Machine learning (ML) is a key technique for big-data-driven modelling and analysis of massive Internet of Things (IoT) based intelligent and ubiquitous computing. For fast-increasing applications and data amounts, distributed learning is a promising emerging paradigm since it is often impractical or inefficient to share/aggregate data to a centralized location from distinct ones. This paper studies the problem of training an ML model over decentralized systems, where data are distributed over many user devices and the learning algorithm run on-device, with the aim of relaxing the burden at a central entity/server. Although gossip-based approaches have been used for this purpose in different use cases, they suffer from high communication costs, especially when the number of devices is large. To mitigate this, incremental-based methods are proposed. We first introduce incremental block-coordinate descent (I-BCD) for the decentralized ML, which can reduce communication costs at the expense of running time. To accelerate the convergence speed, an asynchronous parallel incremental BCD (API-BCD) method is proposed, where multiple devices/agents are active in an asynchronous fashion. We derive convergence properties for the proposed methods. Simulation results also show that our API-BCD method outperforms state of the art in terms of running time and communication costs.

Edge computing provides a promising paradigm to support the implementation of Internet of Things (IoT) by offloading tasks to nearby edge nodes. Meanwhile, the increasing network size makes it impractical for centralized data processing due to limited bandwidth, and consequently a decentralized learning scheme is preferable. Reinforcement learning (RL) has been widely investigated and shown to be a promising solution for decision-making and optimal control processes. For RL in a decentralized setup, edge nodes (agents) connected through a communication network aim to work collaboratively to find a policy to optimize the global reward as the sum of local rewards. However, communication costs, scalability, and adaptation in complex environments with heterogeneous agents may significantly limit the performance of decentralized RL. Alternating direction method of multipliers (ADMM) has a structure that allows for decentralized implementation and has shown faster convergence than gradient descent-based methods. Therefore, we propose an adaptive stochastic incremental ADMM (asI-ADMM) algorithm and apply the asI-ADMM to decentralized RL with edge-computing-empowered IoT networks. We provide convergence properties for the proposed algorithms by designing a Lyapunov function and prove that the asI-ADMM has O(1/k) + O(1/M) convergence rate, where k and M are the number of iterations and batch samples, respectively. Then, we test our algorithm with two supervised learning problems. For performance evaluation, we simulate two applications in decentralized RL settings with homogeneous and heterogeneous agents. The experimental results show that our proposed algorithms outperform the state of the art in terms of communication costs and scalability and can well adapt to complex IoT environments. 

Future communication networks call for new solutions to support their capacity and delay demands by leveraging potentials of the millimeter wave (mmWave) frequency band. However, the beam training procedure in mmWave systems incurs significant overhead as well as huge energy consumption. As such, deriving an adaptive control policy is beneficial to both delay-sensitive and energy-efficient data transmission over mmWave networks. To this end, we investigate the problem of joint beam training and data transmission control for mmWave delay-sensitive communications in this paper. Specifically, the considered problem is firstly formulated as a constrained Markov Decision Process (MDP), which aims to minimize the cumulative energy consumption over the whole considered period of time under delay constraint. By introducing a Lagrange multiplier, we transform the constrained MDP into an unconstrained one, which is then solved via a parallel-rollout-based reinforcement learning method in a data-driven manner. Our numerical results demonstrate that the optimized policy via parallel-rollout significantly outperforms other baseline policies in both energy consumption and delay performance.

Big data, including applications with high security requirements, are often collected and stored on multiple heterogeneous devices, such as mobile devices, drones, and vehicles. Due to the limitations of communication costs and security requirements, it is of paramount importance to analyze information in a decentralized manner instead of aggregating data to a fusion center. To train large-scale machine learning models, edge/fog computing is often leveraged as an alternative to centralized learning. We consider the problem of learning model parameters in a multiagent system with data locally processed via distributed edge nodes. A class of minibatch stochastic alternating direction method of multipliers (ADMMs) algorithms is explored to develop the distributed learning model. To address two main critical challenges in distributed learning systems, i.e., communication bottleneck and straggler nodes (nodes with slow responses), error-control-coding-based stochastic incremental ADMM is investigated. Given an appropriate minibatch size, we show that the minibatch stochastic ADMM-based method converges in a rate of O(1/root k), where k denotes the number of iterations. Through numerical experiments, it is revealed that the proposed algorithm is communication efficient, rapidly responding, and robust in the presence of straggler nodes compared with state-of-the-art algorithms.

Cell-free networks are considered to be a promising distributed network architecture to satisfy the increasing number of users and high rate expectations in beyond-5G systems. However, to further enhance network capacity, an increasing number of high-cost base stations (BSs) is required. To address this problem and inspired by the cost-effective intelligent reflecting surface (IRS) technique, we propose a fully decentralized design framework for cooperative beamforming in IRS-aided cell-free networks. We first transform the centralized weighted sum-rate maximization problem into a tractable consensus optimization problem, and then an incremental alternating direction method of multipliers (ADMM) algorithm is proposed to locally update the beamformer. The complexity and convergence of the proposed method are analyzed, and these results show that the performance of the new scheme can asymptotically approach that of the centralized one as the number of iterations increases. Results also show that IRSs can significantly increase the system sum-rate of cell-free networks and the proposed method outperforms existing decentralized methods.

