Vasileios Giotsas
Research Subjects
Vasileios Giotsas has a total of 99 co-authors. Here are the top co-authors by number of shared publications:
- Shi Zhou
University College London
6 shared publications Christoph Dietzel
Products and Research Department, DE-CIX, Frankfurt, Germany
4 shared publications- Dimitrios D. Vergados
University of Piraeus
4 shared publications - Joseph Doyle
Queen Mary University of London, London, United Kingdom
4 shared publications Kc Claffy
CAIDA / UC San Diego
4 shared publications- Matthew Luckie
UC San Diego, La Jolla, CA, USA
4 shared publications
Latest Publications
- Open Access
- 12 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 888-902
RScope: Unveiling Global ROV Deployments and Dependencies in the Post-ROV Era
The Resource Public Key Infrastructure (RPKI) and Route Origin Validation (ROV) are critical to securing Internet routing, yet measuring real-world ROV deployment and dependencies remains challenging. Existing methods (control-plane analysis and data-plane probing) often struggle to distinguish local ROV adoption from upstream filtering at scale and face diminishing visibility as ROV adoption grows. We present RScope, a measurement framework that dynamically manipulates Route Origin Authorizations (ROAs) from controlled publication points to isolate the impact of individual Relying Party (RP) servers. By selectively invalidating prefixes for specific RPs and probing connectivity changes, RScope maps dependencies among 21,827 ASes, their transit providers, and RP infrastructure. Our findings reveal that 1,127 ASes actively deploy ROV in IPv6, while 1,815 benefit from upstream protection, leaving 69.6% of such "protected" ASes vulnerable to local hijacks. Notably, ROV deployment by major ISPs strongly shape global security, enabling it in the top 100 ASes protects 27.6% of networks, but disabling it in Tier-1s endangers 23.8%. We also uncover operational delays: router enforcement lags ROA updates by 37 minutes on average. These results highlight systemic risks in incremental ROV adoption, emphasizing the need for broader deployment. We open-source tools and datasets to support ongoing routing security efforts.
- 12 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 888-902
Open AccessRScope: Unveiling Global ROV Deployments and Dependencies in the Post-ROV Era
The Resource Public Key Infrastructure (RPKI) and Route Origin Validation (ROV) are critical to securing Internet routing, yet measuring real-world ROV deployment and dependencies remains challenging. Existing methods (control-plane analysis and data-plane probing) often struggle to distinguish local ROV adoption from upstream filtering at scale and face diminishing visibility as ROV adoption grows. We present RScope, a measurement framework that dynamically manipulates Route Origin Authorizations (ROAs) from controlled publication points to isolate the impact of individual Relying Party (RP) servers. By selectively invalidating prefixes for specific RPs and probing connectivity changes, RScope maps dependencies among 21,827 ASes, their transit providers, and RP infrastructure. Our findings reveal that 1,127 ASes actively deploy ROV in IPv6, while 1,815 benefit from upstream protection, leaving 69.6% of such "protected" ASes vulnerable to local hijacks. Notably, ROV deployment by major ISPs strongly shape global security, enabling it in the top 100 ASes protects 27.6% of networks, but disabling it in Tier-1s endangers 23.8%. We also uncover operational delays: router enforcement lags ROA updates by 37 minutes on average. These results highlight systemic risks in incremental ROV adoption, emphasizing the need for broader deployment. We open-source tools and datasets to support ongoing routing security efforts.
- Open Access
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 990-1008
A Global Inference and Assessment of Large Shared IP Addresses
- 11 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 990-1008
Open AccessA Global Inference and Assessment of Large Shared IP Addresses
- Open Access
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 337-364
metAScritic: Reframing AS-Level Topology Discovery as a Recommendation System
Despite prior efforts, the vast majority of the AS-level topology of the Internet remains hidden from BGP and traceroute vantage points. In this work, we introduce metAScritic, a novel system inspired by recommender system literature, designed to infer interconnections within a given metro. metAScritic uses the intuition that the connectivity matrix at a given metro is a low-rank system, since ASes employ similar peering strategies according to their infrastructures, traffic profiles, and business models. This approach allows metAScritic to accurately reconstruct the complete peering connectivity by measuring a strategic subset of interconnections that capture ASes' underlying peering strategies. We evaluate metAScritic's performance across six large metropolitan areas, achieving an average F-score of 0.88 on various validation datasets, including ground truth. metAScritic measures more than 86K edges and infers more than 368K edges, compared to the 13K edges observed for this subset of ASes in public BGP feeds -- an increase of (24X) what is currently seen. We study the impact of our inferred links on Internet properties, illustrating the extent of the Internet's flattening and demonstrating our ability to better predict the impact of route leaks and prefix hijacks, compared to relying only on the existing public view.
