In the ever-evolving field of technologies, the emergence of Artificial General Intelligence (AGI), often referred as strong artificial intelligence (AI), stands as a breakthrough in the realm of machines intelligence, promising to witness a new era of capabilities and possibilities. In particular, AGI ventures into human-level cognition, and expands to thinking, reasoning, and awareness. This imminent evolution is envisioned to be manifested through the embodiment of AI machines, allowing machines to transcend their purely computational nature and interact with the world through the different senses. Accordingly, AI agents will be grounded in the physical environment, going through subjective experiences and acquiring the needed knowledge that will lead to understanding and cognition. In our article, we explore the path towards realizing the true vision of AGI through AI embodiment, where we dig into the different types of thinking required to achieve knowledge, and hence, cognition and understanding. Furthermore, we look through the evolution of generative AI models, and shed lights on the limitations of auto-regression in large language models (LLMs), with the aim to answer the question: is sensory grounding (through 6G) necessary, and enough, to achieve understanding in LLMs? Finally, we identify the main pillars of AGI and unveil how 6G networks will orchestrate the development of AGI systems.

Ensuring the security and reliability of cooperative vehicle-to-vehicle (V2V) communications is an extremely challenging task, due to the dynamic nature of vehicular networks as well as the delay-sensitive wireless medium. The moving target defense (MTD) paradigm has been proposed to overcome the challenges of conventional solutions, based on static network services and configurations. Specifically, the MTD approach involves the dynamic altering of the network configurations to improve resilience to cyberattacks. Nevertheless, the current MTD solution for cooperative networks has several limitations, such as they are not well-suited for highly dynamic environments; they require high synchronization modules that are resource-intensive and difficult to implement; and finally, they rely heavily on the attack-defense models, which may not always be accurate or comprehensive to use. In this paper, we propose an intelligent spatiotemporal diversification MTD scheme to defend against eavesdropping attacks in cooperative V2V networks. Specifically, we design benign random data injection patterns to meet the security and reliability requirements of the vehicular network. Our methodology involves modeling the configuration of vehicular relays and data injection patterns as a Markov decision process, followed by applying deep reinforcement learning to determine the optimal configuration. We then iteratively evaluate the intercept probability and the percentage of transmitted real data for each configuration until convergence is achieved. In order to optimize the security-real data percentage (S-RDP), we developed a two-agent framework, namely MTD-DQN-RSS & MTD-DQN-RSS-RDP. The first agent, MTD-DQN-RSS, tries to minimize the intercept probability by injecting additional fake data, which in turn reduces the overall RDP, while the second agent, MTD-DQN-RSS-RDP, attempts to inject a sufficient amount of fake data to achieve a target S-RDP. Finally, extensive simulation results are conducted to demonstrate the effectiveness of our proposed solution where they improved the system security by almost 28% and 49%, respectively compared to the conventional relay selection approach.

Visible light communication is envisaged as a promising enabling technology for sixth generation (6G) and beyond networks. It was introduced as a key enabler for reliable massive-scale connectivity, mainly thanks to its simple and low-cost implementation which require minor variations to the existing indoor lighting systems. The key features of VLC allow offloading data traffic from the current congested radio frequency (RF) spectrum in order to achieve effective short-range, high speed, and green communications. However, several challenges prevent the realization of the full potentials of VLC, namely the limited modulation bandwidth of light emitting diodes, the interference resulted from ambient light, the effects of optical diffuse reflection, the non-linearity of devices, and the random receiver orientation. Meanwhile, centralized machine learning (ML) techniques have exhibited great potentials in handling different challenges in communication systems. Specifically, it has been recently shown that ML algorithms exhibit superior capabilities in handling complicated network tasks, such as channel equalization, estimation and modeling, resources allocation, opportunistic spectrum access control, non-linearity compensation, performance monitoring, detection, decoding/encoding, and network optimization. Nevertheless, concerns relating to privacy and communication overhead when sharing raw data of the involved clients with a server constitute major bottlenecks in large-scale implementation of centralized ML techniques. This has motivated the emergence of a new distributed ML paradigm, namely federated learning (FL). This method can reduce the cost associated with transferring the raw data, and preserve clients privacy by training ML model locally and collaboratively at the clients side. Thus, the integration of FL in VLC networks can provide ubiquitous and reliable implementation of VLC systems. Based on this, for the first time in the open literature, we provide an overview about VLC technology and FL. Then, we introduce FL and its integration in VLC networks and provide an overview on the main design aspects. Finally, we highlight some interesting future research directions of FL that are envisioned to boost the performance of VLC systems.

