Named Data Networking for Efficient IoT-based Disaster Management in a Smart Campus

Zain Ali, Munam Ali Shah, Ahmad Almogren, Ikram Ud Din, Carsten Maple, Hasan Ali KhattakSustainability 2020 · DOI

TL;DR

This paper proposes NDN-DM, a novel Named Data Networking (NDN) based system for efficient IoT-based disaster management in smart campuses. By suspending the Pending Interest Table (PIT) and using a PUSH mechanism, NDN-DM significantly reduces communication delay and energy consumption during emergencies like fires, outperforming existing NDN-based solutions.

What Problem Does This Paper Address?

Traditional Internet Protocol (IP)-based communication systems struggle with the demands of Internet of Things (IoT) devices, especially in critical situations like disasters. IP's location-dependent nature leads to inefficient bandwidth usage, high processing demands, security vulnerabilities, and excessive memory consumption, which are particularly problematic for resource-constrained IoT devices. While Information-Centric Networking (ICN) and specifically Named Data Networking (NDN) offer a more suitable paradigm by focusing on content rather than location, existing NDN-based Disaster Management Systems (DMSs) often rely on a PULL-based approach. This PULL mechanism, where consumers request data, introduces significant delays, making it unsuitable for time-critical disaster scenarios like fires where rapid dissemination of alerts is crucial. The paper aims to bridge this gap by proposing an efficient, low-delay, and energy-conscious NDN-based DMS for smart campuses.

Methods

The core of the proposed solution, NDN-DM, is built upon the NDN architecture but introduces key modifications to address the limitations of existing PULL-based systems in disaster scenarios.

  • PUSH-based Communication: Instead of waiting for consumer requests (PULL), NDN-DM utilizes a PUSH mechanism. In case of a fire, a producer (e.g., a fire sensor) directly broadcasts an alert message (AM) to neighboring nodes. This ensures immediate dissemination of critical information.
  • Suspension of Pending Interest Table (PIT): A significant innovation is the temporary suspension of the PIT. The PIT is typically used to track pending interests and manage responses. However, in a disaster scenario, the immediate need is to broadcast alerts, not to manage complex interest-response cycles. By suspending the PIT, the system avoids the extra overhead and delay associated with its operation, leading to faster alert delivery and reduced energy consumption.
  • Customized Packet Headers and Table Headers: The paper defines a simplified packet header for alert messages, including a fixed sequence number ('0') to indicate a disaster and a delay field. The Content Store (CS) and Forwarding Information Base (FIB) headers are also customized for efficient disaster management.
  • Mathematical Modeling and Formal Verification: The proposed NDN-DM system is mathematically modeled using timed automata-based transition systems. This model is then formally verified using a real-time model checker to prove that the maximum delay in informing all nodes is within a defined threshold.

Datasets

Not discussed in the paper. The evaluation was performed using simulations.

Key Insights and Ideas

  • PIT Suspension is Key for Disaster Scenarios: The most impactful insight is that for time-critical disaster alerts, the overhead of the PIT in NDN can be detrimental. Suspending it significantly improves performance by reducing delay and energy consumption.
  • PUSH is Superior for Emergency Alerts: Directly pushing alert messages from producers to consumers is far more effective than waiting for requests in disaster situations, ensuring rapid notification.
  • NDN's Content-Centric Nature is Advantageous: NDN's ability to address content directly, rather than relying on IP addresses, simplifies communication and enhances security, making it a strong foundation for DMS.
  • Smart Campus as a Realistic Testbed: Using a smart campus scenario with mobile nodes provides a practical and relevant environment to test the system's effectiveness in a common IoT deployment.

Key Results

The paper presents simulation results using NS2, comparing NDN-DM with a state-of-the-art NDN-based DMS (NDN-DISCA).

  • Reduced End-to-End Delay: NDN-DM achieved a reduction in end-to-end delay of 2% to 10% compared to NDN-DISCA.
  • Minimized Energy Consumption: The proposed scheme minimized energy consumption by up to 20%, with energy improvement ranging from 3% to 20% compared to NDN-DISCA.
  • Improved Performance with Increased Nodes: The benefits of NDN-DM become more pronounced as the number of nodes in the network increases, demonstrating its scalability in handling larger disaster events.

Limitations and Future Work

The authors acknowledge two main limitations:

  • Memory Utilization: While NDN-DM is efficient, smart devices have limited memory. Further improvements in memory utilization are needed.
  • Network Complexity with More Nodes: As the number of nodes increases, the network complexity also grows, which could pose challenges for future scalability.

Future work suggested includes extending the NDN-DM scheme to other disaster scenarios beyond fires, such as smart hospitals, smart homes, and smart cities, and further optimizing memory utilization.

Summary updated

Abstract

Disasters are uncertain occasions that can impose a drastic impact on human life and building infrastructures. Information and Communication Technology (ICT) plays a vital role in coping with such situations by enabling and integrating multiple technological resources to develop Disaster Management Systems (DMSs). In this context, a majority of the existing DMSs use networking architectures based upon the Internet Protocol (IP) focusing on location-dependent communications. However, IP-based communications face the limitations of inefficient bandwidth utilization, high processing, data security, and excessive memory intake. To address these issues, Named Data Networking (NDN) has emerged as a promising communication paradigm, which is based on the Information-Centric Networking (ICN) architecture. An NDN is among the self-organizing communication networks that reduces the complexity of networking systems in addition to provide content security. Given this, many NDN-based DMSs have been proposed. The problem with the existing NDN-based DMS is that they use a PULL-based mechanism that ultimately results in higher delay and more energy consumption. In order to cater for time-critical scenarios, emergence-driven network engineering communication and computation models are required. In this paper, a novel DMS is proposed, i.e., Named Data Networking Disaster Management (NDN-DM), where a producer forwards a fire alert message to neighbouring consumers. This makes the nodes converge according to the disaster situation in a more efficient and secure way. Furthermore, we consider a fire scenario in a university campus and mobile nodes in the campus collaborate with each other to manage the fire situation. The proposed framework has been mathematically modeled and formally proved using timed automata-based transition systems and a real-time model checker, respectively. Additionally, the evaluation of the proposed NDM-DM has been performed using NS2. The results prove that the proposed scheme has reduced the end-to-end delay up from 2 % to 10 % and minimized up to 20 % energy consumption, as energy improved from 3 % to 20 % compared with a state-of-the-art NDN-based DMS.
BibTeX
@inproceedings{ali2020named,
  author = {Zain Ali and Munam Ali Shah and Ahmad Almogren and Ikram Ud Din and Carsten Maple and Hasan Ali Khattak},
  title = {{Named Data Networking for Efficient IoT-based Disaster Management in a Smart Campus}},
  booktitle = {Sustainability},
  year = {2020},
  doi = {10.3390/su12083088},
}

Loading related content...

Figure