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 337-364
Open AccessmetAScritic: Reframing AS-Level Topology Discovery as a Recommendation System
Despite prior efforts, the vast majority of the AS-level topology of the Internet remains hidden from BGP and traceroute vantage points. In this work, we introduce metAScritic, a novel system inspired by recommender system literature, designed to infer interconnections within a given metro. metAScritic uses the intuition that the connectivity matrix at a given metro is a low-rank system, since ASes employ similar peering strategies according to their infrastructures, traffic profiles, and business models. This approach allows metAScritic to accurately reconstruct the complete peering connectivity by measuring a strategic subset of interconnections that capture ASes' underlying peering strategies. We evaluate metAScritic's performance across six large metropolitan areas, achieving an average F-score of 0.88 on various validation datasets, including ground truth. metAScritic measures more than 86K edges and infers more than 368K edges, compared to the 13K edges observed for this subset of ASes in public BGP feeds -- an increase of (24X) what is currently seen. We study the impact of our inferred links on Internet properties, illustrating the extent of the Internet's flattening and demonstrating our ability to better predict the impact of route leaks and prefix hijacks, compared to relying only on the existing public view.
- Open Access
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 165-182
What's in the Dataset? Unboxing the APNIC per AS User Population Dataset
The research measurement community needs methods and datasets to identify user concentrations and to accurately weight ASes against each other for analyzing measurements' coverage. However, academic researchers traditionally lack visibility into how many users are in each network or how much traffic flows to each network and so often fall back on treating all IP addresses or networks equally. As an alternative, some recent studies have used the APNIC per AS Population Estimates dataset, but it is unvalidated and its methodology is not fully public. In this work, we validate its use as a fairly reliable user population indicator. Our approach includes a detailed comparative analysis using a global CDN dataset, providing concrete evidence of the APNIC dataset's accuracy. We find that the APNIC per-AS user estimates closely align with the Content Delivery Network (CDN) per-AS user estimates in 51.2% of countries and correctly identify the largest networks in 93.9% of cases. When we investigate the agreement with CDN traffic volume, the APNIC dataset closely aligns in 36.5% of countries, increasing to 91.0% when focusing only on larger networks. We also evaluate the limitations of the APNIC dataset, particularly its inability to accurately identify user populations for ASes in certain countries. To address this, we introduce new methods to improve its usability by focusing on the statistical representativeness of the underlying data collection process and ensuring consistency across several public datasets.
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 165-182
Open AccessWhat's in the Dataset? Unboxing the APNIC per AS User Population Dataset
The research measurement community needs methods and datasets to identify user concentrations and to accurately weight ASes against each other for analyzing measurements' coverage. However, academic researchers traditionally lack visibility into how many users are in each network or how much traffic flows to each network and so often fall back on treating all IP addresses or networks equally. As an alternative, some recent studies have used the APNIC per AS Population Estimates dataset, but it is unvalidated and its methodology is not fully public. In this work, we validate its use as a fairly reliable user population indicator. Our approach includes a detailed comparative analysis using a global CDN dataset, providing concrete evidence of the APNIC dataset's accuracy. We find that the APNIC per-AS user estimates closely align with the Content Delivery Network (CDN) per-AS user estimates in 51.2% of countries and correctly identify the largest networks in 93.9% of cases. When we investigate the agreement with CDN traffic volume, the APNIC dataset closely aligns in 36.5% of countries, increasing to 91.0% when focusing only on larger networks. We also evaluate the limitations of the APNIC dataset, particularly its inability to accurately identify user populations for ASes in certain countries. To address this, we introduce new methods to improve its usability by focusing on the statistical representativeness of the underlying data collection process and ensuring consistency across several public datasets.
