Artificial intelligence (AI) is envisioned to be natively integrated into the sixth-generation (6G) mobile networks to support a diverse range of intelligent applications. Federated edge learning (FEEL) emerges as a vital enabler of this vision by leveraging the sensing, communication, and computation capabilities of geographically dispersed edge devices to collaboratively train AI models without sharing raw data. This article explores the pivotal role of FEEL in advancing both the “wireless for AI” and “AI for wireless” paradigms, thereby facilitating the realization of scalable, adaptive, and intelligent 6G networks. We begin with a comprehensive overview of learning architectures, models, and algorithms that form the foundations of FEEL. We, then, establish a novel task-oriented communication principle to examine key methodologies for deploying FEEL in dynamic and resource-constrained wireless environments, focusing on device scheduling, model compression, model aggregation, and resource allocation. Furthermore, we investigate the domain-specific optimizations of FEEL to facilitate its promising applications, ranging from wireless air-interface technologies to mobile and the Internet of Things (IoT) services. Finally, we highlight key future research directions for enhancing the design and impact of FEEL in 6G.

In this paper, a novel transmissive reconfigurable meta-surface (RMS) transceiver enabled multi-tier computing network architecture is proposed for improving computing capability, decreasing computing delay and reducing base station (BS) deployment cost, in which transmissive RMS equipped with a feed antenna can be regarded as a new type of multi-antenna system. We formulate a total energy consumption minimization problem by a joint optimization of subcarrier allocation, task input bits, time slot allocation, transmit power allocation and RMS transmissive coefficient while taking into account the constraints of communication resources and computing resources. This formulated problem is a non-convex optimization problem due to the high coupling of optimization variables, which is NP-hard to obtain its optimal solution. To address the above challenging problems, block coordinate descent (BCD) technique is employed to decouple the optimization variables to solve the problem. Specifically, the joint optimization problem of subcarrier allocation, task input bits, time slot allocation, transmit power allocation and RMS transmissive coefficient is divided into three subproblems to solve by applying BCD. Then, the decoupled three subproblems are optimized alternately by using successive convex approximation (SCA) and difference-convex (DC) programming until the convergence is achieved. Numerical results verify that our proposed algorithm is superior in reducing total energy consumption compared to other benchmarks.

The thriving of artificial intelligence (AI) applications is driving the further evolution of wireless networks. It has been envisioned that 6G will be transformative and will revolutionize the evolution of wireless from “connected things” to “connected intelligence”. However, state-of-the-art deep learning and big data analytics based AI systems require tremendous computation and communication resources, causing significant latency, energy consumption, network congestion, and privacy leakage in both of the training and inference processes. By embedding model training and inference capabilities into the network edge, edge AI stands out as a disruptive technology for 6G to seamlessly integrate sensing, communication, computation, and intelligence, thereby improving the efficiency, effectiveness, privacy, and security of 6G networks. In this paper, we shall provide our vision for scalable and trustworthy edge AI systems with integrated design of wireless communication strategies and decentralized machine learning models. New design principles of wireless networks, service-driven resource allocation optimization methods, as well as a holistic end-to-end system architecture to support edge AI will be described. Standardization, software and hardware platforms, and application scenarios are also discussed to facilitate the industrialization and commercialization of edge AI systems.

Driven by the vision of intelligent connection of everything and digital twin towards 6G, a myriad of new applications, such as immersive extended reality, autonomous driving, holographic communications, intelligent industrial internet, will emerge in the near future, holding the promise to revolutionize the way we live and work. These trends inspire a novel technical design principle that seamlessly integrates two originally decoupled functionalities, i.e., wireless communication and sensing, into one system in a symbiotic way, which is dubbed symbiotic sensing and communications (SSaC), to endow the wireless network with the capability to “see” and “talk” to the physical world simultaneously. Noting that the term SSaC is used instead of ISAC (integrated sensing and communications) because the word “symbiotic/symbiosis” is more inclusive and can better accommodate different integration levels and evolution stages of sensing and communications. Aligned with this understanding, this article makes the first attempts to clarify the concept of SSaC, illustrate its vision, envision the three-stage evolution roadmap, namely neutralism, commensalism, and mutualism of SaC. Then, three categories of applications of SSaC are introduced, followed by detailed description of typical use cases in each category. Finally, we summarize the major performance metrics and key enabling technologies for SSaC.

To meet the tremendous demand of Internet of Things (IoT) applications, the 3rd Generation Partnership Project (3GPP) has specified the Narrowband IoT (NB-IoT) standard. However, collisions in the random access (RA) channel of NB-IoT can be severe due to mismatch between frequent random access attempts by a huge number of devices and the extremely limited radio resources. In this paper, a new random access mechanism called time-alignment-value based random access (TARA) is designed to increase the efficiency of the random access channel. The key idea of TARA is to conduct quick retries by an RA-failure device based on the time-alignment (TA) values of colliding UEs in a random access response (RAR) message. Simulation results show that TARA performs much better than the original slotted Aloha mechanism, i.e., higher success probability, higher throughput, and lower access delay. Comparison with other existing schemes further demonstrates high performance and advantages of TARA.

