International audience

Modern geolocation ecosystems rely on diverse technical solutions and architectures, often proprietary, making it difficult to develop a global understanding of the privacy implications of location data production. This challenge is particularly critical given the ubiquity of geolocation in modern digital infrastructures and the central role of location data in privacy concerns. Yet, existing work largely focuses on isolated case studies, resulting in a fragmented understanding of the privacy risks associated with location data production. In this work, we introduce an abstract model of geolocation ecosystems together with a systematic methodology for analyzing their privacy implications, which we apply to nine representative case studies spanning a broad range of architectures, from OS-level geolocation services to object-tracking platforms. This comparative analysis identifies structural design choices that significantly impact users' privacy and reveals common structural privacy risks across heterogeneous ecosystems, which reflect architectural design decisions rather than security vulnerabilities or poor system design. These findings, together with gaps identified in existing defense mechanisms, motivate research directions aimed at strengthening privacy in future geolocation architectures.

Location underpins critical Internet services, yet our primary mechanism for Internet localization, IP-based geolocation, fails to meet the needs of all stakeholders. User location is conflated with network location, leading to a fundamental mismatch between the goals of content providers, infrastructure operators, and regulators. As users increasingly adopt privacy-preserving technologies that obscure their network identity, this mismatch becomes more pronounced, making localization even more challenging. This paper argues that the problem cannot be solved by simply improving the accuracy of incumbent mechanisms that are inappropriately applied today to solve multiple, unrelated problems. Instead, we require a new approach for localization on the Internet.

Machine Learning (ML)-based traffic analysis relies on a data processing pipeline consisting of multiple steps that filter, process, and collect statistics, or features from raw network traffic. These steps are typically performed by in-network measurement systems deployed in existing network fabric (e.g., programmable switches) or using off-the-shelf hardware (e.g., commodity servers). In both deployment scenarios, these systems come with limited processing budgets that must be finely tuned to precisely collect the required features. Unfortunately, the ever growing traffic volume on modern networks can exhaust these budgets, ultimately resulting in packet loss. In this paper, we investigate the impact of packet loss on the performance of ML-based traffic analysis systems. As losses introduce bias in the final features set provided to the machine learning model, we hypothesize that they will negatively impact model performance. We evaluate this hypothesis by analyzing the performance of two different ML models—service classification and QoE analysis—trained on a dataset of video flows, and we measure the impact of two different packet loss models: probabilistic and bursty losses. Our results show that sporadic packet loss has little impact on performance. Conversely, bursty losses, which are more common for packet processing systems, can lead to a significant negative impact.

Monitoring the quality of streaming video applications is important for Internet service providers (ISPs) to detect network issues and facilitate capacity planning. Machine Learning (ML) inference models have emerged as an effective solution to determine service quality using network traffic. However, while much focus has been on enhancing model performance, little attention has been given to deploying these models across entire networks. This paper introduces a new placement approach of quality inference models and their associated tasks to enhance the monitoring of video streaming applications over an entire mobile traffic network. Starting from the observation that inference tasks require the deployment of multiple components to, first, calculate input features from raw traffic, and then execute the inference models, we define the placement problem as an integer programming problem and, given its NP-hardness, we provide a heuristic solution, experimentally close to the optimum, based on the relaxation and the rounding of fractional solutions. We highlight that decoupling these components for the inference of network traffic can be beneficial in terms of total accuracy of the ML inference tasks. Finally, we experimentally show that our solution outperforms state-of-the-art placement techniques by ~30% of accuracy of the deployed inference models.

Edge-based live video analytics are a promising approach to reduce bandwidth overheads caused by the transmission of raw video streams to the cloud. However, the limited resources available on edge devices make it challenging to successfully process video streams in real-time. This gets further exacerbated when attempting to process video streams from mobile cameras. While mobile cameras are a desirable source of information, thanks to them being in the right place at the right time, they are inherently dynamic and unpredictable. To address these challenges, we propose VideoJam, a decentralized load balancing solution for live video analytics. VideoJam uses a set of load balancers to balance incoming video traffic across replicas without the need of centralized coordination. Exploiting the inherent load dynamicity generated by different video sources, VideoJam predicts the incoming load for each processing component and offloads excessive traffic to less-loaded neighbors. Further, VideoJam operates independently of deployed configurations and cameras present in the system, dynamically adapting to handle load changes and balance video traffic across available resources. Our evaluation shows that VideoJam can adapt to different mixes of mobile and fixed cameras, as well as quickly adapting to configuration changes occurring at runtime. Compared to state-of-the-art solutions, VideoJam achieves 2.91× lower response time, while reducing video data loss by more than 4.64× and generating lower bandwidth overheads.

