PhD in telecom resilience | Climate hazards and infrastructure failures

PhD in telecom resilience | Climate hazards and infrastructure failures

Reviewed: 8 November 2026
Assessment: Promising with material uncertainties

Quick Decision

This is a full-time, 36-month doctoral position in Valbonne, advertised from 1 February 2027 at €2,300 gross per month. Its distinctive research challenge is to connect climate hazards with telecom service incidents, network structure and possible cascading failures, then test future scenarios and resilience options. The project names relevant network teams and a concrete data-driven programme. Its central feasibility condition is access to incident, topology and electricity-network data: the notices describe these inputs but do not confirm permissions, agreements or validation partners.

The two official notices give different deadlines: CNRS lists 27 November 2026 at 23:59, with a separate position record labeling that time UTC, while Inria lists 22 November 2026. This conflict remains unresolved. Inria says applications must be submitted through its online portal; confirm that it is the controlling route for this vacancy and use 22 November only as a planning safeguard. Before relying on the advertised national scope, ask whether the restricted datasets and partner validation are already secured. CNRS doctoral notice, Inria recruitment notice 2026-10632

Research Problem and Doctoral Work

What Is Unsolved and Why It Matters

Telecom infrastructure supports economic and public services, but weather extremes and long-term climate shifts can affect equipment, access, backhaul and service quality. The project starts from exploratory small-area analyses that found empirical links between weather and network incidents. It proposes extending this to France-wide assessment and prediction.

The scientific challenge is not only to overlay weather and infrastructure maps. The project must align noisy, incomplete and asynchronous weather, service-quality, incident and topology data; distinguish a climate-triggered fault from secondary effects such as power loss or traffic overload; and account for uncertainty in 2030–2050 scenarios. ARCEP’s 2025 resilience note and an OECD review describe how communications services depend on physical assets, power, routing and cross-sector coordination. A 2026 global study maps mobile-site exposure and estimated direct hazard damage. It provides a useful exposure baseline, but does not establish French outage causality, cascade propagation or access to operator data. ARCEP 2025 network-resilience note, OECD 2025 network-resilience paper, IPCC AR6 WGII Chapter 16, 2026 global mobile-infrastructure study

Work, Methods, and Potential Outputs

The advertised work has three linked parts:

1. Data integration and mapping: extend analysis across France using Météo-France/Copernicus weather data, ARCEP or operator incidents, RIPE Atlas and crowdsourced access-quality measurements, and physical network topology; develop a standardized climate-vulnerability measure for equipment or network nodes.
2. Cause and cascade analysis: design statistical models to identify incident causes and propagation, including possible dependencies on electricity distribution and transmission networks (ENEDIS/RTE).
3. Future scenarios and resilience: simulate network behavior under climate projections for 5G standalone, 6G and fibre, then assess redeployment, equipment hardening or emergency routing options.

The notices specify data families, statistical modeling, scenario simulation and intended risk-assessment support, but not a named causal estimator, vulnerability formula, climate ensemble, network simulator, validation design or operational trial. Those should not be assumed. The DRIAS portal provides regionalized climate projections for France, Copernicus documents CMIP6 projections, Météo-France documents observation APIs, RIPE Atlas documents active measurements, and ARCEP publishes mobile service-quality data. These public resources can inform scoping; they do not prove that project-specific rights to operator incident logs, commercial topology or ENEDIS/RTE data are in place. DRIAS climate projections, Copernicus CMIP6 documentation, Météo-France climate-data API, RIPE Atlas REST API manual, ARCEP mobile-quality data

Analyst interpretation: a promising research angle is to move from asset exposure toward observed service effects and cross-infrastructure causes. Whether that becomes a defensible contribution depends on data permissions, event labels, geographic coverage and a credible validation partner. A 2024 review of subsea cable hazards offers adjacent evidence about hazard diversity and geographic exposure, but subsea systems differ from terrestrial access and backhaul networks. Subsea telecom natural-hazard review

School, Programme and Field Context

The stated work location is Valbonne, within the i3S laboratory’s Sophia Antipolis research environment. i3S describes itself as a joint research unit of CNRS and Université Côte d’Azur (UniCA), with the Inria UniCA centre as a secondary supervisory institution. The CNRS notice says the researcher will be hosted jointly in the COATI project-team and SigNet group. These facts describe the host environment; they do not establish which organization signs the employment contract or awards a degree. i3S laboratory presentation, CNRS doctoral notice

