Trust in Edge-Enabled IoT Security: Features, Challenges and Research Directions
Esin Ece Aydın, Şerif Bahtiyar, Gürkan Gür
cs.CR · cs.AI · cs.CY · cs.ET · cs.NI
Quick verdict
Survey of 55 studies builds a lifecycle×domain taxonomy of edge-IoT trust management; timely and structured, but search bias and ad-hoc ratings limit rigor.
Summary
This systematic review examines trust management in edge-enabled IoT through a PRISMA-based search of Web of Science and Scopus (2020–2026), yielding 55 included studies from 237 screened records. The authors organize the literature along a three-layer (physical/network/application) by four-domain (consumer/commercial/industrial/infrastructure) taxonomy and a five-dimension trust-design framework separating trust target (device/data/service), evidence acquisition, computation, update mechanism, and architecture. They claim to be the first survey covering all three trust targets jointly, supported by a comparison against 18 prior surveys (Table 1). Sixteen trust-related attacks are classified by target and layer, then mapped onto the trust-management lifecycle with subjective detection-difficulty ratings (Tables 3–4); FDI emerges as the most-mitigated attack, followed by BMA, GMA, SA, and OOA (Table 12). Twenty-one research challenges are quantified by mention frequency, led by trust poisoning (C12, 34 mentions), context-aware trust (C9, 29), and AI/FL integration (C10, 27); future directions include privacy-preserving trust computation, 6G sustainability, agent-to-agent trust, and digital-twin/GenAI evidence. Main limitations: the search mandates AI/ML keywords, biasing the corpus; detection-difficulty ratings lack a stated rubric; and the synthesis is purely qualitative, with no cross-study performance or overhead comparison.
Review complete
Assessment at a glance
4
risk flags identified
● high 0
● medium 2
● low 2
5 strengths · 5 weaknesses · 5 author questions
AI text signal = Binoculars zero-shot detector score over 20 sampled passages (Fast-DetectGPT co-reported) — absence does not prove human authorship, and a flag does not imply misconduct.
Levels reflect verification complexity — not a quality score. Evolve never rates your paper.
Key claims under review
- No prior survey covers device, data, and service trust simultaneously; this survey's taxonomy is claimed as the first to do so across four IoT domains and three architectural layers (Table 1).
- Trust-related attacks compromise the entire trust-management process — evidence flow, trust flow, and trust-driven decisions — not merely trust computation (Tables 3–4).
- FDI is the most frequently mitigated attack in the surveyed frameworks, followed by BMA, GMA, SA, and OOA (Table 12, Section 4.2).
- Research focus is shifting from trust-model design toward deployment and adversarial concerns, with trust poisoning (C12, 34 mentions) the most-cited challenge (Figures 8–9).
- Trust lacks a widely accepted definition and fragmented trust targets produce inconsistent evaluation, motivating a unified lifecycle treatment (Section 1).
Strengths
Corrects a real taxonomy flaw
Section 1.1 shows [99]'s classification conflates evidence source with computation mechanism; the five-dimension design taxonomy (Figure 5) treats them as orthogonal, enabling cleaner comparison across the domain tables (Tables 5–9).
Process-level attack mapping
Tables 3–4 classify 16 attacks by trust target and architectural layer — explicitly allowing hybrid assignments, improving on [57]'s single-layer mapping — and map each attack to affected lifecycle stages, origin, attacker behavior, and detection difficulty.
Quantified challenge landscape
Transparent selection counts
Forward-looking coverage
Weaknesses
Mandatory AI clause biases the corpus
Rubric-less subjective ratings
Non-auditable percentages
No quantitative cross-study synthesis
Presentation lapses
Full Review Report
$1.99 · one-timeUnlock the complete preliminary analysis Evolve produced for this paper.
Report preview
Scope & Claims
This 34-page survey (arXiv:2609.24669, formatted as an ACM manuscript) reviews trust management in edge-enabled IoT around three research questions: characterization across domains/layers/trust-targets (RQ1), how trust scores drive operational decisions and which attacks target trust (RQ2), and open gaps and directions (RQ3). Its central positioning claim — that no prior survey covers device, data, and service trust jointly — is argued through Table 1, comparing 18 surveys (2019–2026) on five axes. A secondary methodological claim is that prior taxonomies (notably [99]) conflated trust-evidence source with trust-computation mechanism; this survey separates them.
The remaining sections cover Methodology Assessment, Statistical Evidence, Results & Reproducibility, Limitations & Risks, and Verdict & Recommendations.
- All 5 strengths & 5 weaknesses, analyzed in detail
- 4 risk flags with severity, section & reviewer questions
- Methods & statistical evidence assessment
- 5 questions a peer reviewer would ask
- Full six-dimension evaluation report & recommendations
One-time purchase · this paper only
One-time purchase, this paper only. AI-generated preliminary analysis — not peer review.
Frequently asked questions
What is this paper about?
It is a PRISMA-based systematic review of 55 studies (2020–2026, from Web of Science and Scopus) on trust management in edge-enabled IoT. It organizes the field with a three-layer × four-domain taxonomy and a five-dimension trust-design framework covering device, data, and service trust, maps 16 trust-related attacks onto the trust-management lifecycle, and ranks 21 research challenges with future directions.
How is it different from earlier IoT trust surveys?
The authors compare 18 prior surveys (2019–2026) and claim to be the first covering all three trust targets — device, data, and service — simultaneously. It also adds process-level attack mapping across lifecycle stages with hybrid layer assignments, and quantifies challenge frequency (e.g., trust poisoning is the most-cited challenge at 34 mentions).
What are the top research gaps identified?
Defending the trust pipeline itself (trust poisoning, C12, 34 mentions), context-aware multi-dimensional trust (C9, 29), AI/federated-learning integration (C10, 27), resource-constrained operation (C16, 27), cold-start and sparse-history trust models (C7, 26), and reproducible evaluation methodology — which the authors flag as rarely addressed (C15, only 4 mentions).
Does the paper provide code or data?
No. It publishes its database search queries and PRISMA flow counts (256 identified, 237 screened, 55 included), but no extraction dataset, analysis code, or benchmark; the synthesis is qualitative.
About this review
This is an AI-generated preliminary review by Evolve.Pub, produced from the openly available text of the paper. It is not peer review and does not represent the authors' views. The canonical record of the paper is at the publisher link above.
- Extracted text is truncated: the consumer-IoT taxonomy (Table 5), portions of Sections 4.1 and 5.2, and the conclusion (Section 6) are missing, so Consumer-IoT domain findings could not be independently verified.
- Figures 1–9 are only partially recoverable from embedded text; Figure 7 bar values and Figure 9(a) per-year counts were reconstructed from text labels and may be incomplete.
- Table 1's per-survey checkmarks are garbled in extraction; assessment of the 18-survey comparison relies mainly on the surrounding prose.
- Placeholder ACM DOI (10.1145/nnnnnnn.nnnnnnn) and 'Manuscript submitted to ACM' indicate a preprint; a published version may differ.
Go deeper on this paper
Wrote a paper? Stress-test it before reviewers do.
Upload your manuscript for the same structured scrutiny — reasoning, statistics, evidence, figures, and reproducibility.