Millions of video cameras are deployed globally across major cities for learning-based video analytic (VA) applications, such as object detection. Video streams from the cameras are either sent over the wide-area network to be processed by the cloud or are (at least partially) processed in a local edge workstation, incurring significant latency and elevated financial costs. In this paper, to minimize reliance on the cloud and overcome the unavailability of high-compute workstations on edge, we investigate the use of heterogeneous and distributed embedded devices as edge nodes shared by multiple cameras to fully serve the video processing needs of a VA application (without requiring cloud support). We present OVIDA, an edge-only orchestrator to deploy VA application(s) on a distributed edge environment to maximize accuracy. Given the resource-constrained nature of edge nodes, OVIDA disaggregates the VA application pipeline into multiple modules. OVIDA's core functionality and contributions are: (i) optimizing the placement and replication of the VA application modules across the edge nodes to maximize the throughput, and in turn, accuracy; and (ii) an adaptive model selection algorithm for VA modules based on accuracy-throughput tradeoff to maximize accuracy in response to varying load conditions. To further improve performance, OVIDA employs a central-queue-based design (instead of the usual push-based design), which also obviates the need for complex load balancing algorithms. We implement OVIDA on top of Kubernetes and evaluate its performance for three VA applications, supported over a heterogeneous edge cluster under varying network conditions. When compared against several baselines in our evaluation, we achieve throughput and accuracy gains of at least 51% and 28%.

Recently, much attention has been devoted to the development of generative network traces and their potential use in supplementing real-world data for a variety of data-driven networking tasks. Yet, the utility of existing synthetic traffic approaches are limited by their low fidelity: low feature granularity, insufficient adherence to task constraints, and subpar class coverage. As effective network tasks are increasingly reliant on raw packet captures, we advocate for a paradigm shift from coarse-grained to fine-grained traffic generation compliant to constraints. We explore this path employing controllable diffusion-based methods. Our preliminary results suggest its effectiveness in generating realistic and fine-grained network traces that mirror the complexity and variety of real network traffic required for accurate service recognition. We further outline the challenges and opportunities of this approach, and discuss a research agenda towards text-to-traffic synthesis.

Edge Computing exploits computational capabilities deployed at the very edge of the network to support applications with low latency requirements. Such capabilities can reside in small embedded devices that integrate dedicated hardware - e.g., a GPU - in a low cost package. But these devices have limited computing capabilities compared to standard server grade equipment. When deploying an Edge Computing based application, understanding whether the available hardware can meet target requirements is key in meeting the expected performance. In this paper, we study the feasibility of deploying Augmented Reality applications using Embedded Edge Devices (EEDs). We compare such deployment approach to one exploiting a standard dedicated server grade machine. Starting from an empirical evaluation of the capabilities of these devices, we propose a simple theoretical model to compare the performance of the two approaches. We then validate such model with NS-3 simulations and study their feasibility. Our results show that there is no one-fits-all solution. If we need to deploy high responsiveness applications, we need a centralized server grade architecture and we can in any case only support very few users. The centralized architecture fails to serve a larger number of users, even when low to mid responsiveness is required. In this case, we need to resort instead to a distributed deployment based on EEDs.

One of the most fundamental capabilities of swarm robotics is their ability to cooperate. This implies that swarm robots must exchange information with each other or with a centralized controller. However, this communication is often assumed to be perfect, an assumption that does not reflect real-world conditions, where impairments can affect the Packet Delivery Ratio (PDR) over wireless links. One essential application of swarm robotic cooperation is exploration and mapping in a timely and accurate manner. This paper studies how communication impairments can have a drastic impact on the performance of robotic swarms in critical missions such as exploration. We use an improved version of the Atlas algorithm to simulate the effect of various PDRs on the exploration mission execution performance, with the key indicator being mapping completion time. Our results show that the time it takes to complete area exploration increases exponentially as the PDR decreases linearly. Based on our results, we emphasise the importance of considering methods that minimize the delay caused by lossy communication when designing and implementing algorithms for robotic swarm exploration.

We envision swarms of mm-scale micro-robots to be able to carry out critical missions such as exploration and mapping for hazard detection and search and rescue. These missions share the need to reach full coverage of the explorable space and build a complete map of the environment. To minimize completion time, robots in the swarm must be able to exchange information about the environment with each other. However, communication between swarm members is often assumed to be perfect, an assumption that does not reflect real-world conditions, where impairments can affect the Packet Delivery Ratio (PDR) of the wireless links. This paper studies how communication impairments can have a drastic impact on the performance of a robotic swarm. We present Atlas 2.0, an exploration algorithm that natively takes packet loss into account. We simulate the effect of various PDRs on robotic swarm exploration and mapping in three different scenarios. Our results show that the time it takes to complete the mapping mission increases significantly as the PDR decreases: on average, halving the PDR triples the time it takes to complete mapping. We emphasise the importance of considering methods to compensate for the delay caused by lossy communication when designing and implementing algorithms for robotics swarm coordination.