UniCA appears at 711–720 overall in QS World University Rankings 2027 and 701–800 overall in THE World University Rankings 2027; THE places it at 301–400 for Computer Science in 2026. These are university-level rankings, not assessments of i3S, COATI, SigNet or this doctorate. QS UniCA profile, THE UniCA profile

The recruitment notices do not identify the doctoral school, registration institution, degree-awarding institution or formal doctoral specialty. UniCA has an École Doctorale STIC and publishes general doctoral-registration guidance, but no official recruitment page assigns this position to that school. UniCA’s general process describes ADUM application, doctoral-school approval and administrative registration; it applies only if UniCA is the confirmed registration route. The topic concerns telecom resilience, but the formal degree specialty remains unconfirmed. Do not infer the awarding body or specialty from the host laboratory, team names or employer classification. UniCA Doctoral School STIC, UniCA doctoral registration guidance

Supervisor and Lab Assessment

Supervisor Research Directions

The Inria notice names Joanna Moulierac as PhD supervisor for COATI and lists Guillaume Urvoy-Keller, Dino Lopez-Pacheco and Clément Courageux-Sudan with SigNet. The CNRS page confirms the joint COATI/SigNet host arrangement. SigNet’s member directory identifies Urvoy-Keller as group leader; its research page describes network-oriented work. The split between formal thesis direction, daily supervision, data ownership and partner contact is not specified, so ask directly. Inria recruitment notice, SigNet members, SigNet research

Recent Important Publications

Three recent team-associated papers indicate adjacent modeling and measurement experience, not a ready-made climate-resilience method:

• “Towards Estimating the Carbon Footprint of Video Streaming” (2025): Guillaume Urvoy-Keller, Joanna Moulierac and co-authors compare five footprint-estimation approaches. The paper traces large differences mainly to system boundaries and modeling assumptions and uses a controlled server deployment to examine them. This is relevant to defining and validating metrics, but it does not study climate hazards or outages. Paper record
• “Charting 5G Energy Efficiency: Flexible Energy Modeling for Sustainable Networks” (2024): Anderson L. de Araujo, Luc Deneire, Guillaume Urvoy-Keller and André L. F. de Almeida develop a granular 5G radio-access energy model and compare baseband-computation estimates with an implementation on an Intel target. This is adjacent infrastructure modeling, not climate-risk research. Paper record
• “A Wi-Fi Energy Model for Scalable Simulation” (2023): Clément Courageux-Sudan, Anne-Cécile Orgerie and Martin Quinson present an energy model intended for scalable Wi-Fi simulation. It is relevant to network simulation expertise but not a climate-resilience study. DBLP record

The 2026 mobile-infrastructure climate study and 2024 subsea-hazard review provide field context and are listed below; neither confirms this project’s datasets or deliverables.

Lab, Resources and Collaboration Environment

The 2025 COATI activity report covers network optimization and management, including causal analysis of network malfunctions. SigNet’s directory and research page document the team’s people and network interests. This is a relevant environment for network measurement and modeling, but publication overlap does not guarantee a particular thesis method, industry access or supervision arrangement. COATI 2025 activity report, SigNet members, SigNet research

Funding, Application and Living Cost

Both official notices advertise a full-time 36-month position, starting 1 February 2027, at €2,300 gross per month. Inria identifies PEPR CARECloud funding; the CARECloud project page says it is part of PEPR Cloud, funded by ANR under France 2030. The collected opportunity record says it is not funded through an EU research-framework programme; this is consistent with the stated French national programme. CNRS doctoral notice, Inria recruitment notice, CARECloud project

The application dates remain contradictory: CNRS gives 27 November 2026, 23:59, and a separate position record labels that time UTC, while Inria gives 22 November 2026. Inria specifically says to submit online through its website and warns that applications sent by other routes may not be processed. That makes the Inria portal the clearest stated route, but it does not resolve which deadline controls; verify both with the recruiter and treat 22 November as a planning safeguard only.

The benefits wording also differs: CNRS lists 44 annual leave/RTT days, while Inria lists seven weeks of annual leave plus 10 RTT days. Both mention 75% public-transport reimbursement. Confirm the contracting entity and applicable terms. Gross pay is stated; net pay is not.

