In this article, we study the optimization problem in low Earth orbit (LEO) satellite network scenarios. We present a 3D model of the LEO satellite utilizing a spherical coordinate system, allowing us to determine a practical elevation angle for the ground-to-satellite channel model. Our objective is to maximize the minimum average throughput by formulating an optimization problem that considers the association between the ground terminal (GT) and LEO satellite, as well as the transmit power of the GT. To address this problem, we propose an iterative algorithm that combines the block coordinate descent method with the successive convex approximation algorithm. Simulation results corroborate that our proposed optimal method achieves a substantial throughput gain of up to 59.63 % compared to conventional methods, while ensuring fairness. Additionally, we present two significant findings: 1) the efficiency of our method is demonstrated through Top-view 2D (and 3D) snapshots, showcasing the optimal decisions made, and 2) the impact of key parameters such as the number of resource blocks, interference, and network density on the network performance.

The burgeoning field of on-device AI communication, where devices exchange information directly through embedded foundation models, such as language models (LMs), requires robust, efficient, and generalizable communication frameworks. However, integrating these frameworks with existing wireless systems and effectively managing noise and bit errors pose significant challenges. In this work, we introduce a practical on-device AI communication framework, integrated with physical layer (PHY) communication functions, demonstrated through its performance on a link-level simulator. Our framework incorporates end-to-end training with channel noise to enhance resilience, incorporates vector quantized variational autoencoders (VQ-VAE) for efficient and robust communication, and utilizes pre-trained encoder-decoder transformers for improved generalization capabilities. Simulations, across various communication scenarios, reveal that our framework achieves a 50% reduction in transmission size while demonstrating substantial generalization ability and noise robustness under standardized 3GPP channel models.

Large-scale channel prediction, i.e., estimation of the pathloss from geographical/morphological/building maps, is an essential component of wireless network planning. Ray tracing (RT)-based methods have been widely used for many years, but they require significant computational effort that may become prohibitive with the increased network densification and/or use of higher frequencies in B5G/6G systems. In this paper, we propose a data-driven, model-free pathloss map prediction (PMP) method, called PMNet. PMNet uses a supervised learning approach: it is trained on a limited amount of RT data and map data. Once trained, PMNet can predict pathloss over location with high accuracy (an RMSE level of$10^{-2}$) in a few milliseconds. We further extend PMNet by employing transfer learning (TL). TL allows PMNet to learn a new network scenario quickly ($\times 5.6$faster training) and efficiently (using$\times 4.5$less data) by transferring knowledge from a pre-trained model, while retaining accuracy. Our results demonstrate that PMNet is a scalable and generalizable ML-based PMP method, showing its potential to be used in several network optimization applications.

This study presents a novel deep reinforcement learning (DRL)-based handover (HO) protocol, called DHO, specifically designed to address the persistent challenge of long propagation delays in low-Earth orbit (LEO) satellite networks’ HO procedures. DHO skips the Measurement Report (MR) in the HO procedure by leveraging its predictive capabilities after being trained with a pre-determined LEO satellite orbital pattern. This simplification eliminates the propagation delay incurred during the MR phase, while still providing effective HO decisions. The proposed DHO outperforms the legacy HO protocol across diverse network conditions in terms of access delay, collision rate, and handover success rate, demonstrating the practical applicability of DHO in real-world networks. Furthermore, the study examines the trade-off between access delay and collision rate and also evaluates the training performance and convergence of DHO using various DRL algorithms.

Pathloss prediction is an essential component of wireless network planning. While ray tracing based methods have been successfully used for many years, they require significant computational effort that may become prohibitive with the increased network densification and/or use of higher frequencies in 5G/BSG (beyond 5G) systems. In this paper, we propose and evaluate a data-driven and model-free pathloss prediction method, dubbed PMNet. This method uses a supervised learning approach: training a neural network (NN) with a limited amount of ray tracing (or channel measurement) data and map data and then predicting the pathloss over location with no ray tracing data with a high level of accuracy. Our proposed pathloss map prediction-oriented NN architecture, which is empowered by state-of-the-art computer vision techniques, outperforms other architectures that have been previously proposed (e.g., UNet, RadioUNet) in terms of accuracy while showing generalization capability. Moreover, PMNet trained on a 4-fold smaller dataset surpasses the other baselines (trained on a 4-fold larger dataset), corroborating the potential of PMNet.11The trained model and codes are publicly available on the Github page: https://github.com/abman23/PMNet

