Search

Recent progress in parallel computing machines, deep neural networks, and training techniques have contributed to the significant advances in artificial intelligence (AI) with respect to tasks such as object classification, speech recognition, and natural language processing. The development of such deep learning-based techniques has enabled AI-based networks to outperform humans in the recognition of objects in images. The graphics processing unit (GPU) has been the primary component used for parallel computing during the inference and training phases of deep neural networks. In this study, we perform training using a desktop or a server with one or more GPUs and inference using hardware accelerators on embedded devices. Performance, power consumption, and requirements of embedded system present major hindrances to the application of deep neural network-based systems using embedded controllers such as drones, AI speakers, and autonomous vehicles. In particular, power consumption of a commercial GPU commonly surpasses the power budget of a stand-alone embedded system. To reduce the power consumption of hardware accelerators, reductions in the precision of input data and hardware weight have become popular topics of research in this field. However, precision and accuracy share a trade-off relationship. Therefore, it is essential to optimize precision in a manner that does not degrade the accuracy of the inference process. In this context, the primary issues faced by hardware accelerators are loss of accuracy and high power consumption.

Advances in the architecture of inference accelerators and quantization techniques of neural networks allow effective on-device inference in embedded devices. Privacy issues for user data, as well as increasing needs of user-specific services, have led to a need for on-device training. The dot product operation required in backpropagation can be computed efficiently by multiplier–accumulators (MACs) in the inference accelerator if forward and backward propagation of the neural network have the same precision. This chapter introduces a quantization technique to enable computation by the digital neuron inference accelerator with the same precision as that using the forward path. Updating the 5-bit weights with gradients of higher precision is challenging. To address this issue, this chapter also introduces a probabilistic weight update. It also describes the hardware implementation of the probabilistic weight-update scheme. The proposed training technique achieves 98.15% recognition accuracy on the MNIST dataset.

With continuously changing operational and business needs of the organizations, Decentralized Autonomous Organizations (DAO) is the current need of the organizations. Centralized Autonomous Organization (CAO) lack transparency and are managed by few efficient managers whereas Decentralized autonomous Organization's (DAO) is novel scalable, self-organizing coordination on the blockchain, controlled by smart contracts and its essential operations are automated agreeing to rules and principles assigned in code without human involvement. In this chapter we discuss the needs for Decentralized Autonomous Organizations (DAO) and key efforts in this field. We then introduce a prospective solution employing blockchain Ethereum, which incorporates a Turing complete programming language with smart contract computing functionality. A solution is elaborated that permits the formation of organizations where participants preserve straight real-time check of contributed collects and governance policies are formalized, automatized and imposed using software. Basic code for smart contract is composed to make a Decentralized Autonomous Organization (DAO) on the Ethereum blockchain. We also explain the working of DAOs code, centering on fundamental establishment and governance characteristics, which includes organization, formation and voting rights. DAOs are considered to agree to the expectation of the business work in the future. But there is still lack of operational base for DAOs in the blockchain community.

The advancement of new technology has taken a huge leap in the last few decades. It is bringing a drastic change in each step of human life that are capable of performing intelligent tasks. The internet of things and blockchain are disruptive technologies that have received a huge attention from industry, academic and financial technologies. There is a risk of privacy leakage of sensitive information in the centralized IoT system because the centralized servers can access the plain text data from the IoT devices. There is an extensive interest in applying the blockchain in the IoT system to provide IoT data privacy and decentralized access model. However, the previous blockchain-based IoT systems have issues related to privacy leakage of sensitive information to the servers as the servers can access the plaintext data from the IoT devices. So, we present the potential of integration of blockchain based-IoT with homomorphic encryption that can secure the IoT data with high privacy in a decentralized mode. In addition, we provide comparison of the recent technologies toward the securing and preserving the privacy of the IoT data using blockchain and homomorphic encryption technology. We have also highlighted the research challenges and possible future research directions in the integrated blockchain-based IoT with homomorphic encryption.

The intelligent vehicle communication network is prone to cyberthreats, which are difficult to solve using traditional centralized security approaches. Blockchain is an immutable peer-to-peer distributed database containing cryptographically secured information. Blockchain shows successful use cases in financial applications, smart contact, protecting digital copyright of media contents. It extends to all industries including the secure IoT devices, embedded systems, etc. The superior feature of blockchain is its decentralized, immutable, auditable database that secures transactions by protecting privacy. In this chapter, we contemplate the environment of the intelligent vehicle communication network and issues regarding methods of building a blockchain-based trust network among intelligent vehicles. We present the use cases of blockchain in intelligent vehicles in the phase of ongoing research or that under development from automotive industries and academic institutes. We also deliberate the challenging issue of blockchain for intelligent vehicles.