Geng integrity report: Verifiable attribute-based encryption scheme with outsourced decryption for fog computing (DOI: 10.3969/j.issn.1007-130X.2023.03.009)

Academic fraud report · Geng Detector

Summary

This Geng (耿同学) integrity report flags the paper "Verifiable attribute-based encryption scheme with outsourced decryption for fog computing" (Duan Yahong, Wang Zheng, Zhao Juanjuan, Wang Long; Journal of Computer Engineering & Science, 2023) as highly suspicious (🟠). Five issues are raised: (1) Algorithm 5, the smart-contract access-verification procedure, appears mathematically incomplete — the Proof input contains only hashCode, hashCode', and Address, with no access to the user's attribute set S or the policy matrix M needed to construct constants c_x and read ciphertext components C'_x; (2) the system flow contradicts ABE's offline-access-control purpose by requiring the CA to generate the attribute private key only after on-chain verification, effectively degenerating ABE into an online identity-based scheme; (3) any on-chain computation over S would leak user attributes, conflicting with the claimed privacy-preservation goal; (4) Section 5.2 presents experimental figures without numeric data, and the prototype cannot run if Algorithm 5 is broken, casting doubt on the reported measurements; (5) OCR artifacts and typographical errors (e.g., "Averifiable", "fogcomputing") suggest weak proofreading. Confidence is moderate; findings derive from textual and logical analysis only and have not been independently verified by running code or contacting the authors.

Verdict

🟠 Highly suspicious. Multiple logical inconsistencies in the smart-contract design, a self-contradictory system flow, and unverifiable experimental claims. No evidence of outright data fabrication is presented, but the core algorithm appears structurally broken.

Key findings

Algorithm 5 is logically incomplete: Proof={hashCode, hashCode', Address} lacks S and M, making computation of the constant set {c_x} and reading of {C'_x} from the contract impossible.
Smart contracts on-chain normally store only hashes and addresses; full ciphertext CT retrieval for verification is not justified (gas/storage costs ignored).
System flow forces the CA to issue attribute private keys only after on-chain verification, contradicting ABE's offline-access-control property and the stated goal of "reducing CA burden".
On-chain use of S for verification would expose user attributes, contradicting the paper's privacy-protection claim.
Section 5.2 provides Figures 3–5 but no numeric data; the prototype cannot function correctly if Algorithm 5 is broken, undermining the 10-run experiment.
Typographical/OCR issues ("Averifiable", "fogcomputing", formula garbling) suggest weak proofreading.

Evidence highlights

DOI: 10.3969/j.issn.1007-130X.2023.03.009; Journal of Computer Engineering & Science, 2023.
Authors listed: 段亚红, 王峥, 赵涓涓, 王龙.
Algorithm 5 input set Proof={hashCode, hashCode', Address} — does not contain S or M.
Section 3.2 places key generation as stage 4, "after ciphertext correctness is confirmed"; Section 3.3 (3) ⑤ states "CA generates the attribute private key when Proof is verified True".
Experimental hardware mentioned: AMD Ryzen 5 2500U; 10 system runs reported but no tabular values supplied.
OCR artifacts include "Averifiable", "fogcomputing", and garbled exponent in K = g^{α+kα b+ka} h^{bc}.

Notes

Findings are based on logical analysis of the published text and Algorithm 5 specification; no source code, prototype logs, or author correspondence were examined.
The five issues correspond to Geng's categorizations: #1, #3 → 第六式 (logical gaps); #2 → 第六式 (flow contradiction); #4 → 第二式 (data/reproducibility); #5 → 第一/第三式 (presentation).
The report explicitly notes it is AI-assisted and intended for academic discussion; final misconduct determination requires institutional investigation.
Recommended follow-up: post on PubPeer, request runnable source code from the authors, report to the journal editorial office.