Decentralized security and data integrity of blockchain using deep learning techniques

Abstract

Since the introduction of blockchain, cryptocurrencies have become very attractive as an alternative digital payment method and a highly speculative investment. With the rise in computational power and the growth of available data, the artificial intelligence concept of deep neural networks had a surge of popularity over the last years as well. With the introduction of the long short-term memory (LSTM) architecture, neural networks became more efficient in understanding long-term dependencies in data such as time series. In this research paper, we combine these two topics, by using LSTM networks to make a prognosis of decentralized blockchain security. In particular, we test if LSTM based neural networks can produce profitable trading signals for different blockchains. We experiment with different preprocessing techniques and different targets, both for security regression and trading signal classification. We evaluate LSTM based networks. As data for training we use historical security data in one-minute intervals from August 2019 to August 2020. We measure the performance of the models via back testing, where we simulate trading on historic data not used for training based on the model’s predictions. We analyze that performance and compare it with the buy and hold strategy. The simulation is carried out on bullish, bearish and stagnating time periods. In the evaluation, we find the best performing target and pinpoint two preprocessing combinations that are most suitable for this task. We conclude that the CNN LSTM hybrid is capable of profitably forecasting trading signals for securing blockchain, outperforming the buy and hold strategy by roughly 30%, while the performance was better. The LTSM method used by current system for encrypting passwords is efficient enough to mitigate modern attacks like man in the middle attack (MITM) and DDOS attack with 95.85% accuracy.