To process and transfer large amounts of data in emerging wireless services, it has become increasingly appealing to exploit distributed data communication and learning. Specifically, edge learning (EL) enables local model training on geographically disperse edge nodes and minimizes the need for frequent data exchange. However, the current design of separating EL deployment and communication optimization does not yet reap the promised benefits of distributed signal processing, and sometimes suffers from excessive signalling overhead, long processing delay, and unstable learning convergence. In this paper, we provide an overview on practical distributed EL techniques and their interplay with advanced communication optimization designs. In particular, typical performance metrics for dual-functional learning and communication networks are discussed. Also, recent achievements of enabling techniques for the dual-functional design are surveyed with exemplifications from the mutual perspectives of “communications for learning” and “learning for communications.” The application of EL techniques within a variety of future communication systems are also envisioned for beyond 5G (B5G) wireless networks. For the application in goal-oriented semantic communication, we present a first mathematical model of the goal-oriented source entropy as an optimization problem. In addition, from the viewpoint of information theory, we identify fundamental open problems of characterizing rate regions for communication networks supporting distributed learning-and-computing tasks. We also present technical challenges as well as emerging application opportunities in this field, with the aim of inspiring future research and promoting widespread developments of EL in B5G.

Edge Learning for B5G Networks With Distributed Signal Processing: Semantic Communication, Edge Computing, and Wireless Sensing

Mobile edge computing (MEC) and device-to-device (D2D) offloading are two promising paradigms in the industrial Internet of Things (IIoT). In this article, we investigate task co-offloading, where computing-intensive industrial tasks can be offloaded to MEC servers via cellular links or nearby IIoT devices via D2D links. This co-offloading delivers small computation delay while avoiding network congestion. However, erratic movements, the selfish nature of devices and incomplete offloading information bring inherent challenges. Motivated by these, we propose a co-offloading framework, integrating migration cost and offloading willingness, in D2D-assisted MEC networks. Then, we investigate a learning-based task co-offloading algorithm, with the goal of minimal system cost (i.e., task delay and migration cost). The proposed algorithm enables IIoT devices to observe and learn the system cost from candidate edge nodes, thereby selecting the optimal edge node without requiring complete offloading information. Furthermore, we conduct simulations to verify the proposed co-offloading algorithm. 

Task Co-Offloading for D2D-Assisted Mobile Edge Computing in Industrial Internet of Things

Mobile edge computing (MEC) and device-to-device (D2D) offloading are two promising paradigms in the industrial Internet of Things (IIoT). In this article, we investigate task co-offloading, where computing-intensive industrial tasks can be offloaded to MEC servers via cellular links or nearby IIoT devices via D2D links. This co-offloading delivers small computation delay while avoiding network congestion. However, erratic movements, the selfish nature of devices and incomplete offloading information bring inherent challenges. Motivated by these, we propose a co-offloading framework, integrating migration cost and offloading willingness, in D2D-assisted MEC networks. Then, we investigate a learning-based task co-offloading algorithm, with the goal of minimal system cost (i.e., task delay and migration cost). The proposed algorithm enables IIoT devices to observe and learn the system cost from candidate edge nodes, thereby selecting the optimal edge node without requiring complete offloading information. Furthermore, we conduct simulations to verify the proposed co-offloading algorithm.

Driven by the rapid development of the Internet of Things (IoT) technology, the issue of trust has become increasingly apparent and received considerable scholarly attention in recent years. With the occurrence of various security incidents, data leakage accidents, and service fraud, which seriously affects the quality of service and system efficiency of IoT, trust is fast becoming a key issue in IoT system. In addition to the security and efficiency, trust further contains the reliability, attack resistance, fairness, flexibility, and incentive, resulting in a wide range of new researches from the trust management framework to the quantification method. In this work, various dimensions of trust, including definition, composition, aggregation, and computation are introduced and analyzed. We focus on comprehensive comparison of the state-of-the-art trust management researches and the related applications. Besides, we explore and discuss the important challenges, including performance bottleneck, bidirectional trust, dynamic changes of context, privacy preserving, and cross-domain issue. Some potential enabling technologies for trust management are further analyzed. The objective of this article is to comprehend the trust issue and the composition of trust management in IoT, and illustrate the difference of existing work, thereby motivating further research interest in this field.

Trust Management for Internet of Things: A Comprehensive Study

Vehicular ad-hoc network enhances driving safety and enables various intelligent transportation applications by adopting the revolutionary vehicular wireless communication technology. This has attracted a lot of attentions from both academia and industry in recent years. Given the sophistication of vehicular manufacturing and the heterogeneity of intelligent transport terminals, performing vehicular authentication is of great importance. The existing schemes have largely considered vehicle security and authentication within a single administrative domain, which lacks supervision of the authority and entity in the intelligent transportation system. In this article, we propose a multidomain vehicular authentication architecture by introducing blockchain technique to build distributed trust and share cross-domain information among multiple administrative domains. To guarantee the anonymity and traceability, a pseudonym-based privacy-preserving authentication method is proposed. Specifically, considering the supervision of authority and the resilience to key escrow, we design a two-phase pseudonym distribution mechanism with the assistance of a roadside unit (RSU) proxy. We conduct in-depth security analysis by comparing with existing works and deploy experiments to show the efficiency and feasibility of the proposed scheme in the multidomain scenario. 

