Unmanned aerial vehicles (UAVs) will be an integral part of the next generation wireless communication networks. Their adoption in various communication-based applications is expected to improve coverage and spectral efficiency, as compared to traditional ground-based solutions. However, this new degree of freedom that will be included in the network will also add new challenges. In this context, the machine-learning (ML) framework is expected to provide solutions for the various problems that have already been identified when UAVs are used for communication purposes. In this article, we provide a detailed survey of all relevant research works, in which ML techniques have been used on UAV-based communications for improving various design and functional aspects such as channel modeling, resource management, positioning, and security.

No abstract

No abstract

Non-orthogonal multiple access (NOMA) aims to increase the spectral efficiency of fifth generation (5G) networks by relaxing the orthogonal use of radio-resources. In this work, a network with multiple half-duplex (HD) buffer-aided (BA) relays is considered, where the source transmits with fixed rate towards two users. The users might demand the same rate by the source (e.g., two cellular users requiring the same service), or they could have different rate requirements (e.g., a cellular user coexisting with an Internet of Things (IoT) device). By deploying multiple BA relays, increased reliability and additional degrees of freedom are provided. Leveraging the spectral efficiency of NOMA and the increased diversity gain of BA relaying, two relay selection algorithms with broadcasting are proposed for power-domain (PD) NOMA and hybrid NOMA/OMA, namely BA-NOMA and BA-NOMA/OMA, respectively. BA-NOMA can improve the performance in terms of outage probability when the power allocation factor α is selected such that robustness against channel uncertainties due to, e.g., outdated channel state information (CSI), is provided. Moreover, BA-NOMA/OMA further improves the sum-rate by switching to OMA when the relays cannot serve the users through NOMA. For both cases, a theoretical analysis for the outage probability is conducted and the asymptotic performance is studied. Finally, numerical results and comparisons with other state-of-the-art algorithms are provided for the outage probability, average throughput and average delay.

Buffer-Aided (BA) relaying has shown tremendous performance improvements in terms of throughput and outage probability, although it has been criticized of suffering from long delays that are restrictive for applications, such as video streaming, Web browsing, and file sharing. In this paper, we propose novel relay selection policies aiming at reducing the average delay by incorporating the buffer size of the relays into the decision making of the relay selection process. More specifically, we first propose two new delay-aware policies. One is based on the hybrid relay selection algorithm, where the relay selection takes into account the queue sizes so that the delay is reduced and the diversity is maintained. The other approach is based on the max − link relay selection algorithm. For the max − link algorithm, a delay-aware only approach starves the buffers and increases the outage probability of the system. Thus, for max − link, we propose a delay-and diversity-aware BA relay selection policy targeting the reduction of the average delay, while maintaining the diversity of the transmission. The proposed policies are analyzed by means of Markov Chains and expressions for the outage, throughput, and delay are derived. The asymptotic performance of the policies is also discussed. The improved performance in terms of delays and the use of the proposed algorithms are demonstrated via extensive simulations and comparisons, signifying, at the same time, the need for adaptive mechanisms to handle the interplay between delay and diversity.