To handle the data explosion in the era of Internet-of-things, it is of interest to investigate the decentralized network, with the aim at relaxing the burden at the central server along with preserving data privacy. In this work, we develop a fully decentralized federated learning (FL) framework with an inexact stochastic parallel random walk alternating direction method of multipliers (ISPW-ADMM). Performing more efficient communication and enhanced privacy preservation compared with the current state-of-the-art, the proposed ISPW-ADMM can be partially immune to the effect of time-varying dynamic network and stochastic data collection, while still in fast convergence. Benefiting from the stochastic gradients and biased first-order moment estimation, the proposed framework can be applied to any decentralized FL tasks over time-varying graphs. Thus, to demonstrate the practicability of such a framework in providing fast convergence, high communication efficiency, noise robustness for a specific on-board mission to some extent, we study the extreme learning machine-based FL model beamforming design in unmanned aerial vehicle communications, as verified by the numerical simulations.

Millimeter Wave (mmWave) communications with full-duplex (FD) have the potential of increasing the spectral efficiency, relative to those with half-duplex. However, the residual self-interference (SI) from FD and high pathloss inherent to mmWave signals may degrade the system performance. Meanwhile, hybrid beamforming (HBF) is an efficient technology to enhance the channel gain and mitigate interference with reasonable complexity. However, conventional HBF approaches for FD mmWave systems are based on optimization processes, which are either too complex or strongly rely on the quality of channel state information (CSI). We propose two learning schemes to design HBF for FD mmWave systems, i.e., extreme learning machine based HBF (ELM-HBF) and convolutional neural networks based HBF (CNN-HBF). Specifically, we first propose an alternating direction method of multipliers (ADMM) based algorithm to achieve SI cancellation beamforming, and then use a majorization-minimization (MM) based algorithm for joint transmitting and receiving HBF optimization. To train the learning networks, we simulate noisy channels as input, and select the hybrid beamformers calculated by proposed algorithms as targets. Results show that both learning based schemes can provide more robust HBF performance and achieve at least 22.1% higher spectral efficiency compared to orthogonal matching pursuit (OMP) algorithms. Besides, the online prediction time of proposed learning based schemes is almost 20 times faster than the OMP scheme. Furthermore, the training time of ELM-HBF is about 600 times faster than that of CNN-HBF with 64 transmitting and receiving antennas.

In multi-task learning (MTL), related tasks learn jointly to improve generalization performance. To exploit the high learning speed of feed-forward neural networks (FNN), we apply the randomized single-hidden layer FNN (RSF) to the MTL problem, where the output weights of RSFs for all the tasks are learned collaboratively. We first present the RSF based MTL problem in the centralized setting, which is solved by the proposed MTL-RSF algorithm. Due to the fact that many data sets of different tasks are geo-distributed, decentralized machine learning is studied. We formulate the decentralized MTL problem based on RSF as majorized multi-block optimization with coupled bi-convex objective functions. To solve the problem, we propose the DMTL-RSF algorithm, which is a hybrid Jacobian and Gauss-Seidel Proximal multi-block alternating direction method of multipliers (ADMM). Further, to reduce the computation load of DMTL-RSF, DMTL-RSF with first-order approximation (FO-DMTL-RSF) is presented. Theoretical analysis shows that the convergence to the stationary point of proposed decentralized algorithms can be guaranteed conditionally. Through simulations, we demonstrate the convergence of presented algorithms, and also show that they can outperform existing MTL methods. Moreover, by adjusting the dimension of hidden feature space, there exists a trade-off between communication load and learning accuracy for DMTL-RSF.

Wireless content caching in cellular networks is an efficient way to reduce the service delay and alleviate backhaul pressure. For the benefits of sharing spectral and storage resources, clustering in cached networks has recently attracted significant research interests. Meanwhile, since the multimedia content (e.g., video) of caching networks may require a huge transmission rates, millimeter wave (mmWave) communication is considered to be an efficient transmission scheme for cache-enabled networks. We investigate the ergodic rate and average service delay for typical user terminal (UT) in the clustered cache-enabled small cell networks (SCN) and ultra dense networks (UDN) with mmWave channels. In SCN, each cluster consists of cache-enabled UTs, and in the UDN a cluster is formed by cache-enabled UTs and small base stations (SBSs) with non-uniform caching capacity. The clusters are assumed to be discs and content sharing is only possible within clusters through mmWave device-to-device (D2D) tier and SBS tier communications. With stochastic geometry methods, the distributions of content sharing distance and signal-to-interference-noise-ratio (SINR) of typical UT in a cluster are derived for both SCN and UDN scenarios. To minimize the average service delay in high SINR region, we provide an algorithm to jointly optimize caching scheme for SBSs and UTs. By simulations, we validate our theoretical analysis and the performance of proposed caching scheme. The numerical results also show that there exists best radius in the design of cluster for UDNs.