Indoor localization is a challenging task. Compared to outdoor environments where GPS is dominant, there is no robust and almost-universal approach. Recently, machine learning (ML) has emerged as the most promising approach for achieving accurate indoor localization. Nevertheless, its main challenge is requiring large datasets to train the neural networks. The data collection procedure is costly and laborious, requiring extensive measurements and labeling processes for different indoor environments. The situation can be improved by Data Augmentation (DA), a general framework to enlarge the datasets for ML, making ML systems more robust and increasing their generalization capabilities. This paper proposes two simple yet surprisingly effective DA algorithms for channel state information (CSI) based indoor localization motivated by physical considerations. We show that the number of measurements for a given accuracy requirement may be decreased by an order of magnitude. Specifically, we demonstrate the algorithms' effectiveness by experiments conducted with a measured indoor WiFi measurement dataset: As little as 10% of the original dataset size is enough to get the same performance as the original dataset. We also showed that if we further augment the dataset with the proposed techniques, test accuracy is improved more than three-fold.

This paper describes our pathloss prediction system submitted to the ICASSP 2023 First Pathloss Radio Map Prediction Challenge. We describe the architecture of PMNet, a neural network we specifically designed for pathloss prediction. Moreover, to enhance the prediction performance, we apply several machine learning techniques, including data augmentation, fine-tuning, and optimization of the network architecture. Our system achieves an RMSE of 0.02569 on the provided RadioMap3Dseer dataset, and 0.0383 on the challenge test set, placing it in the 1st rank of the challenge.

A mega-constellation of low-altitude earth orbit (LEO) satellites (SATs) are envisaged to provide a global coverage SAT network in beyond fifth-generation (5G) cellular systems. LEO SAT networks exhibit extremely long link distances of many users under time-varying SAT network topology. This makes existing multiple access protocols, such as random access channel (RACH) based cellular protocol designed for fixed terrestrial network topology, ill-suited. To overcome this issue, in this paper, we propose a novel contention-based random access solution for LEO SAT networks, dubbed emergent random access channel protocol (eRACH). In stark contrast to existing model-based and standardized protocols, eRACH is a model-free approach that emerges through interaction with the non-stationary network environment, using multi-agent deep reinforcement learning (MADRL). Furthermore, by exploiting known SAT orbiting patterns, eRACH does not require central coordination or additional communication across users, while training convergence is stabilized through the regular orbiting patterns. Compared to RACH, we show from various simulations that our proposed eRACH yields 54.6% higher average network throughput with around two times lower average access delay while achieving 0.989 Jain’s fairness index.

A mega-constellation of low-altitude earth orbit (LEO) satellites (SATs) are envisaged to provide a global coverage SAT network in beyond fifth-generation (5G) cellular systems. However, such wide coverage rather makes it difficult to apply existing multiple access protocols, such as random access channel (RACH). To overcome this issue, in this paper, we propose a novel random access solution for LEO SAT networks, called as S-RACH. In contrast to existing standardized protocols, S-RACH is a model-free approach using deep reinforcement learning (DRL). Compared to RACH, we show from various simulations that our proposed S-RACH yields around 2x lower average access delay.

Integrating terrestrial networks with burgeoning high-altitude platform stations (HAPSs) will be a disruptive challenge for beyond-5G systems provisioning large-scale three-dimensional connectivity. Here, we study the problem of forwarding packets between terrestrial terminals and backhaul through multi-HAPS relaying. Considering the limited wireless backhaul, which is the practical constraint of HAPS relaying, dual-hop mixed radiofrequency/free-space optical (RF/FSO) networks are investigated, where backhaul-to-relay and relay-to-user communications employ FSO and RF links, respectively. To maximize the end-to-end network throughput, including downlink and uplink rate, we formulate the optimization problem for variables; the association between aerial and terrestrial terminals, transmit power, and deployment of multiple HAPSs, respectively. We tackle this problem using an iterative algorithm with proposed surrogate functions to efficiently obtain the locally optimal solution. Simulation results corroborate that our proposed optimal method achieves up to 11.3% spectral efficiency compared to the conventional heuristic method. Furthermore, we answer three questions; 1) what the wireless backhaul requirements are, 2) how the number of HAPSs and terrestrial terminals impact the network, and 3) what if a certain user terminal has a particular demand on data rate.

