Blockchain naturally fits multiple industry sectors due its characteristics of decentralization, enhanced security, tamper-proof, improved traceability and transparency. However, there is a significant concern of blockchain’s performance, since blockchain trades off its performance for a completely distributed feature, which enhances its security. In this paper, we investigate the state-of-the-art progress of blockchain, mainly from a performance and security perspective. We extracted 42 primary papers from major scientific databases and 34 online technical articles. The objective is to understand the current research trends, challenges and future directions. We briefly introduce the key technologies of blockchain including distributed ledger, cryptography, consensus, smart contracts and benchmarks. We next summarize the performance and security concerns raised in the investigation. We discuss the architectural choices, performance metrics, database management enhancements, and hybrid blockchains, and try to identify the effort that the state-of-the-art has made to balance between the performance and security. We also make experiments on Ethereum and survey other popular blockchain platforms on the scalability feature of blockchain. We later discuss the potential applications and present the lessons learned. Finally, we attempt to identify the open issues and possible research directions.

As a major strategy to ensure the safety of IT infrastructure, anomaly detection plays a more important role in cloud computing platform which hosts the entire applications and data. On top of the classic Markov chain model, we proposed in this paper a feasible multi-order Markov chain based framework for anomaly detection. In this approach, both the high-order Markov chain and multivariate time series are adopted to compose a scheme described in algorithms along with the training procedure in the form of statistical learning framework. To curb time and space complexity, the algorithms are designed and implemented with non-zero value table and logarithm values in initial and transition matrices. For validation, the series of system calls and the corresponding return values are extracted from classic Defense Advanced Research Projects Agency (DARPA) intrusion detection evaluation data set to form a two-dimensional test input set. The testing results show that the multi-order approach is able to produce more effective indicators: in addition to the absolute values given by an individual single-order model, the changes in ranking positions of outputs from different-order ones also correlate closely with abnormal behaviours.

With the expected explosive use of the Internet of Everything in sixth generation (6G), the cybertwin network is able to convert user information to digital assets and provide extensive services. However, protecting and enhancing privacy of the processed and transmitted data in cybertwin-driven 6G is still in its infancy. Federated learning (FL) is a nascent distributed machine learning paradigm that is able to facilitate privacy protection in cybertwin networks. In a cybertwin network, imbalanced data distribution of the clients can increase the bias of the global model and sacrifice the performance of the FL model. Prior research work dealing with imbalanced data requires extra data information exchanged between clients and the server, which increases the risk of privacy leakage. To avoid privacy leakage, we design an estimation algorithm to determine the distribution of local data collected at the clients without the awareness of specific raw data. We consider two scenarios in FL: 1) the server could receive the individual trained model for each selected device and 2) the server could receive the aggregated model from the selected clients. We formulate two device selection problems to improve the training performance of the aforementioned scenarios. We develop two online learning algorithms to tackle the selection problems for both individual model uploading and aggregated model uploading. The proposed algorithms are conducted on the server, thereby avoiding privacy leakage and extra computation at the clients. We validate the effectiveness of the proposed client selection algorithms with sufficient experiments in cybertwin-driven 6G networks.

The rapid development in wind power comes with new technical challenges. Reliable and accurate wind power forecast is of considerable significance to the electricity system’s daily dispatching and production. Traditional forecast methods usually utilize wind speed and turbine parameters as the model inputs. However, they are not sufficient to account for complex weather variability and the various wind turbine features in the real world. Inspired by the excellent performance of convolutional neural networks (CNN) in computer vision, we propose a novel approach to predicting short-term wind power by converting time series into images and exploit a CNN to analyze them. In our approach, we first propose two transformation methods to map wind speed and precipitation data time series into image matrices. After integrating multi-dimensional information and extracting features, we design a novel CNN framework to forecast 24-h wind turbine power. Our method is implemented on the Keras deep learning platform and tested on 10 sets of 3-year wind turbine data from Hangzhou, China. The superior performance of the proposed method is demonstrated through comparisons using state-of-the-art techniques in wind turbine power forecasting.