UniCA’s 2026–27 schedule lists €398 doctoral registration fees and a €105 CVEC, subject to relevant rules and exemptions. These are not confirmed liabilities for this position, because its registration institution is not named. The current CROUS 2026–27 schedule lists example monthly rent plus charges of €374.33 for a 21 m² Newton studio and €413.81 for a 21 m² Dolines studio in Valbonne. That is about 16–18% of the advertised gross salary before deductions and all other expenses; the ratio does not establish affordability. Employee eligibility and room availability are unknown, and a complete local budget has not been verified. UniCA 2026–27 fee schedule, CROUS 2026–27 Sophia/Valbonne tariffs

Applicant Fit and Preparation

Required Background and Skills

The employer profile asks for a Master’s or engineering degree in Computer Science, Data Science or Telecommunications; strong data-analysis, machine-learning/deep-learning and time-series skills; a good understanding of telecom networks and Internet protocols; and interest in interdisciplinary work across computing, networks, climate and the environment. The notices reviewed do not set a French or other language threshold. No language requirement or level is inferred; ask the recruiter about the working language, documents and interview. CNRS candidate profile, Inria keys to success

Preferred Experience and Suggested Preparation

The notices do not clearly separate a preferred-experience list from core conditions, and do not state a prior-research-years requirement in the employer text reviewed. The following are analyst-suggested preparation, not additional eligibility rules:

• Show a reproducible time-series or network-data project, including how missing and asynchronous observations were handled.
• Be ready to explain how you would distinguish correlation from a plausible causal mechanism and validate a cascade claim.
• Demonstrate familiarity with telecom protocols, service-quality measures and the limits of public probe data.
• Prepare questions about scenario choice and uncertainty; the advert does not name a particular climate ensemble or model.

APEC’s occupational profile describes adjacent telecom-engineering work such as coverage analysis, network deployment, service-quality measurement and fault investigation across operators and other large network users. It is an occupation profile, not a climate-resilience labour-market or doctoral-placement study; it does not support a project-specific employment or earnings forecast. APEC telecom-engineer profile

Next Actions and Questions to Verify

Before Applying

• Ask whether the 22 November Inria date or 27 November at 23:59 CNRS date (a separate record labels this UTC) controls, and whether the Inria online portal is the required submission route. Plan around the earlier date until the discrepancy is resolved.
• Confirm that applications remain open, identify the contracting entity, and request the required documents and any language expectations.
• Ask which doctoral school and registration institution apply, which institution awards the doctorate, and what formal doctoral specialty will appear in the registration. The notices do not answer these questions.
• Ask whether operator incident records, topology and ENEDIS/RTE data are already available under agreements and permissions, or would need to be obtained during the thesis.

At Interview

• Who is the formal thesis director, and how are supervision and daily responsibilities split across COATI and SigNet?
• Which incident labels, topology coverage and operator or power-grid partners are committed? How will causal and cascade models be validated?
• Which climate projections, time horizons and uncertainty treatment are planned? Are resilience options tested in simulation or with operational partners?

Before Accepting an Offer

• Get written confirmation of employer, registration school, degree-awarding institution and formal doctoral specialty.
• Confirm net monthly pay, leave/RTT, transport benefits, tuition/CVEC liability and any exemption.
• Check whether an employed doctoral researcher can access CROUS housing and whether a room is available; budget for costs beyond rent.

Evidence and Sources

Core Institutional and Project Sources

• CNRS doctoral recruitment notice
• Inria recruitment notice 2026-10632
• i3S laboratory presentation
• UniCA Doctoral School STIC
• UniCA doctoral registration guidance
• QS UniCA profile
• THE UniCA profile

Supervisor and Publication Sources

• COATI 2025 activity report
• SigNet members
• SigNet research
• 2025 video-streaming footprint paper
• 2024 flexible 5G energy-model paper
• 2023 scalable Wi-Fi energy-model paper

Climate, Network and Data Sources

• ARCEP 2025 network-resilience note
• OECD 2025 network-resilience paper
• IPCC AR6 WGII Chapter 16
• DRIAS climate projections
• Copernicus CMIP6 documentation
• Météo-France climate-data API documentation
• RIPE Atlas REST API manual
• ARCEP mobile-quality data
• 2026 global mobile-infrastructure climate-risk study
• 2024 subsea telecom natural-hazard review

Funding, Application, Cost and Market Sources

• UniCA 2026–27 fee schedule
• UniCA doctoral funding overview
• CARECloud project
• APEC telecom-engineer occupation profile
• CROUS 2026–27 Sophia/Valbonne housing tariffs