Asynchronous transmission for 5G massive machine type communication (mMTC) reduces power consumption and provides longer battery life of mMTC devices compared to synchronous transmission because of less signaling interaction between the base station and mMTC devices. To reduce frequency leakage from other user devices, several waveforms based on modification on orthogonal frequency division multiple (OFDM) with low out-of-band (OOB) emission property has been proposed in 5G new radio. However, those waveforms are typically of high peak to average power ratio (PAPR) and it has been shown for those waveforms there is OOB emission regrowth for mMTC devices caused by nonlinear power amplify (PA). We propose a low PAPR filter bank single carrier frequency division multiple access (SC-FDMA) waveform, called LP-FBSC, which helps to resist OOB emission regrowth caused by nonlinear PA. Two filters are applied in one stream to achieve low PAPR and low OOB emission performance, where one filter is poly phase network (PPN) structure in time domain for low OOB emission purpose, and the other is applied at frequency domain with element-wise multiplication for low PAPR purpose. In addition, the proposed method introduces multiple streams for purpose of even further reduced PAPR, where the modulated data and filter coefficients of the other streams are determined from those of first stream. For Pi/2-BPSK and Pi/4-QPSK modulation, the PAPR for LP-FBSC can be as low as 2.6 dB and 4.2 dB while low OOB emission property being maintained.

Multi-panel MIMO is a promising technology in millimeter wave communications. Due to its partially hybrid structure and non-uniform antenna array, existing channel estimation and hybrid precoding cannot be directly applied to multipanel MIMO. In this paper, we study channel estimation and hybrid precoding for multi-panel MIMO. We first transform the channel response vector into angular domain and then reconstruct channel state information (CSI) by using the estimated angular CSI. Moreover, by exploiting the sparse nature of mmWave channels, a sparsity-based channel estimation method is proposed to reduce training overhead and the computational complexity of hybrid precoding. Numerical results show that the proposed channel estimation and hybrid precoding scheme achieve satisfactory spectral efficiency performance with greatly reduced training overhead and low computational complexity.

As the clock frequency of a chip increases, the on-chip decoupling capacitor must be placed closer to the load to be effective. A method of efficiently defining the location of capacitor placement to meet specified noise limits is presented. Based on a single RL line model for the power distribution system, charging radius of the decoupling capacitor is calculated under the constraint of the target noise but not the constraint of being fully charged. Under the constraints of the length of the charging path and the target noise, discharging radius of the decoupling capacitor is figured out. The conversion coefficient that converts the radii of a decoupling capacitor in a single RL line model into the equivalent one in a meshed model is presented based on the expression that can exactly compute the impedance between any two points on an infinite meshed network. In this paper, it is shown that the conversion coefficient is a constant when the radius of the decoupling capacitor is taken as the per unit length of the meshed model.

We propose a fast algorithm to compute an initial triangular square-root of the estimation error covariance matrix for BLAST, which are then applied to develop a square-root algorithm for BLAST. The speedups of our square-root BLAST algorithm over the previous square-root BLAST algorithm in the number of multiplications and additions are 3.78-5.8 and 3.95-5 respectively, and the ratios between the computational complexity of our BLAST algorithm and that of the linear MMSE detection algorithm in the number of multiplications and additions are 1.10-0.71 and 0.90-0.71 respectively, which means that for the first time, the nonlinear MMSE BLAST detector with successive interference cancellation can have even lower complexity than the linear MMSE detector. Moreover, our BLAST algorithm is also numerically stable and hardware friendly, since it uses unitary transformations to avoid the matrix inversions, and gets the initial square-root which is equivalent to a Cholesky factor of the estimation error covariance matrix.

The communication system is currently undergoing a convergence to IP services. As a result, voice over internet protocol (VoIP) will be commonplace in the near future. In order to maximize the voice capacity, the overhead associated with controlling VoIP transmissions must be carefully managed. The current efforts in B3G (beyond 3G) standards development to efficiently control VoIP transmissions by grouping VoIP users into scheduling groups, assigning the group a set of shared time-frequency resources, and using bitmap signaling to allocate resources were detailed in [1]. This paper introduces several improvement mechanisms which enhance the basic concept outlined in [1]. System level simulations are used to validate the improved signaling technique and show that the technique can efficiently support 133 VoIP users per megahertz in a VoIP only system and 64 VoIP users per megahertz plus 1.05 Mbps for the traffic mixed considered in a mixed VoIP/data system.

With channel state information (CSI) at the transmitter, beamforming can be used for spatial diversity and multiple spatial access. Due to latency and feedback bandwidth limitation, the CSI at the transmitter is often known with some ambiguity. In this paper, we develop a robust method for downlink beamforming that takes the ambiguity of the CSI into consideration. It is shown by computer simulation that, compared with the existing method, the required signal-to-noise ratio (SNR) for a 1% bit-error rate (BER) is reduced by over 2 dB for a system with 4 transmit antennas and 2 users when the variance of the CSI is -20 dB. The performance gain increases with the number of transmit antennas when the number of users is fixed. The required SNR for a 1% BER is reduced by over 4 dB if the the number of transmitter antennas is 8. We also study the impact of power allocation on the downlink beamforming.

Transmit and receive antenna arrays can be used to form multiple-input and multiple-output (MIMO) systems for improving the reliability and capacity of data transmission. Layered space-time coding with decision-feedback detection is a promising technique for future wireless communications. In this paper, we investigate power allocation at the transmitter to improve the performance of the decision-feedback detection. The proposed power allocation method may only depend on or is even independent of the signal-to-noise ratio (SNR) of the MIMO systems. When the SNR at the transmitter is not available, the power of each data stream is allocated according to the required bit-error-rate (BER) of the system. Computer simulation shows that the proposed method can improve the performance of a 2-input and 2-output system by 4 dB at 1% BER and that of a 4-input and 4-output system by 3.5 dB.