pith. sign in

arxiv: 2605.09910 · v1 · submitted 2026-05-11 · 💻 cs.NI

CloudEmu: A Trace-Driven Cloud-Native Emulation Testbed for Vehicle Video Uplink over Cellular Networks

Takashi Torii , Soto Anno , Masaki Okada , Takuma Tsubaki , Nobuhiro Azuma , Takuya Tojo This is my paper

Pith reviewed 2026-05-12 04:28 UTC · model grok-4.3

classification 💻 cs.NI
keywords cloud emulationtrace-driven replayvehicle video uplinkcellular networksnetwork testbedautonomous vehiclesposition replayvideo communication
0
0 comments X

The pith

CloudEmu replays real vehicle cellular and position traces on virtual Linux nodes to enable repeatable video uplink tests under identical conditions.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents CloudEmu, a trace-driven cloud-native emulation testbed for vehicle video uplink communication over cellular networks. It targets the need for reliable low-latency video from autonomous vehicles by replaying time-synchronized traces collected once from real vehicles. The system runs these traces on commodity virtual nodes to couple traffic behavior with position along a route. This produces controlled experiments that avoid the cost of repeated physical trials while retaining measured network dynamics. Readers would care because it bridges the gap between expensive field tests and less faithful simulations for developing production video stacks.

Core claim

CloudEmu replays time-synchronized cellular and position traces collected from vehicles on commodity Linux-based virtual vehicle and video-receiver nodes. A Linux-based emulation framework couples traffic replay with position replay, tying network dynamics directly to each point along the route. This enables repeatable route-aware experiments without repeated on-road trials and supports running production-grade video-uplink stacks for low-cost controlled comparisons under identical replayed conditions.

What carries the argument

CloudEmu, the trace-driven cloud-native emulation framework that couples traffic replay with position replay on virtual Linux nodes to link network dynamics to specific route points.

If this is right

  • Video uplink stacks can be validated and compared under identical replayed network conditions without deploying physical vehicles.
  • Experiments scale through cloud virtualization while preserving the realism of field-measured cellular dynamics.
  • Route-specific performance can be isolated and tested repeatedly by replaying position traces tied to network events.
  • Production systems become testable for remote autonomous-vehicle monitoring at lower cost and with better reproducibility.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same replay approach could extend to testing other real-time vehicle data streams such as sensor uploads under variable cellular conditions.
  • Scaling the number of virtual nodes might allow emulation of multi-vehicle scenarios to study shared-network contention.
  • Adaptive video encoding logic could be validated against the replayed trace variations to tune for route segments.

Load-bearing premise

Replaying collected traces on virtual Linux nodes will produce end-to-end performance dynamics sufficiently close to real deployments to support valid comparisons and validation of video uplink stacks.

What would settle it

A side-by-side run of the same production video-uplink stack on CloudEmu versus physical vehicles using identical traces, checking whether latency, loss, and throughput metrics match within acceptable bounds.

Figures

Figures reproduced from arXiv: 2605.09910 by Masaki Okada, Nobuhiro Azuma, Soto Anno, Takashi Torii, Takuma Tsubaki, Takuya Tojo.

Figure 1
Figure 1. Figure 1: Overview of CloudEmu architecture. II. PROPOSED SYSTEM: CLOUDEMU As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Example screens on a vehicle-side and video-receiver-side. while also limiting throughput, we double-count queuing and overestimate the emulated delay, especially under heavy load. To address this issue, we propose a congestion-induced delay correction method. It flags intervals where throughput stays below Bth and delay exceeds Dth for at least Tth, then replaces delay and jitter with averages over window… view at source ↗
Figure 4
Figure 4. Figure 4: Overall demonstration setups. CloudEmu is deployed on Amazon EC28 instances for this demo; during the demo, we use a laptop to access CloudEmu and output both the vehicle-side and video-receiver-side views to an external monitor, as shown in [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figure 3
Figure 3. Figure 3: Emulated throughput / delay with and without our correction method. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
read the original abstract

We present CloudEmu, a trace-driven, cloud-native cellular-emulation testbed for vehicle video uplink communication. Reliable, low-latency video uplink over cellular networks is essential for remote monitoring of autonomous vehicles. However, existing testbeds fall into two extremes. Physical-vehicle platforms provide realism but are costly and make validation under identical network conditions difficult, whereas simulations are inexpensive and reproducible but generally cannot replay field-measured end-to-end performance dynamics without substantial calibration or readily run production video-uplink stacks. A software-defined, cloud-native emulation approach can combine the fidelity of trace-driven replay with the agility and scalability that network softwarization principles offer. To this end, we propose CloudEmu that replays time-synchronized cellular and position traces, collected once from vehicles, on commodity Linux-based virtual vehicle and video-receiver nodes. A Linux-based emulation framework couples traffic replay with position replay, tying network dynamics to each point along the route and enabling repeatable, route-aware experiments without repeated on-road trials. Our demo deploys a production-grade video-uplink stack on CloudEmu, allowing attendees to experience low-cost, repeatable trials and controlled comparisons under identical replayed network conditions.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper presents CloudEmu, a trace-driven cloud-native emulation testbed for vehicle video uplink over cellular networks. It replays time-synchronized cellular and position traces collected from real vehicles onto commodity Linux-based virtual vehicle and receiver nodes, coupling traffic replay with position replay to enable repeatable, route-aware experiments. The approach is positioned as combining the fidelity of trace-driven replay with the agility and scalability of network softwarization, and a production-grade video-uplink stack is deployed in a demo for controlled comparisons under identical replayed conditions.