With the advancements of high-resolution cameras, highly sensitive photodetectors, and energy-efficient light emitting-diodes, optical wireless communication (OWC) has emerged as key-enabling technology for 6G and beyond. Similarly, digital twin (DT) technology has been recently proposed to meet the diverse requirements of emerging applications and provide efficient optimization of the overall system resources by enabling selfsustaining and proactive online learning. In DT, various technology disciplines, including big data, artificial intelligence, intelligent computing, communication, and security-aware technologies, are integrated to create a virtual replica that reflects the physical system. Motivated by this, the presented article discusses DT and its role in the reliable and ubiquitous implementation of OWC networks. We first provide an overview of this emerging research area by shedding light on the key enabling technologies and potential applications in the context of smart-autonomous OWC networks. Subsequently, we provide recent advancements and design aspects related to DT-assisted OWC systems. Finally, future research directions and the impact of different system factors on the overall network performance are discussed. To the best of the authors’ knowledge, this is the first in-depth review in the literature on the integration of DT with OWC networks.

Fifth-generation and beyond (5G and xG) wireless networks are envisioned to meet the requirements of vertical applications like high traffic throughput, ultra-massive connectivity, extremely low latency, and high quality of service. Disruptive technologies, such as massive multiple-input multiple-output, millimeter wave, and multiple access are being deployed to meet these requirements. However, their deployment poses several challenges, including a lack of network transparency, management decentralization, and reliability. Moreover, the heterogeneity of future networks raises security concerns, e.g., confidentiality, privacy, and trustworthiness. Indeed, due to the emergence of novel paradigms, e.g., quantum computing, traditional security approaches are no longer sufficient to protect over-the-air communications. Hence, 5G/xG networks must consider smart security techniques to operate seamlessly and efficiently. Within this context, physical layer security (PLS) and blockchain represent promising solutions to complement existing methods. By exploiting the characteristics of wireless links, PLS can enhance the security of communications, while blockchain may enable networks’ decentralization, integrity, and trustworthiness. Motivated by these advancements, we provide an in-depth review of the existing PLS and blockchain literature. Then, for the first time in the literature, we present a framework to integrate PLS with blockchain in 5G/xG systems. We first provide a thorough discussion about the potential of PLS and blockchain for 5G/xG systems. Then, we present our vision of a cross-layer architecture that leverages PLS and blockchain. Through a case study, we demonstrate the high potential of cross-layer design to improve the security of vehicular networks. Finally, we identify related challenges and shed light on future research directions.

Spatial Modulation (SM) has been proposed as a multiple-input-multiple-output (MIMO)-based technique to overcome the inter-channel interference experienced in conventional MIMO systems. It has been further shown that SM enhances energy efficiency and reduces the receiver’s complexity. Nevertheless, under high antenna correlation scenarios, the detection performance of the antenna indices degrades significantly. To address this critical concern, in this paper, we propose three autoencoder-based frameworks for spatial modulation. The first scenario, similar to conventional spatial modulation, trains the encoder for data modulation and the decoder for data demodulation as well as antenna index detection. The performance of this framework deteriorates in high antenna correlation scenarios. Therefore, two novel solutions are presented to embed the antenna index into the transmitted signal in order to reduce the receiver’s reliance on the channel conditions. The first framework adds a  phase-shift keying-based antenna signature, while the other trains the encoder to learn an appropriate antenna index embedding. Simulation results show that the two enhanced frameworks result in a significantly enhanced performance, compared to conventional spatial modulation, in terms of block error rate and power efficiency under a high correlation setup (about 18 dB and 24 dB gain, respectively, at a Rician factor of 20 dB).

The evolution of generative artificial intelligence (GenAI) constitutes a turning point in reshaping the future of technology in different aspects. Wireless networks in particular, with the blooming of self-evolving networks, represent a rich field for exploiting GenAI and reaping several benefits that can fundamentally change the way how wireless networks are designed and operated nowadays. To be specific, large language models (LLMs), a subfield of GenAI, are envisioned to open up a new era of autonomous wireless networks, in which a multimodal large model trained over various Telecom data, can be fine-tuned to perform several downstream tasks, eliminating the need for dedicated AI models for each task and paving the way for the realization of artificial general intelligence (AGI)-empowered wireless networks. In this article, we aim to unfold the opportunities that can be reaped from integrating LLMs into the Telecom domain. In particular, we aim to put a forward-looking vision on a new realm of possibilities and applications of LLMs in future wireless networks, defining directions for designing, training, testing, and deploying Telecom LLMs, and reveal insights on the associated theoretical and practical challenges.

The recent progress of artificial intelligence (AI) opens up new frontiers in the possibility of automating many tasks involved in Telecom networks design, implementation, and deployment. This has been further pushed forward with the evolution of generative artificial intelligence (AI), including the emergence of large language models (LLMs), which is believed to be the cornerstone toward realizing self-governed, interactive AI agents. Motivated by this, in this paper, we aim to adapt the paradigm of LLMs to the Telecom domain. In particular, we fine-tune several LLMs including BERT, distilled BERT, RoBERTa and GPT-2, to the Telecom domain languages, and demonstrate a use case for identifying the \ac{3gpp} standard working groups. We consider training the selected models on 3GPP techincal documents (Tdoc) pertinent to years 2009-2019 and predict the Tdoc categories in years 2020-2023. The results demonstrate that fine-tuning BERT and RoBERTa model achieves 84.6% accuracy, while GPT-2 model achieves 83% in identifying 3GPP working groups. The distilled BERT model with around 50% less parameters achieves similar performance as others. This corroborates that fine-tuning pretrained LLM can effectively identify the categories of Telecom language. The developed framework shows a stepping stone towards realizing intent-driven and self-evolving wireless networks from Telecom languages, and paves the way for the implementation of generative AI in the Telecom domain.

