Slotted ALOHA Based Practical Byzantine Fault Tolerance (PBFT) Blockchain Networks: Performance Analysis and Optimization

Ziyi Zhou, Oluwakayode Onireti, Lei Zhang, Muhammad Ali ImranSensors 2024 · DOI

TL;DR

This paper introduces a novel framework for analyzing and optimizing the performance of Practical Byzantine Fault Tolerance (PBFT) consensus in wireless blockchain networks, specifically for IoT applications. By incorporating the challenges of wireless communication, like channel contention using Slotted ALOHA, the authors derive key performance indicators and define a "viable area" to ensure network reliability and efficiency.

What Problem Does This Paper Address?

The Internet of Things (IoT) is rapidly expanding, connecting billions of devices. While blockchain technology offers a promising solution for enhancing security, trust, and decentralization in IoT systems, traditional blockchain consensus mechanisms often struggle with the resource constraints and unique communication challenges of wireless IoT networks. Specifically, existing PBFT models typically assume ideal, wired communication, failing to account for issues like fluctuating channel quality, network congestion, and message collisions inherent in wireless environments. This paper tackles the problem of adapting and analyzing PBFT for these realistic wireless scenarios, aiming to ensure its effectiveness and efficiency in a decentralized IoT ecosystem.

Methods

The authors propose a framework that models wireless PBFT networks using a spatio-temporal Poisson Point Process (PPP) for node distribution and transaction arrival. A key innovation is the integration of Slotted ALOHA as the medium access control (MAC) protocol for wireless communication between PBFT nodes. This allows them to analyze the impact of channel contention on the PBFT consensus phases (pre-prepare, prepare, and commit).

They derive:

  • Success probabilities for each PBFT phase, considering collisions.
  • End-to-end success probability (Ps), which is crucial for transaction finality.
  • Key Performance Indicators (KPIs) derived from Ps, including:
  • Average transaction throughput.
  • Optimal transmission interval (v*).
  • Transaction confirmation delay.
  • Average transmit power.
  • Viable Area: A novel concept defining the minimum network coverage area (or number of nodes) required to guarantee PBFT's safety, liveness, and resilience for a given success probability and number of faulty nodes.

The paper also analyzes the impact of PBFT's view change mechanism on transaction confirmation delay in wireless settings.

Datasets

Not discussed in the paper. The evaluation relies on simulated and analytical results based on defined system parameters.

Key Insights and Ideas

  • Wireless Challenges Matter: The paper clearly demonstrates that ignoring wireless communication realities (like Slotted ALOHA collisions) significantly impacts PBFT performance and reliability.
  • Slotted ALOHA Integration: The novel integration of Slotted ALOHA into the PBFT consensus model provides a more realistic performance evaluation for wireless IoT networks.
  • The "Viable Area" Concept: This is a significant contribution, offering a practical metric to determine the minimum network resources (coverage, nodes) needed for a reliable PBFT operation, directly impacting energy efficiency and cost.
  • Trade-offs are Crucial: The research highlights inherent trade-offs. For instance, increasing the transmission interval can improve success probability by reducing collisions, but it also increases delay. Similarly, a larger viable area might mean more nodes, increasing potential collisions but also potentially improving resilience.
  • View Change Impact: The analysis of view change in wireless PBFT reveals a substantial increase in confirmation delay, a critical factor for real-time IoT applications.

Key Results

  • The derived analytical formulas for end-to-end success probability (Ps) show a tight match with simulation results, validating the model.
  • Increasing the number of faulty nodes ($f$) reduces end-to-end success probability, while increasing the transmission interval ($v$) improves it by mitigating collisions.
  • An optimal transmission interval ($v^*$) exists that maximizes transaction throughput. This optimal interval is influenced by the number of nodes ($n$) and faulty nodes ($f$).
  • The viable area analysis shows that a lower transmission interval is required for the network to be viable when the end-to-end success probability requirement or the number of faulty nodes is reduced.
  • The effective transaction confirmation delay increases significantly due to view changes in wireless PBFT, and this increase is amplified by a higher number of faulty nodes.

Limitations and Future Work

  • Static IoT Network Scenario: The current analysis assumes a static network. Future work should consider the performance of wireless PBFT networks in mobile IoT systems or environments with evolving mobility.
  • Focus on Collisions: The analysis primarily addresses the impact of collisions on system effectiveness. The paper suggests that investigating the effects of malicious attacks on the security of the system is a crucial next step.
  • Other Wireless Protocols: The authors recommend exploring the implementation and performance of other wireless protocols with PBFT.
  • Optimization Algorithm: While Algorithm 1 describes a simple exhaustive search for the viable area, more efficient optimization algorithms could be explored.

Summary updated

Abstract

Practical Byzantine Fault Tolerance (PBFT) is one of the most popular consensus mechanisms for the consortium and private blockchain technology. It has been recognized as a candidate consensus mechanism for the Internet of Things networks as it offers lower resource requirements and high performance when compared with other consensus mechanisms such as proof of work. In this paper, by considering the blockchain nodes are wirelessly connected, we model the network nodes distribution and transaction arrival rate as Poisson point process and we develop a framework for evaluating the performance of the wireless PBFT network. The framework utilizes slotted ALOHA as its multiple access technique. We derive the end-to-end success probability of the wireless PBFT network which serves as the basis for obtaining other key performance indicators namely, the optimal transmission interval, the transaction throughput and delay, and the viable area. The viable area represents the minimum PBFT coverage area that guarantees the liveness, safety, and resilience of the PBFT protocol while satisfying a predefined end-to-end success probability. Results show that the transmission interval required to make the wireless PBFT network viable can be reduced if either the end-to-end success probability requirement or the number of faulty nodes is lowered.
BibTeX
@inproceedings{zhou2024slotted,
  author = {Ziyi Zhou and Oluwakayode Onireti and Lei Zhang and Muhammad Ali Imran},
  title = {{Slotted ALOHA Based Practical Byzantine Fault Tolerance (PBFT) Blockchain Networks: Performance Analysis and Optimization}},
  booktitle = {Sensors},
  year = {2024},
  doi = {10.3390/s24237688},
}
Figure