Trusted Routing Based on Dynamic Trust Mechanism in Mobile Ad-Hoc Networks

Sancheng Peng, Weijia Jia, Guojun Wang, Jie Wu, Minyi GuoIEICE Trans. Information and Systems 2010 · DOI

Introduction

Mobile Ad-Hoc Networks (MANETs) are decentralized, self-organizing systems that rely on node cooperation for data routing. This inherent openness makes them vulnerable to various security threats, such as black hole and Denial of Service (DoS) attacks. Traditional security measures like encryption are often impractical due to the limited computational resources of MANET nodes and the absence of a central authority. This paper proposes DTM-DSR, a secure routing protocol that replaces rigid security with a dynamic trust mechanism inspired by human social behaviors.

The Dynamic Trust Mechanism (DTM)

The core of the proposed system is a Dynamic Trust Mechanism that evaluates node trustworthiness in real-time. It addresses two critical social phenomena: trust decay over time and the risk of dishonest recommendations.

Trust Updating Algorithm

Direct trust between nodes is governed by a "slow rise, fast drop" principle. It incorporates three primary factors:

  1. Time Aging: Trust values naturally decrease if no interactions occur over a period.
  2. Reward Factor: Successful interactions lead to a gradual increase in trust.
  3. Penalty Factor: Failed interactions or malicious behavior result in a rapid, significant drop in trust.

Collaborative Filtering (CF)

To mitigate the impact of malicious nodes providing false ratings (collusion or slandering), the system employs Collaborative Filtering. A node validates a recommendation by comparing the recommender's historical rating patterns with its own. Recommendations from nodes with similar "trust preferences" are weighted more heavily, effectively filtering out dishonest feedback.

DTM-DSR Protocol Implementation

The DTM is integrated into the Dynamic Source Routing (DSR) protocol. Nodes maintain neighbor tables with trust values and use these to populate Route Request (RREQ) and Route Reply (RREP) packets.

Figure 2: Route discovery and maintenance using path trust values ($T_{route}$).

As shown in Figure 2, nodes calculate an accumulated route trust ($T_{route}$). The protocol prioritizes paths with the highest $T_{route}$ and maintains a "BlackList" for neighbors whose trust falls below a critical threshold ($T_{low}$), ensuring that potentially malicious nodes are bypassed during route selection.

Performance Evaluation

The protocol was evaluated using the GloMoSim 2.03 simulation environment, modeling a 50-node MANET with 5 malicious nodes that randomly dropped 10% to 30% of packets. DTM-DSR was compared against standard DSR and a Bayesian-based trust model (Bayesian-DSR).

Figure 4: Throughput comparison between DTM-DSR, Bayesian-DSR, and standard DSR across different pause times.

Key Results

  • Throughput: DTM-DSR consistently achieved higher throughput (approximately 32 packets/s) compared to Bayesian-DSR and standard DSR (Figure 4). This is attributed to its ability to accurately identify and avoid packet-dropping nodes.
  • Packet Loss Ratio: The protocol demonstrated a significantly lower packet loss ratio. By proactively avoiding untrustworthy nodes, it minimizes the need for costly route rediscoveries triggered by packet drops.
  • End-to-End Delay: DTM-DSR reduced average delay by selecting reliable paths from the outset, avoiding the timeouts associated with malicious node activity in standard DSR.

Conclusion

DTM-DSR provides a robust framework for secure routing in MANETs by leveraging socially-inspired trust dynamics and collaborative filtering. While the protocol effectively counters basic packet-dropping attacks and dishonest recommendations, future work will focus on optimizing computational overhead and defending against more sophisticated threats like Sybil attacks.

Summary updated

BibTeX
@article{peng2010trusted,
  author = {Sancheng Peng and Weijia Jia and Guojun Wang and Jie Wu and Minyi Guo},
  title = {{Trusted Routing Based on Dynamic Trust Mechanism in Mobile Ad-Hoc Networks}},
  journal = {IEICE Trans. Information and Systems},
  year = {2010},
  doi = {10.1587/transinf.e93.d.510},
}
Figure