A Blockchain-Based Multidomain Authentication Scheme for Conditional Privacy Preserving in Vehicular Ad-Hoc Network

The emerging mobile-edge computing paradigm provides opportunities for the resource-hungry mobile devices (MDs) to migrate computation. In order to satisfy the requirements of MDs in terms of latency and energy consumption, recent researches proposed diverse computation offloading schemes. However, they either fail to consider the potential computing resources at the edge, or ignore the selfish behavior of users and the dynamic resource adaptability. To this end, we study the computation offloading problem and take into consideration the dynamic available resource of idle devices and the selfish behavior of users. Furthermore, we propose a game theoretic offloading method by regarding the computation offloading process as a resource contention game, which minimizes the individual task execution cost and the system overhead. Utilizing the potential game, we prove the existence of Nash equilibrium (NE), and give a lightweight algorithm to help the game reach a NE, wherein each user can find an optimal offloading strategy based on three contention principles. Additionally, we conduct analysis of computational complexity and the Price of Anarchy (PoA), and deploy three baseline methods to compare with our proposed scheme. Numerical results illustrate that our scheme can provide high-quality services to users, and also demonstrate the effectiveness, scalability and dynamic resource adaptability of our proposed algorithm in a multiuser network.

D2D-Enabled Mobile-Edge Computation Offloading for Multiuser IoT Network

With the rapid development of Internet of Things, the demand of data sharing is not only to ensure the data integrity, but also to protect user's individual privacy. Secure multi-party computation, as a technology paradigm based on cryptography technology, enables privacy protection in data sharing among multiple parties, while the implement of secure multi-party computation is limited due to the inefficient, complex computation protocol and frequent interaction. Fortunately, the emergence of blockchain technology provides excellent properties of decentralization, verifiability and high reliability for secure multi-party computation. In this paper, we propose a blockchain-based secure multi-party computation architecture for data sharing, where we design an aggregator consortium for data storage, verification and joint computation. Moreover, we introduce a straightforward secure multi-party computation scheme based on homomorphic encryption. In addition, we analyze and discuss this scheme in terms of data security, access flexibility, attack resistance and potential applications.

Block-SMPC: A Blockchain-based Secure Multi-party Computation for Privacy-Protected Data Sharing

Federated learning (FL), as a disruptive machine learning (ML) paradigm, enables the collaborative training of a global model over decentralized local data sets without sharing them. It spans a wide scope of applications from the Internet of Things (IoT) to biomedical engineering and drug discovery. To support low-latency and high-privacy FL over wireless networks, in this article, we propose a reconfigurable intelligent surface (RIS)-empowered over-the-air FL system to alleviate the dilemma between learning accuracy and privacy. This is achieved by simultaneously exploiting the channel propagation reconfigurability with RIS for boosting the received signal power, as well as the waveform superposition property with over-the-air computation (AirComp) for fast model aggregation. By considering a practical scenario, where high-dimensional local model updates are transmitted across multiple communication blocks, we characterize the convergence behaviors of the differentially private federated optimization algorithm. We further formulate a system optimization problem to optimize the learning accuracy while satisfying privacy and power constraints via the joint design of transmit power, artificial noise, and phase shifts at RIS, for which a two-step alternating minimization framework is developed. Simulation results validate our systematic, theoretical, and algorithmic achievements and demonstrate that RIS can achieve a better tradeoff between privacy and accuracy for over-the-air FL systems.

Differentially Private Federated Learning via Reconfigurable Intelligent Surface

With the rapid increase of the industrial data and the development of the industrial Internet of Things (IIoT) paradigm, the efficiency and the quality of service of the emerging applications have been improved. However, the contradiction between data sharing and privacy preserving is still an obstacle in the IIoT. To this end, in this article, we propose a privacy-preserving and publicly auditable multiparty computation scheme for industrial data sharing and computing, which avoids privacy leakage and computation misbehavior by separating the data ownership, data use, and data verification. Using the blockchain technology, a transparent management platform is provided to recognize and trace the illegal data and computation behavior. Moreover, we integrate the noninteractive zero-knowledge proof in the multiparty interaction mechanism, wherein the verification of data consistency and computation validity is executed publicly on the blockchain. Finally, we implement experiment to evaluate the performance of the computation latency, communication overhead and the influence of encryption parameter, and the numerical results illustrate the efficiency and feasibility of our scheme.

Yuhan Yang

Papers

D2D-Enabled Mobile-Edge Computation Offloading for Multiuser IoT Network

Trust Management for Internet of Things: A Comprehensive Study

Block-SMPC: A Blockchain-based Secure Multi-party Computation for Privacy-Protected Data Sharing

Differentially Private Federated Learning via Reconfigurable Intelligent Surface

Blockchain-Enabled Multiparty Computation for Privacy Preserving and Public Audit in Industrial IoT