[Paper Review] PROTECT: Proximity-based Trust-advisor using Encounters for Mobile Societies

Udayan Kumar, Gautam Thakur+1 more|arXiv (Cornell University)|Apr 25, 2010

Opportunistic and Delay-Tolerant Networks8 references4 citations

TL;DR

This paper proposes PROTECT, a decentralized trust-advisory framework for mobile and delay-tolerant networks (DTNs) that infers trust from encounter frequency, duration, location-based behavior vectors, and behavior matrices. Using real-world campus traces, it demonstrates that trust filters remain stable over 9 weeks and significantly reduce network unreachability caused by selfish nodes—cutting it from 14.6% to 12.2% under high selfishness, proving trust can restore connectivity without degrading performance.

ABSTRACT

Many interactions between network users rely on trust, which is becoming particularly important given the security breaches in the Internet today. These problems are further exacerbated by the dynamics in wireless mobile networks. In this paper we address the issue of trust advisory and establishment in mobile networks, with application to ad hoc networks, including DTNs. We utilize encounters in mobile societies in novel ways, noticing that mobility provides opportunities to build proximity, location and similarity based trust. Four new trust advisor filters are introduced - including encounter frequency, duration, behavior vectors and behavior matrices - and evaluated over an extensive set of real-world traces collected from a major university. Two sets of statistical analyses are performed; the first examines the underlying encounter relationships in mobile societies, and the second evaluates DTN routing in mobile peer-to-peer networks using trust and selfishness models. We find that for the analyzed trace, trust filters are stable in terms of growth with time (3 filters have close to 90% overlap of users over a period of 9 weeks) and the results produced by different filters are noticeably different. In our analysis for trust and selfishness model, our trust filters largely undo the effect of selfishness on the unreachability in a network. Thus improving the connectivity in a network with selfish nodes. We hope that our initial promising results open the door for further research on proximity-based trust.

Motivation & Objective

• To address the challenge of trust establishment in dynamic, infrastructure-less mobile networks such as DTNs and ad hoc networks.
• To leverage mobility patterns—specifically encounters, location, and temporal behavior—to infer trust without centralized authority.
• To evaluate the stability and performance of multiple trust filters in real-world mobile social networks.
• To analyze how trust mitigates the negative effects of selfish node behavior on network connectivity in DTNs.
• To enable trust-aware routing and forwarding decisions in mobile peer-to-peer networks using decentralized, encounter-based metrics.

Proposed method

• The framework uses four trust advisor filters: encounter frequency (FE), encounter duration (DE), behavior vector (BV) based on count and duration of encounters, and behavior matrix (BM) based on location preferences.
• Each device maintains a local history of encounters, including timestamps and location data (e.g., via GPS or AP association), enabling fully distributed trust computation.
• Trust scores are computed based on spatio-temporal patterns, with BV and BM filters capturing behavioral similarity and location consistency across encounters.
• The system evaluates trust stability via overlap of recommended nodes over time (e.g., 90% overlap over 9 weeks) and compares filter performance using epidemic routing in DTN simulations.
• Selfishness is modeled by introducing node behavior where a fraction of nodes refuse to forward messages, and trust is used to counteract this effect.
• Performance is measured using metrics like unreachability, delay, and overhead in synthetic DTN routing scenarios based on real WLAN traces.

Experimental results

Research questions

• RQ1 How stable are trust recommendations derived from encounter-based metrics over time in real mobile societies?
• RQ2 To what extent do different trust filters (FE, DE, BV, BM) produce divergent or overlapping recommendations?
• RQ3 Can trust filters mitigate the impact of selfish node behavior on network connectivity in DTNs?
• RQ4 How does the performance of epidemic routing vary when using trust-based forwarding versus random or selfish selection?
• RQ5 Does trust have a greater impact in networks with higher levels of selfishness?

Key findings

• Three trust filters—DE, FE, and BV-Count—produced stable trust lists with over 90% overlap in recommended nodes over a 9-week period, indicating long-term reliability.
• Despite differences in recommended nodes, all filters performed similarly in epidemic routing, suggesting robustness and flexibility in network performance.
• The FE filter reduced unreachability in a selfish network (S=0.8, T=40%) from 14.5975% to 12.1892%, approaching the 10.8998% level of a non-selfish network.
• Trust had the greatest impact on reducing unreachability when selfishness was high (S>0.4), with the most significant improvement occurring between T=0% and T=20%.
• The delay and overhead metrics followed similar trends, confirming that trust-based forwarding maintains efficient routing even under selfishness.
• The behavior-matrix (BM) and behavior-vector (BV-D) filters performed slightly worse than DE and FE, but still reduced unreachability by over 15% compared to no trust.