In this paper, we investigate a mobile relaying system assisted by an unmanned aerial vehicle (UAV) with a finite size of the buffer. Under the buffer size limit and delay constraints at the UAV relay, we consider a dual-hop mixed free-space optical/radio frequency (FSO/RF) relaying system (i.e., the source-to-relay and relay-to-destination links employ FSO and RF links, respectively). Taking an imbalance in the transmission rate between RF and FSO links into consideration, we address the trajectory design of the UAV relay node to obtain the maximum data throughput at the ground user terminal. Specifically, we classify two relaying transmission schemes according to the delay requirements, i.e., i) delay-limited transmission and ii) delay-tolerant transmission. Accordingly, we propose an iterative algorithm to effectively obtain the locally optimal solution to our throughput optimization problems and further present the complexity analysis of this algorithm. Through this algorithm, we present the resulting trajectories over the atmospheric condition, the buffer size, and the delay requirement. In addition, we show the optimum buffer size and the throughput-delay tradeoff for a given system. The numerical results validate that the proposed buffer-aided and delay-considered mobile relaying scheme obtains 223.33% throughput gain compared to the conventional static relaying scheme.

A mega-constellation of low-earth orbit (LEO) satellites has the potential to enable long-range communication with low latency. Integrating this with burgeoning unmanned aerial vehicle (UAV) assisted non-terrestrial networks will be a disruptive solution for beyond 5G systems provisioning large-scale three-dimensional connectivity. In this article, we study the problem of forwarding packets between two faraway ground terminals, through an LEO satellite selected from an orbiting constellation and a mobile high-altitude platform (HAP) such as a fixed-wing UAV. To maximize the end-to-end data rate, the satellite association and HAP location should be optimized, which is challenging due to a huge number of orbiting satellites and the resulting time-varying network topology. We tackle this problem using deep reinforcement learning (DRL) with a novel action dimension reduction technique. Simulation results corroborate that our proposed method achieves up to 5.74x higher average data rate compared to a direct communication baseline without SAT and HAP.

In this work, we address the optimization of vertical backhaul framework where multiple aerial base station (ABS) are deployed to conFigure the intermediate backhaul links for terrestrial base stations (TBS) in wireless networks. Here, we focus on maximizing the downlink network throughput in mixed RF/FSO backhaul networks by optimizing resource allocation and placement. Specifically, the ABS-TBS association, transmit power, and positions of ABSs are alternately and iteratively optimized. To solve the corresponding mixed-integer non-convex optimization problem, we propose an efficient iterative algorithm by applying the block coordinate descent method and successive convex optimization techniques. Although we obtain suboptimal solutions due to the non-convexity of the problems, simulation results indicate that the proposed scheme demonstrates the significant network throughput gain compared with the conventional scheme.

In this work, we address the trajectory optimization of a fixed-wing unmanned aerial vehicle (UAV) using free space optical communication (FSOC). Here, we focus on maximizing the flight time of the UAV by considering practical constraints for wireless UAV communication, including limited propulsion energy and required data rates. We find optimized trajectories in various atmospheric environments (e.g., moderate-fog and heavy-fog conditions), while also considering the channel characteristics of FSOC. In addition to maximizing the flight time, we consider the energy efficiency maximization and operation-time minimization problem to find the suboptimal solutions required to meet those constraints. Furthermore, we introduce a low-complexity approach to the proposed framework. In order to address the optimization problem, we conduct a bisection method and sequential programming and introduce a new feasibility check algorithm. Although our design considers suboptimal solutions owing to the nonconvexity of the problems, our simulations indicate that the proposed scheme exhibits a gain of approximately 44.12% in terms of service time when compared to the conventional scheme.

In this paper, we consider an unmanned aerial vehicle (UAV) aided mobile relaying system under a buffer constraint at the relay node. We propose a new relaying protocol employing mixed free-space optical/radio frequency (FSO/RF) communication, i.e., the source-relay and relay-destination link utilize FSO communication and RF communication, respectively, under the buffer constraint which is required to consider practical relay system. We study the trajectory optimization problem of buffer-constrained UAV-relaying in order to maximize the end-to-end data throughput. Taking the conditions of the mixed FSO/RF systems (e.g., a full-duplex relaying network that works in decode-and-forward, atmospheric attenuation, transmit power, and bandwidth in FSO and RF links) into consideration, we characterize the channel and throughput models. Furthermore, corresponding to the buffer-aided relaying, we derive a limited buffer constraint regarding the state of the queue in the buffer of relay. We solve the optimal trajectory problem of the UAV to maximize the throughput of user terminal using quadratically constrained programming. As a result, we propose an iterative algorithm that efficiently finds a local optimum solution for the throughput maximization problems. Our numerical results show that proposed buffer-aided mobile relaying achieves 161.3% throughput gains compared to a static relaying scheme.