Deep learning (DL) has been widely proposed for botnet attack detection in Internet of Things (IoT) networks. However, the traditional centralized DL (CDL) method cannot be used to detect the previously unknown (zero-day) botnet attack without breaching the data privacy rights of the users. In this article, we propose the federated DL (FDL) method for zero-day botnet attack detection to avoid data privacy leakage in IoT-edge devices. In this method, an optimal deep neural network (DNN) architecture is employed for network traffic classification. A model parameter server remotely coordinates the independent training of the DNN models in multiple IoT-edge devices, while the federated averaging (FedAvg) algorithm is used to aggregate local model updates. A global DNN model is produced after a number of communication rounds between the model parameter server and the IoT-edge devices. The zero-day botnet attack scenarios in IoT-edge devices is simulated with the Bot-IoT and N-BaIoT data sets. Experiment results show that the FDL model: 1) detects zero-day botnet attacks with high classification performance; 2) guarantees data privacy and security; 3) has low communication overhead; 4) requires low-memory space for the storage of training data; and 5) has low network latency. Therefore, the FDL method outperformed CDL, localized DL, and distributed DL methods in this application scenario.

Transaction-based energy (TE) management and control has become an increasingly relevant topic, attracting considerable attention from industry and the research community alike. As a result, new techniques are emerging for its development and actualization. This paper presents a comprehensive review of TE involving peer-to-peer (P2P) energy trading and also covering the concept, enabling technologies, frameworks, active research efforts and the prospects of TE. The formulation of a common approach for TE management modelling is challenging given the diversity of circumstances of prosumers in terms of capacity, profiles and objectives. This has resulted in divergent opinions in the literature. The idea of this paper is therefore to explore these viewpoints and provide some perspectives on this burgeoning topic on P2P TE systems. This study identified that most of the techniques in the literature exclusively formulate energy trade problems as a game, an optimization problem or a variational inequality problem. It was also observed that none of the existing works has considered a unified messaging framework. This is a potential area for further investigation.

Rising awareness and emergence of smart technologies have inspired new thinking in energy system management. Whilst integration of distributed energy resources in micro-grids (MGs) has become the technique of choice for consumers to generate their energy, it also provides a unique opportunity to explore energy trading and sharing amongst them. This paper investigates peer-to-peer (P2P) communication architectures for prosumers’ energy trading and sharing. The performances of common P2P protocols are evaluated under the stringent communication requirements of energy networks defined in IEEE 1547.3-2007. Simulation results show that the structured P2P protocol exhibits a reliability of 99.997% in peer discovery and message delivery whilst the unstructured P2P protocol yields 98%, both of which are consistent with the requirements of MG applications. These two architectures exhibit high scalability with a latency of 0.5 s at a relatively low bandwidth consumption, thus, showing promising potential in their adoption for prosumer to prosumer communication.

Multisource energy data, including from distributed energy resources and its multivariate nature, necessitate the integration of robust data predictive frameworks to minimise prediction error. This work presents a hybrid deep learning framework to accurately predict the energy consumption of different building types, both commercial and domestic, spanning different countries, including Canada and the UK. Specifically, we propose architectures comprising convolutional neural network (CNN), an autoencoder (AE) with bidirectional long short-term memory (LSTM), and bidirectional LSTM BLSTM). The CNN layer extracts important features from the dataset and the AE-BLSTM and LSTM layers are used for prediction. We use the individual household electric power consumption dataset from the University of California, Irvine to compare the skillfulness of the proposed framework to the state-of-the-art frameworks. Results show performance improvement in computation time of 56% and 75.2%, and mean squared error (MSE) of 80% and 98.7% in comparison with a CNN BLSTM-based framework (EECP-CBL) and vanilla LSTM, respectively. In addition, we use various datasets from Canada and the UK to further validate the generalisation ability of the proposed framework to underfitting and overfitting, which was tested on real consumers’ smart boxes. The results show that the framework generalises well to varying data and constraints, giving an average MSE of ∼0.09 across all datasets, demonstrating its robustness to different building types, locations, weather, and load distributions.

Blockchain (BC) and artificial intelligence (AI) are often utilised separately in energy trading systems (ETS). However, these technologies can complement each other and reinforce their capabilities when integrated. This paper provides a comprehensive review of consensus algorithms (CA) of BC and deep reinforcement learning (DRL) in ETS. While the distributed consensus underpins the immutability of transaction records of prosumers, the deluge of data generated paves the way to use AI algorithms for forecasting and address other data analytic related issues. Hence, the motivation to combine BC with AI to realise secure and intelligent ETS. This study explores the principles, potentials, models, active research efforts and unresolved challenges in the CA and DRL. The review shows that despite the current interest in each of these technologies, little effort has been made at jointly exploiting them in ETS due to some open issues. Therefore, new insights are actively required to harness the full potentials of CA and DRL in ETS. We propose a framework and offer some perspectives on effective BC-AI integration in ETS.

Consumer-centric Internet of Things (CIoT) will play a pivotal role in the fifth industrial revolution (Industry 5.0) but it exhibits vulnerabilities that can render it susceptible to various cyberattacks. Recent studies have explored the potential of Federated Learning (FL) for privacy-preserving intrusion detection in IoT. However, the development of the FL models relied on unrealistic and irrelevant network traffic data, while also exhibiting limitations in terms of covered attack types and classification scenarios. In this paper, we develop Federated Deep Learning (FDL) models using three recent and highly relevant datasets, covering a wide range of attack types as well as binary and multi-class classification scenarios. Our findings demonstrate that the FDL models not only achieve high classification performance, comparable to traditional Centralized Deep Learning (CDL) models, in terms of accuracy (99.60±0.46%), precision (92.50±8.40%), recall (95.42±6.24%), and F1 score (93.51±7.76%) but also exhibit superior computational efficiency compared to their CDL counterparts. The FDL approach reduces the training time by 30.52-75.87%. These classification performance and computational efficiency were achieved through multiple rounds of distributed local training in FDL. Therefore, the proposed FDL framework presents a robust security solution for designing and deploying a resilient CIoT.