- Open Access
- 18 August 2024
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Networking
- Vol. 2 (CoNEXT3) , 1-21
Geofeeds: Revolutionizing IP Geolocation or Illusionary Promises?
Despite more than twenty years of efforts, the research community is still looking for a publicly available Internet-scale IP geolocation dataset with an explainable methodology. Recently, a new hope has appeared, with the emergence of geofeeds. Geofeeds are a self published IP geolocation dataset where operators give the geolocation of their IP addresses, with the underlying idea being that other network providers can tune their services to better serve the IP address depending on its geolocation. In this paper, we analyze whether the hope of finally obtaining the golden geolocation dataset is a real possibility or a mirage. Two years after the standardization of geofeeds, we look at how they are adopted by operators, and what is their accuracy, and how we can use them for operational and research scenarios. First, geofeeds are in the early adoption process with 1.50% and 0.70% of the allocated IPv4 and IPv6 prefixes covering the geofeed prefixes. Second, even if we cannot use geofeeds as ground truth as we found 0.9%, 4.0%, and 8.5% of the client, router, and server IP addresses with an erroneous geofeed, most of them look correct and provide at least a geolocation hint for building an internet scale IP geolocation dataset. Finally, we provide some recommendations on how to use geofeeds and how we could improve the format and the process of sharing geofeeds to improve their quality.
- 18 August 2024
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Networking
- Vol. 2 (CoNEXT3) , 1-21
Open AccessGeofeeds: Revolutionizing IP Geolocation or Illusionary Promises?
Despite more than twenty years of efforts, the research community is still looking for a publicly available Internet-scale IP geolocation dataset with an explainable methodology. Recently, a new hope has appeared, with the emergence of geofeeds. Geofeeds are a self published IP geolocation dataset where operators give the geolocation of their IP addresses, with the underlying idea being that other network providers can tune their services to better serve the IP address depending on its geolocation. In this paper, we analyze whether the hope of finally obtaining the golden geolocation dataset is a real possibility or a mirage. Two years after the standardization of geofeeds, we look at how they are adopted by operators, and what is their accuracy, and how we can use them for operational and research scenarios. First, geofeeds are in the early adoption process with 1.50% and 0.70% of the allocated IPv4 and IPv6 prefixes covering the geofeed prefixes. Second, even if we cannot use geofeeds as ground truth as we found 0.9%, 4.0%, and 8.5% of the client, router, and server IP addresses with an erroneous geofeed, most of them look correct and provide at least a geolocation hint for building an internet scale IP geolocation dataset. Finally, we provide some recommendations on how to use geofeeds and how we could improve the format and the process of sharing geofeeds to improve their quality.
- Open Access
- 20 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 15-22
Investigating Location-aware Advertisements in Anycast IP Networks
Anycast routing offers transparent service replication by distributing traffic across multiple Points of Presence (PoP). By advertising the same IP prefix from each PoP via BGP, traffic is routed to the nearest server, minimizing user latency. Despite its perceived benefits, prior research suggests IP anycast often falls short, with clients routed to distant replicas, increasing latency. Selective announcements made by anycast ASes contribute to this inefficiency, serving as a traffic engineering strategy to control incoming traffic flows. In this work, we aim to shed light on the prevalence and rationale behind selective announcements in anycast networks. Through empirical evidence, we identify their significant adoption, primarily driven by the geolocation-agnostic BGP best path selection process. In particular, we observe that 84.06% of anycast ASes announce at least one of their anycast prefixes to a specific subset of their neighbors, whereas 80% of these selective announcers announce 100% of their prefixes selectively. This research represents an initial step towards comprehending the effects of selective and location-based routing policies in anycast IP networks.
