Decentralized Identity in Practice: Benchmarking Latency, Cost, and Privacy

Decentralized Identifiers (DIDs) are increasingly deployed on distributed ledgers, yet systematic cross-platform evidence on their operational behavior remains limited. We present an empirical benchmarking study of three prominent ledger-based DID methods - Ethereum, Hedera, and XRP Ledger - using reference Software Development Kits (SDKs) under a unified experimental setup. We measure latency, transaction cost, and on-chain metadata exposure, normalizing latency by each platform’s block or consensus interval and cost by its native value transfer fee. Privacy leakage is quantified using a Metadata-Leakage Score (MLS), an entropy-based measure expressed in bits per operation. Our results reveal distinct architectural trade-offs. Ethereum enables near-instant, off-chain DID creation, but incurs the highest latency and cost for on-chain lifecycle operations. XRPL delivers deterministic and stable latency with fixed, low fees, yet exhibits higher metadata leakage due to more verbose transaction payloads. Hedera achieves the lowest on-chain latency and low fees with minimal metadata leakage, while occasional variance arises from SDK-side processing and confirmation pipelines. Overall, the findings show that ledger architecture and SDK workflows play a major role in shaping DID latency, cost, and metadata exposure, complementing the effects of the underlying consensus mechanism. These results provide evidence-based insights to support informed selection and configuration of DID systems under performance and privacy constraints.

💡 Research Summary

This paper presents the first systematic, cross‑platform benchmarking of decentralized identifier (DID) methods that are implemented on public distributed ledgers. The authors focus on three widely‑deployed DID methods—did:ethr on Ethereum, did:hedera on Hedera Hashgraph, and did:xrpl on the XRP Ledger—because each represents a distinct architectural pattern: contract‑based, event‑stream‑based, and ledger‑native respectively. Using the official JavaScript SDKs for each platform, they construct a unified experimental framework that isolates the DID‑specific overhead from the underlying consensus mechanisms.

The study is driven by three research questions: (1) What are the absolute and relative latency, cost, and privacy (metadata leakage) characteristics of core DID operations under standardized conditions? (2) Which design choices in the ledger architecture and SDK workflow explain the observed performance and privacy outcomes? (3) How do the evaluated platforms compare when balancing responsiveness, economic efficiency, and privacy for practical deployments?

Methodology
The authors measure five core DID lifecycle operations—Create, Resolve, Update, Revoke, and Delete—across 100 independent runs per operation per platform. Latency is recorded from transaction submission to final consensus confirmation. To enable fair comparison across heterogeneous ledgers, they introduce two normalized metrics: relative latency (ω_rel), defined as the operation latency divided by the platform’s average block or consensus interval, and relative cost (ν_rel), defined as the operation fee divided by the fee of a standard token transfer on the same network. In addition, they propose a Metadata‑Leakage Score (MLS), an entropy‑based metric that quantifies the amount of identifying information exposed in on‑chain transaction payloads. MLS is expressed in bits per operation; higher values indicate greater fingerprintability. The Create operation is excluded from MLS because did:ethr creation is entirely off‑chain and therefore yields zero on‑chain metadata.

Results

• Latency: Hedera exhibits the lowest relative latency (ω_rel ≈ 0.5), reflecting its fast (~2 s) consensus and event‑stream design. XRPL follows with ω_rel ≈ 0.8, benefiting from a fixed 4‑second ledger close interval. Ethereum shows the highest latency (ω_rel ≈ 2.3) for on‑chain updates and deletions, driven by its ≈13‑second block time and the overhead of smart‑contract execution.
• Cost: Relative cost mirrors the latency pattern. Hedera and XRPL incur modest fees (ν_rel ≈ 0.2–0.3), while Ethereum’s gas costs are substantially higher (ν_rel ≈ 5.8), often amounting to tens of dollars per operation on mainnet.
• Privacy (MLS): XRPL has the highest MLS (≈12.4 bits/op) because its transaction format includes verbose service endpoint and key fields, making each operation highly distinguishable. Hedera achieves the lowest MLS (≈4.1 bits/op) due to a compact payload that omits unnecessary identifiers. Ethereum’s MLS (≈6.7 bits/op) sits in the middle.
• SDK‑Level Variability: The authors note that SDK implementations contribute non‑trivial variance. Hedera’s SDK performs asynchronous signing and submission, reducing average latency but occasionally triggering retries under network congestion. XRPL’s SDK optimizes payload serialization but lacks native multi‑signature support, which can increase cost for certain security policies. Ethereum’s SDK includes complex gas‑limit estimation logic; mis‑estimation leads to transaction re‑submission and added latency.

Discussion
The findings demonstrate that ledger architecture, not merely the underlying consensus algorithm, dominates the trade‑offs among latency, cost, and privacy. Contract‑based Ethereum offers flexibility and ecosystem maturity but suffers from high gas fees, slower finality, and moderate metadata exposure. Event‑stream‑based Hedera delivers fast, cheap, and privacy‑preserving operations, making it attractive for large‑scale identity deployments where cost and privacy are paramount. Ledger‑native XRPL provides deterministic latency and low fees but at the expense of higher on‑chain fingerprintability.

From a privacy engineering perspective, the MLS metric highlights that reducing on‑chain payload entropy—e.g., storing only key hashes on‑chain and keeping service endpoints off‑chain (IPFS, secure APIs)—can substantially lower fingerprintability without affecting functional correctness. The authors suggest that SDKs should incorporate payload minimization and optional compression to improve both privacy and efficiency.

Practical Guidance

• For cost‑sensitive, high‑throughput identity services (e.g., national ID systems), Hedera appears most suitable.
• When integration with existing Ethereum‑based DeFi or NFT ecosystems is required, developers should consider layer‑2 solutions (Optimism, zkSync) to mitigate gas costs while preserving compatibility.
• XRPL is a good fit for financial‑oriented identity verification where deterministic latency is valuable, provided that additional encryption or off‑chain storage is employed to curb metadata leakage.

Conclusion
By providing a reproducible benchmarking framework and introducing normalized latency, cost, and entropy‑based privacy metrics, the paper offers a quantitative foundation for comparing DID implementations across heterogeneous ledgers. The results underscore that architectural choices at the ledger and SDK layers jointly shape the operational profile of decentralized identity systems. Future work should extend the methodology to layer‑2 scaling solutions, cross‑chain DID interoperability, and richer adversarial models that incorporate off‑chain side‑channels.