Challenges

Current IoT ecosystems rely heavily on centralized communication models, such as pure cloud-based architectures, which excel in providing huge data storage and high levels of processing power. However, there are issues in the cloud-based IoT, including extensive security risks, poor scalability, privacy concerns, large data storage.

high costs, and inadequate infrastructures. With an ever-growing need for more and more IoT nodes, scalability becomes a major concern in such centralized approaches for constructing an IoT application. Extensive security risks have become a major concern in recent years, due to poor security safeguards in network edge devices such as printers, smoke detectors, etc. Poor scalability results in communication jitters and timeliness problems in real-time applications, including critical safety systems. Sending all the data to the cloud also results in privacy concerns, for example, face recognition systems that send all detected facial images to the cloud servers. In real- world applications, scalability, security, and privacy concerns eventually hinder the adoption of the IoT. A very promising solution is the edge-based blockchain (eBC) technology, as proposed in this chapter. This proposal has a distributed architecture with a tamper-proof ledger mechanism. Thus, it can address the three problems, including security (via a tamper-proof ledger), scalability (via distributed processing), and privacy (via edge-based processing).

Blockchain is an emerging technology that started from cryptocurrencies, such as Bitcoin, but has now been widely applied to various different fields, with the IoT being one of the more prominent applications. For IoT nodes to be authenticated when connecting to an IoT infrastructure, a safe decentralized mechanism such as blockchain makes the whole process seamless, automatic, and safe. Thus, the use of blockchain in the IoT has great potential and is very much required. Besides authentication, all interactions among IoT nodes can be validated easily by the blockchain mechanism, via hashing and storage in a tamper-proof ledger. The conventional issue of extensive security risks in a growing complex multi-node architecture also disappears when Blockchain is used, because the greater the number of IoT nodes authenticated, the safer all the transactions will be, because the transactions must be validated by a majority of the existing authenticated nodes. Thus, an IoT platform operated by a decentralized blockchain setup is so robust that it is almost impossible for a hacker to successfully compromise the system, since that then requires knocking down a majority of the nodes. Consensus-based control distributes security responsibilities among the nodes in the blockchain network, preventing hackers from tampering with the network and also protecting the IoT network from being destroyed by DDoS attacks. In this work, we show how edge-based blockchain technology applied to a Docker/Kubernetes-based containerized IoT architecture makes it secure and robust against attacks.

Decentralization also makes such a solution more scalable. Scalability is one of the biggest concerns in the deployment of network security systems on an evergrowing network, with more and more nodes being connected in real-time. In our proposed edge-based IoT security architecture, notification about the addition of new device nodes, the deletion of existing ones, and functional changes in any node goes out to all IoT nodes in real-time, through blockchain smart contracts. This reactive approach of our proposed edge-based blockchain system in our IoT architecture allows the system to be adaptable and flexible enough to expand and evolve without upgrading the entire platform.

Besides security and scalability, privacy is also enhanced in our proposed edge computing IoT architecture, through the containerized technology, where privacy across applications is strictly enforced. For example, access to the image data captured by a camera is only provided to the containers that have been authenticated and authorized. Access to object features extracted by the application in a container is restricted within the container. All access to ledger information in the distributed blockchain architecture is logged within the ledger; thus, privacy is ensured through security.

Scalability and Interoperability

Due to the heavy computations and communications required, there are often performance bottlenecks in a system that uses blockchain technology. With increasing network size, this issue becomes even worse. As a result, the efficiency of the consensus algorithm and the speed of transaction response both decrease. Therefore, blockchain transaction records and related data information require some kind of management, which can cope with better scalability.

Interoperability is the ability to connect and transfer information across different blockchain systems, for the smooth sharing of information. The performance of block- chain technology is reduced by the overhead of interacting with different blockchain networks, for example, a Bitcoin system with an Etherum system.

Limitation with Storage Facility

In a blockchain-based IoT system, the ledger records of the blockchain that are stored on each node increase in size, as time passes. Due to the limited storage capacities in IoT nodes, the required storage size will exceed the given capacities of memory storage in the IoT nodes.

Data Privacy and Confidentiality

Since an IoT is often an integration of various devices, services, and networks, the data stored or transferred on a device is vulnerable to privacy violation by compromising nodes in an IoT network. An attacker can attempt to impact data integrity by modifying the stored data or injecting false data into it for malicious purposes.

Authentication and Authorization

All devices in an IoT network must be authenticated for privileged access to services. Diversity and heterogeneity in the underlying architectures and environments supporting IoT devices lead to the existence of different authentication mechanisms for the IoT. As a result, there is no global standard protocol for authentication in the IoT.