to generalized SM, which completes the investigation for virtually all SM-type systems. Simulation results show that the proposed designs improve the performance of SM-/GSM-MIMO systems and outperform existing precoding methods with a potentially higher complexity cost.

In multiple-input multiple-output (MIMO) systems adopting space shift keying (SSK), the use of adaptive precoder on transmitter offers the opportunity to improve its performance significantly. One major challenge in precoder operation is the difficulty in obtaining the full channel state information on transmitter (CSIT). In this work, we investigate the precoder design for SSK-MIMO systems with limited feedback through using the codebook for the precoding. We formulate a distortion metric to evaluate the quality of a codebook based on the maximum minimum Euclidean distance criterion and propose a codebook design criterion accordingly. Two effective codebook design algorithms are proposed based on thorough analysis of the criterion. Simulation results show the performance improvement of the proposed codebook-based precoding, and also support our analyses on the codebook design.

This work investigates the configuration selection in IA-aided UL CoMP systems which pursues the maximal achievable sum rate. The configuration is defined by the number of data streams transmitted in each user. Intuitively, the solution can be found by exhaustively calculating the achievable sum rate for all the possible configurations, and select the configuration leading to the maximal achievable sum rate. The exhaustive search method is infeasible due to its high complexity. Based on the characteristics of IA-aided UL CoMP systems, an algorithm is proposed to obtain the configuration with extremely low complexity, and the proposed approach achieves comparable performance as in exhaustive method.

The authors investigate linear transceiver design in uplink (UL) coordinated multipoint transmission and reception (CoMP) multiple‐input multiple‐output (MIMO) systems with joint detections. A two‐stage design algorithm is proposed by exploiting the technique of interference alignment, optimising the structure of the effective channel and employing power loading, with the goal to achieve high throughput and convergence performance. In contrast to conventional CoMP transceiver design, which is investigated under a predefined number of data streams transmitted by each user, the authors further investigate the selection of the number of data streams, called the configuration selection, and propose a corresponding low‐complexity algorithm. By combining the proposed linear transceiver algorithm and low‐complexity configuration selection algorithm, this work presents a new practical framework for linear transceiver design in UL CoMP MIMO systems. The simulation results confirm that the proposed algorithms achieve higher sum‐rate performance than prior linear transceivers used in UL CoMP MIMO systems. Furthermore, the proposed transceiver offers comparable performance to existing IA‐aided transceivers with significantly faster convergence.