- 20 July 2024
- Published by Association for Computing Machinery (ACM)
- p. 15-22
Open AccessInvestigating Location-aware Advertisements in Anycast IP Networks
Anycast routing offers transparent service replication by distributing traffic across multiple Points of Presence (PoP). By advertising the same IP prefix from each PoP via BGP, traffic is routed to the nearest server, minimizing user latency. Despite its perceived benefits, prior research suggests IP anycast often falls short, with clients routed to distant replicas, increasing latency. Selective announcements made by anycast ASes contribute to this inefficiency, serving as a traffic engineering strategy to control incoming traffic flows. In this work, we aim to shed light on the prevalence and rationale behind selective announcements in anycast networks. Through empirical evidence, we identify their significant adoption, primarily driven by the geolocation-agnostic BGP best path selection process. In particular, we observe that 84.06% of anycast ASes announce at least one of their anycast prefixes to a specific subset of their neighbors, whereas 80% of these selective announcers announce 100% of their prefixes selectively. This research represents an initial step towards comprehending the effects of selective and location-based routing policies in anycast IP networks.
- 5 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 21 (3) , 3623-3635
These are Not the PLCs You are Looking for: Obfuscating PLCs to Mimic Honeypots
Industry 4.0 and the trend of connecting legacy Industrial Control Systems (ICSs) to public networks have exposed these systems to various online threats. To combat these threats, honeypots have been widely used to provide proactive monitoring, detection and deception security capabilities. However, skilled attackers are now adept at fingerprinting and avoiding honeypots. Therefore, we take a fundamentally different approach in this paper. Instead of the honeypot representing a real system, we deploy it as a deterrent. Through obfuscation, the aim is to make an attacker believe the real system is a honeypot and collect threat intelligence data on the attacker. To achieve this, we introduce a new obfuscation technique that allows real ICSs to present themselves as honeypots. By taking advantage of honeypot fingerprinting techniques, we are able to deter attackers from interacting with the real Programmable Logic Controller (PLC) within the industrial network. The approach is implemented and evaluated using different penetration testing tools and an expert evaluation highlighting the benefits of obfuscation in that potential adversaries would be misled into assuming the PLC is a honeypot.
- 5 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 21 (3) , 3623-3635
These are Not the PLCs You are Looking for: Obfuscating PLCs to Mimic Honeypots
Industry 4.0 and the trend of connecting legacy Industrial Control Systems (ICSs) to public networks have exposed these systems to various online threats. To combat these threats, honeypots have been widely used to provide proactive monitoring, detection and deception security capabilities. However, skilled attackers are now adept at fingerprinting and avoiding honeypots. Therefore, we take a fundamentally different approach in this paper. Instead of the honeypot representing a real system, we deploy it as a deterrent. Through obfuscation, the aim is to make an attacker believe the real system is a honeypot and collect threat intelligence data on the attacker. To achieve this, we introduce a new obfuscation technique that allows real ICSs to present themselves as honeypots. By taking advantage of honeypot fingerprinting techniques, we are able to deter attackers from interacting with the real Programmable Logic Controller (PLC) within the industrial network. The approach is implemented and evaluated using different penetration testing tools and an expert evaluation highlighting the benefits of obfuscation in that potential adversaries would be misled into assuming the PLC is a honeypot.
- Open Access
- 24 October 2023
- Published by Association for Computing Machinery (ACM)
- p. 16-29
Replication: 20 Years of Inferring Interdomain Routing Policies
- 24 October 2023
- Published by Association for Computing Machinery (ACM)
- p. 16-29
Open AccessReplication: 20 Years of Inferring Interdomain Routing Policies
- Open Access
- 21 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Mapping the Commercial VPN Ecosystem from Cost-Quality Perspective
Reliance on commercial virtual private networks has increased manifold in the recent past. This paradigm shift of global internet users is primarily driven by their need of online privacy, censorship circumvention, and accessing geo-filtered content among other motivations. Variety of VPN services underscored by their opaque nature captures huge user base which is technically unsophisticated, and has very limited means to verify claims of a given VPN service. Capitalizing on extensive literature and online review, we have selected a sample of 60 x VPNs (from categories of free as well as paid services) which adequately represents the commercial VPN market from cost-quality perspective. An elaborate eco-system analysis reveals significant variations against the claims of service providers. Our crystallized infrastructure testing methodology executed over selected VPNs reveals: (1) Anonymity index of free versus paid VPNs differs widely at all levels i.e., IP, subnet, country and ISP; (2) 54% of selected VPN’s server locations in attractive geographic locations are virtual instead of physical; (3) 4/10 VPN tested services have been found to share their vantage points at subnet level; and (4) Selective co-location of 22% vantage points have also been observed.