The sixth generation (6G) of wireless networks are envisioned to support a plethora of human-centric applications and offer connectivity to a massive number of devices with diverse requirements, thus enabling massive Machine Type Communications. Nevertheless, with the rapid growth of the number of connected devices as well as the ever-increasing network traffic, network energy consumption has become a major challenge. Additionally, 6G is expected to catalyze the emergence of new applications that are characterized by their harsh environmental conditions, with ultra-small and low-cost wireless devices. Therefore, there is a pressing need for developing sustainable solutions that take into consideration all these requirements in order to realize the full potential of 6G networks. Within this context, zero-energy devices (ZEDs) have emerged as a prominent solution for the next generation green communication architecture. Such devices eliminate the need for recharging plugins and replacing batteries by integrating disruptive technologies, such as radio frequency energy harvesting, backscatter communications, low power computing, and ultra-low power receivers. Motivated by this, this article provides an in-depth review of the existing literature on the newly emerging ZEDs for future networks. We further identify different relevant use cases and provide an extensive overview on the key enabling technologies and their requirements for realizing ZED-empowered networks. Finally, we discuss potential future research directions and challenges that are envisioned to enhance the performance and efficiency of ZED-based networks.

Cooperative communications is a core research area in wireless vehicular networks (WVNs), thanks to its capability to mitigate fading and improve spectral efficiency. In a cooperative scenario, the performance of the system is improved by selecting the best relay for data transmission among a group of available relays. However, due to the mobility of WVNs, the best relay is often selected in practice based on outdated channel state information (CSI), which in turn affects the overall system performance. Therefore, there is a need for a robust relay selection scheme (RSS) that improves the overall achievable performance of an outdated CSI. Motivated by this and considering the advantageous features of autoregressive moving average (ARMA), in the present work we model a cooperative vehicular communication scenario with relay selection as a Markov decision process (MDP) and propose two deep Q-networks (DQNs), namely DQN-RSS and DQN-RSS-ARMA. In the proposed framework, two deep reinforcement learning (RL)-based RSS are trained based on the intercept probability, aiming to select the optimal vehicular relay from a set of multiple relays. We then compare the proposed RSS with the conventional methods and evaluate the performance of the network from the security point of view. Simulation results show that DQN-RSS and DQN-RSS-ARMA perform better than conventional approaches, and they reduce intercept probability by approximately 15\% and 30\%, respectively, compared to the ARMA approach.

Future wireless networks must incorporate awareness, adaptability, and intelligence as fundamental building elements in order to meet the wide range of requirements of the next-generation communication systems. Wireless sensing techniques can be used to gather awareness from the radio signals present in the surroundings. However, threats from hostile attackers, such as jamming, eavesdropping, and manipulation, are also present along with this. This paper describes in detail an RF-jamming detection testbed and provides experimentally measured data. The RF jamming detection testbed, in particular, makes use of the spectral scan capability of wireless network interfaces and JamRF, a jamming toolkit. In order to facilitate future progress in the experimentation of jamming detection and avoidance systems, we explain the methodology used for the development of our testbed and discuss the reasoning behind the choices made throughout its evolution. Furthermore, we provide various types of measurement data obtained with the testbed. The data set is expected to facilitate and promote the experimental evaluation of jamming detection, jamming avoidance, and anti-jamming schemes developed by the wireless security community. As an illustration, we trained five different machine learning algorithms and got results with an overall jamming detection accuracy of more than 90% for all algorithms.

The wireless networks of the future are evolving to enable reliable communication for resource-constrained, miniaturized internet of things (IoT) devices, which will place stringent demands on future sixth-generation (6G) mobile networks. These demands include low cost, very low latency, improved spectrum, and power efficiency, higher reliability, and significantly improved data rates. Emphasizing that these devices have limited functionality and may be placed in inaccessible locations, replacing or recharging batteries can be a daunting task, so energy-efficient solutions should be developed to ensure uninterrupted, seamless wireless communication for the power-constrained IoT devices. In this paper, we consider the integration of long-range (LoRa) modulation into backscatter communications (BackCom), and we develop a mathematical framework in order to investigate the error rate performance of the considered system model. In particular, we derive novel exact and approximated closed-form expressions for the symbol error rate (SER), under the assumption of canceled radio-frequency (RF) interference. The obtained analytical results, corroborated by numerical results, confirm the advantages of integrating LoRa into BackCom system as a low-complex technique in order to extend the transmission distance in power-limited backscatter devices.