Connected-vehicle and roadside telemetry enable low-latency safety navigation, and traffic-optimisation services at the edge, but finegrained mobility streams (locations, speeds, events, and contexts) create high re-identification and linkage risk when accessed by multiple stakeholder domains. We present a Zero Touch Network and Service Management (ZSM) integrated, policy-driven data-collection service that operationalises mobility data governance through intent-based automation. Stakeholders submit high-level collection intents (purpose, fields, spatial/temporal granularity, latency, and utility targets); a policy engine evaluates and rewrites intents into compliant, effective intents; and a plan generator compiles them into executable data-collection pipelines deployable within a ZSM closed loop. Experiments on Beijing taxi mobility traces execute 87,500 DP-protected releases and achieve 1.26\% relative error for Road Safety Authority (RSA) at ϵ = 8.0, while DP-Stochastic Gradient Descent (DP-SGD) risk scoring reaches 0.97 ± 0.03 test accuracy at ϵ = 0.5,
Kumar, V, Yadav, AK, Pnadey, PK, Misra, M, Liyanage, M & Braeken, A 2026, 'Zero-Touch Mobility Data Governance with Differential Privacy in ZSM-Based Vehicular Edge Services'. https://doi.org/10.1145/3779208.3804881
Kumar, V., Yadav, A. K., Pnadey, P. K., Misra, M., Liyanage, M., & Braeken, A. (2026). Zero-Touch Mobility Data Governance with Differential Privacy in ZSM-Based Vehicular Edge Services. https://doi.org/10.1145/3779208.3804881
@conference{0e8a96f970f54fbdba77999d9268f04e,
title = "Zero-Touch Mobility Data Governance with Differential Privacy in ZSM-Based Vehicular Edge Services",
abstract = "Connected-vehicle and roadside telemetry enable low-latency safety navigation, and traffic-optimisation services at the edge, but finegrained mobility streams (locations, speeds, events, and contexts) create high re-identification and linkage risk when accessed by multiple stakeholder domains. We present a Zero Touch Network and Service Management (ZSM) integrated, policy-driven data-collection service that operationalises mobility data governance through intent-based automation. Stakeholders submit high-level collection intents (purpose, fields, spatial/temporal granularity, latency, and utility targets); a policy engine evaluates and rewrites intents into compliant, effective intents; and a plan generator compiles them into executable data-collection pipelines deployable within a ZSM closed loop. Experiments on Beijing taxi mobility traces execute 87,500 DP-protected releases and achieve 1.26\% relative error for Road Safety Authority (RSA) at ϵ = 8.0, while DP-Stochastic Gradient Descent (DP-SGD) risk scoring reaches 0.97 ± 0.03 test accuracy at ϵ = 0.5,",
author = "Vishal Kumar and Yadav, \{Awaneesh Kumar\} and Pnadey, \{Pradumn Kumar\} and Manoj Misra and Madhusanka Liyanage and An Braeken",
note = "Publisher Copyright: {\textcopyright} 2026 Copyright held by the owner/author(s).",
year = "2026",
month = jun,
day = "4",
doi = "10.1145/3779208.3804881",
language = "English",
}