Significance. If the emulation framework can be shown to reproduce end-to-end cellular performance dynamics (latency, jitter, loss, throughput under load) with sufficient fidelity to real vehicle deployments, CloudEmu would provide a valuable, low-cost platform for validating and comparing video uplink stacks without repeated physical trials or the calibration overhead of pure simulation.

major comments (2)
  1. [Abstract] Abstract: The central claim that replaying collected traces on virtual Linux nodes produces end-to-end performance dynamics sufficiently close to real deployments to support valid comparisons and validation of video uplink stacks is unsupported; the manuscript contains no evaluation results, comparisons to physical traces, or validation data whatsoever.
  2. [System description] System description (Linux-based emulation framework): No mechanism is provided for mapping cellular/position traces to network conditions (e.g., via netem/tc, custom kernel modules, or equivalent), nor for preserving production-stack behaviors such as cellular modem interactions, real-time scheduling, or video encoder adaptation; these omissions are load-bearing for the fidelity premise.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive review of our manuscript on CloudEmu. We appreciate the identification of areas where the current presentation requires strengthening to better support the claims. Below we respond point by point to the major comments and indicate the revisions we will make.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The central claim that replaying collected traces on virtual Linux nodes produces end-to-end performance dynamics sufficiently close to real deployments to support valid comparisons and validation of video uplink stacks is unsupported; the manuscript contains no evaluation results, comparisons to physical traces, or validation data whatsoever.

    Authors: We agree that the manuscript does not currently include quantitative evaluation results, direct comparisons to physical traces, or validation data demonstrating closeness of emulated dynamics to real deployments. The paper's primary contribution is the design of the trace-driven emulation testbed and its demonstration with a production video-uplink stack for repeatable experiments. To address this, we will revise the abstract to moderate the fidelity claim and add a dedicated evaluation subsection presenting preliminary validation results (e.g., comparisons of latency, jitter, loss, and throughput under replayed traces versus available real measurements). This will provide the necessary support for the claims while preserving the system-focused nature of the work. revision: yes

  2. Referee: [System description] System description (Linux-based emulation framework): No mechanism is provided for mapping cellular/position traces to network conditions (e.g., via netem/tc, custom kernel modules, or equivalent), nor for preserving production-stack behaviors such as cellular modem interactions, real-time scheduling, or video encoder adaptation; these omissions are load-bearing for the fidelity premise.

    Authors: The manuscript currently describes the emulation framework at the architectural level, emphasizing the coupling of time-synchronized traffic and position replay on Linux virtual nodes. We acknowledge that concrete mechanisms for mapping traces to network conditions (such as netem/tc rules derived from cellular metrics) and provisions for production-stack behaviors (real-time scheduling, encoder adaptation) are not specified. In the revised version we will expand the system description section with these details, including how trace data is translated into traffic-control parameters and any Linux-level configurations used to approximate modem and scheduling effects within the virtual environment. revision: yes

Circularity Check

0 steps flagged

No circularity: system description with no derivations or self-referential claims

full rationale

The paper is a pure system-description manuscript presenting CloudEmu as a trace-driven emulation testbed. It contains no equations, fitted parameters, predictions, first-principles derivations, or load-bearing self-citations. The central claim—that replaying synchronized cellular/position traces on virtual Linux nodes yields repeatable end-to-end dynamics—is advanced by architectural description rather than by any reduction to prior inputs or self-referential logic. No step matches any of the enumerated circularity patterns; the work is self-contained as an engineering artifact.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No mathematical model, free parameters, or new physical entities are introduced; the contribution is an engineering testbed whose correctness depends on unstated assumptions about trace fidelity.

pith-pipeline@v0.9.0 · 5532 in / 988 out tokens · 27138 ms · 2026-05-12T04:28:42.280347+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

  1. [1]

    Jiang, Frank J and others , booktitle=

    work page

  2. [2]

    Assessing the impact of

    Charpentier, Vincent and others , booktitle=. Assessing the impact of

    work page

  3. [3]

    Zhang, Ethan and Masoud, Neda , booktitle=

    work page

  4. [4]

    A Real-World Testbed for

    Elmoghazy, Ammar and others , booktitle=. A Real-World Testbed for

    work page

  5. [5]

    Simulation of tele-operated driving over

    Cislaghi, Valerio and others , booktitle=. Simulation of tele-operated driving over

    work page

  6. [6]

    Hofbauer, Markus and others , booktitle=

    work page

  7. [7]

    WEVJ , volume=

    Teleoperated Driving with Virtual Twin Technology: A Simulator-Based Approach , author=. WEVJ , volume=. 2024 , publisher=

    work page 2024

  8. [8]

    Network emulation for tele-driving application development , author=. Proc. of COMSNETS , year=

    work page

  9. [9]

    Remote driving as the Failsafe: Qualitative investigation of Users' perceptions and requirements towards the

    Li, Shuo and others , journal=. Remote driving as the Failsafe: Qualitative investigation of Users' perceptions and requirements towards the. 2024 , publisher=

    work page 2024

  10. [10]

    IEEE Access , volume=

    A survey on remote operation of road vehicles , author=. IEEE Access , volume=. 2022 , publisher=

    work page 2022

  11. [11]

    Multipath

    De Coninck, Quentin and Bonaventure, Olivier , booktitle=. Multipath

    work page

  12. [12]

    Video Quality Control Method Based on Mobile Network Quality Measurements and Predictions for Reliable Remote Monitoring in Autonomous Driving , author=. Proc. of VTC2025-Spring , pages=

    work page

  13. [13]

    Multipath Transmission System using Multi-layer Bandwidth Prediction for Autonomous Driving , author=. Proc. of VTC2025-Spring , pages=

    work page

  14. [14]

    Sentosa, William and others , booktitle=

    work page

  15. [15]

    ETSI TR 138 901

    Study on channel model for frequencies from 0.5 to 100 GHz (Release 19). ETSI TR 138 901. 2025

    work page 2025