- 21 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Open AccessMapping the Commercial VPN Ecosystem from Cost-Quality Perspective
Reliance on commercial virtual private networks has increased manifold in the recent past. This paradigm shift of global internet users is primarily driven by their need of online privacy, censorship circumvention, and accessing geo-filtered content among other motivations. Variety of VPN services underscored by their opaque nature captures huge user base which is technically unsophisticated, and has very limited means to verify claims of a given VPN service. Capitalizing on extensive literature and online review, we have selected a sample of 60 x VPNs (from categories of free as well as paid services) which adequately represents the commercial VPN market from cost-quality perspective. An elaborate eco-system analysis reveals significant variations against the claims of service providers. Our crystallized infrastructure testing methodology executed over selected VPNs reveals: (1) Anonymity index of free versus paid VPNs differs widely at all levels i.e., IP, subnet, country and ISP; (2) 54% of selected VPN’s server locations in attractive geographic locations are virtual instead of physical; (3) 4/10 VPN tested services have been found to share their vantage points at subnet level; and (4) Selective co-location of 22% vantage points have also been observed.
- Open Access
- 31 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 59935-59953
On Improving the Accuracy of Internet Infrastructure Mapping
This study identifies a method to create fine-grained multilayer maps of the Internet’s structure, which are currently lacking. We begin with an investigation of current techniques for geolocating hosts using passive, active, and hybrid methods. This is followed by a survey of the fundamental problems that IP geolocation techniques must address. The survey points to the obvious difficulties in using Delay-Distance models and suggests that the use of Return Trip Times can lead to highly misleading results. We therefore develop a new procedure that combines state-of-the-art methods to avoid many of the fundamental problems in Internet topology mapping, whilst creating finer-grained internet maps than those currently available. This procedure is tested on the UK infrastructure by conducting a series of tests using distributed measurement points provided by the RIPE Atlas platform. Our results show that we can accurately geolocate routers between two endpoints to create a fine-grained map of the internet infrastructure involved in our measurements. Researchers have long recognized the scarcity of ground truth datasets where IP geolocation is a concern. As a byproduct of our new method reported in this paper, we create a validation dataset that maps hundreds of IP addresses to geo-coordinate landmarks or vantage points, which is highly desirable for IP geolocation research. Finally, we discuss some limitations of this method, and we summarise the next steps toward accurate and complete Internet infrastructure maps.
- 31 May 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 59935-59953
Open AccessOn Improving the Accuracy of Internet Infrastructure Mapping
This study identifies a method to create fine-grained multilayer maps of the Internet’s structure, which are currently lacking. We begin with an investigation of current techniques for geolocating hosts using passive, active, and hybrid methods. This is followed by a survey of the fundamental problems that IP geolocation techniques must address. The survey points to the obvious difficulties in using Delay-Distance models and suggests that the use of Return Trip Times can lead to highly misleading results. We therefore develop a new procedure that combines state-of-the-art methods to avoid many of the fundamental problems in Internet topology mapping, whilst creating finer-grained internet maps than those currently available. This procedure is tested on the UK infrastructure by conducting a series of tests using distributed measurement points provided by the RIPE Atlas platform. Our results show that we can accurately geolocate routers between two endpoints to create a fine-grained map of the internet infrastructure involved in our measurements. Researchers have long recognized the scarcity of ground truth datasets where IP geolocation is a concern. As a byproduct of our new method reported in this paper, we create a validation dataset that maps hundreds of IP addresses to geo-coordinate landmarks or vantage points, which is highly desirable for IP geolocation research. Finally, we discuss some limitations of this method, and we summarise the next steps toward accurate and complete Internet infrastructure maps.