Blockchain applications and limitations in Cyber-Physical Systems – CPS

Cyber-Physical Systems (CPS) integrate sensing, networking, and computation capabilities where millions of networkable devices sense the environment to capture information. Using network infrastructure, mainly the Internet, the information is transferred to cloud servers where it is processed to offer personalized services to the users. The data processing outcome is fed into decision making algorithms that may trigger particular actions by the devices.

The potential of CPS is immense across many economic sectors.

For this reason, CPS underpins the fourth industrial revolution that promises transformation of industries and consumer markets through sensing, robotics, artificial intelligence, and other emerging technologies. Realising the many opportunities offered by CPS involves unique challenges including centralization, security, privacy, heterogeneity in devices, lack of control or auditability over or data, persistence/ sustainability, and trust.

In recent years, blockchain, a distributed ledger of chained blocks, has attracted significant attention as a means to address the aforementioned challenges in CPS due to its salient features including security, anonymity, decentralization, trust, and immutability. The basic communication primitive between the blockchain participants is known as a transaction. Blockchains are managed in a distributed manner, where all transactions and blocks are verified by the participating nodes. Particular nodes in the network, known as miners or validators, may choose to append new blocks to the blockchain in return for some incentives, normally in the form of tokens or coins. Appending a new block involves following a consensus algorithm, whereby all validators must agree on the addition of the new block, which introduces high security and trust.

Blockchain CPS adoption challenges:

Scalability

While blockchain holds significant potential for addressing the challenges of CPS its adoption for CPS involves several challenges including: 1) scalability: The pure distributed nature of the blockchain limits its scalability for large scale networks, such as CPS, as the blockchain packet overhead increases quadratically with the number of nodes in the network..

Delay

2) delay: storing a transaction in the blockchain involves following a consensus algorithm which in turn increases the delay associated with storing a transaction in the blockchain. Additionally, the users must wait for a confirmation time that protects the users against double spending attack, ..

Computational resource consumption

3) computational resource consumption: Blockchain removes the centralization by requiring all the participating nodes to participate in verifying all new transactions and blocks. On the flip side, this potentially increases the processing overhead on the participants, particularly in large scale networks with millions of transactions,

Memory Overhead

4) Memory overhead: modification or removal of the previously stored data is impossible which makes blockchain immutable and auditable. However, this potentially increases the size of the blockchain database in time, e.g., the Bitcoin blockchain database size is currently 285 GB,

Throughput

5) Throughput: Throughput is defined as the total number of transactions that can be stored in blockchain per second. The blockchain throughput is affected by the underlying consensus algorithm, e.g., Bitcoin throughput is 7 transaction per second. However, with millions of nodes, CPS demands high throughput which is far beyond the throughput in existing blockchain instantiations,

Privacy

6) Privacy: The immutability of the blockchain makes it impossible for the users to remove the history of their transactions from the blockchain. While the data (if any) that is stored in transactions is encrypted, its persistence indefinitely creates long-term privacy risks for the data owners. Malicious nodes may attempt to deanonymize a user by analyzing the history of the transactions in the blockchain which in turn may compromise the user privacy, and

Reliance on trusted third parties

7) reliance on trusted third parties: In non-monetary applications of blockchain, the ownership of the asset is transferred through the blockchain and the physical asset is transferred off-the-chain. To ensure that both sides of a trade commit to their obligations, most of the existing works rely on a Trusted Third Party (TTP), e.g., smart city manager or energy companies. However, reliance on a TTP raises challenges pertaining to centralization such as single point of failure and lack of privacy as discussed earlier.

The book Blockchain for Cyberphysical Systems by Ali Dorri, Salil Kanhere, Raja Jurdak studies these fundamental challenges involved in adopting blockchain for CPS and explores the existing solutions.

In addition to the outlined challenges, applying blockchain in various application domains involves application-specific challenges. Thus, we explore how to apply blockchain in various application domains including smart grid, smart vehicles, supply chains, and IoT data market place. For each application, we study the existing frameworks in the literature, highlight their limitation, and outline the future research directions.

Authors Bio:

Ali Dorri is a Research Fellow at Queensland University of Technology. He has served as reviewer for over 35 conferences and journals.

Salil Kanhere is a professor at The University of New South Wales. He received his Ph.D. in electrical engineering from Drexel University.

Raja Jurdak is Professor and Chair at Queensland University of Technology. He received his Ph.D in information and computer science from the University of California, Irvine.