Exploration and mapping is a fundamental capability of a swarm of robots: robots enter an unknown area, explore it, and collectively build a map of it. This capability is important regardless of whether the robots are crawling, flying, or swimming. Existing exploration and mapping algorithms tend to either be inefficient, or rely on having a dense swarm of robots. This paper introduces Atlas, an exploration and mapping algorithm for sparse swarms of robots, which completes a full exploration even in the extreme case of a single robot. We develop an open-source simulator and show that Atlas outperforms the state-of-the-art in terms of exploration speed and completeness of the resulting map.

In the recent years it has become increasingly evident that the current end-to-end host-centric communication paradigm will not be capable of meeting the ongoing demand for massive data rates and ultra-low latency. With the advent of fifth generation of cellular architecture (5G) to support these requirements on the wireless edge of the network, the need for core network solutions to play a complementary role is conspicuous. In this paper we present and tackle some of the challenges of deploying a Future Internet Architecture (FIA), called MobilityFirst (MF), specifically for 5G use case scenarios. We report our findings of the deployment based on a setup on a small- scale testbed (ORBIT) and a nation-wide distributed testbed (GENI), and illustrate some results for the use case of device multihoming, in comparison with current TCP/IP based solution, i.e. Multipath TCP.

The introduction of Dynamic Adaptive Streaming over HTTP (DASH) helped reduce the consumption of resources in video delivery, but its client-based rate adaptation is unable to optimally use the available end-to-end network bandwidth. We consider the problem of optimizing the delivery of video content to mobile clients while meeting the constraints imposed by the available network resources. Observing the bandwidth available in the network's two main components, core network, transferring the video from the servers to edge nodes close to the client, and the edge network, which is in charge of transferring the content to the user via wireless links, we aim to find an optimal solution by exploiting the predictability of future user requests of sequential video segments, as well as the knowledge of available infrastructural resources at the core and edge wireless networks in a given future time window. Instead of regarding the bottleneck of the end-to-end connection as our throughput, we distribute the traffic load over time and use intermediate nodes between the server and the client for buffering video content to achieve higher throughput, and ultimately significantly improve the Quality of Experience for the end user in comparison with current solutions.

This paper presents a novel approach to achieving scalable push multicast services using the distributed global name resolution service associated with emerging name-based network architectures. The proposed named-object multicast (NOMA) scheme employs unique names to identify multicast groups while using the global name resolution service (GNRS) to store the tree structure and maintain current mappings to mobile end-user addresses. The NOMA scheme achieves improved scalability and performance over conventional multicast protocols such as PIM-SM and MDSP by taking advantage of the GNRS to simplify tree management and limit control overhead. Performance evaluation results including comparisons with IP multicast are given using a combination of analysis and NS-3 simulation. The results show good scalability properties along with low control overhead for medium to large multicast groups. In addition, NOMA seamlessly handles mobility for end-hosts subscribed to a group, avoiding data losses upon mobility events. Results further demonstrate how separating names from addresses enables NOMA to dynamically forward traffic to mobile users. In conclusion, we describe a proof-of-concept prototype developed for further experimental validation of the proposed NOMA multicast routing scheme.

This paper proposes a network-aware resource management scheme that improves the quality of experience (QoE) for adaptive video streaming in CDNs and Information-Centric Networks (ICN) in general, and Dynamic Adaptive Streaming over HTTP (DASH) in particular. By utilizing the DASH manifest, the network (by way of a logically centralized controller) computes the available link resources and schedules the chunk dissemination to edge caches ahead of the end-user's requests. Our approach is optimized for multi-rate DASH videos. We implemented our resource management scheme, and demonstrated that in the scenario when network conditions evolve quickly, our approach can maintain smooth high quality playback. We show on actual video server data and in our own simulation environment that a significant reduction in peak bandwidth of 20% can be achieved using our approach.

Traffic from mobile wireless networks has been growing at a fast pace in recent years and is expected to surpass wired traffic very soon. Service providers face significant challenges at such scales including providing seamless mobility, efficient data delivery, security, and provisioning capacity at the wireless edge. In the Mobility First project, we have been exploring clean slate enhancements to the network protocols that can inherently provide support for at-scale mobility and trustworthiness in the Internet. An extensible data plane using pluggable compute-layer services is a key component of this architecture. We believe these extensions can be used to implement in-network services to enhance mobile end-user experience by either off-loading work and/or traffic from mobile devices, or by enabling en-route service-adaptation through context-awareness (e.g., Knowing contemporary access bandwidth). In this work we present details of the architectural support for in-network services within Mobility First, and propose protocol and service-API extensions to flexibly address these pluggable services from end-points. As a demonstrative example, we implement an in network service that does rate adaptation when delivering video streams to mobile devices that experience variable connection quality. We present details of our deployment and evaluation of the non-IP protocols along with compute-layer extensions on the GENI test bed, where we used a set of programmable nodes across 7 distributed sites to configure a Mobility First network with hosts, routers, and in-network compute services.

Streaming services have grown rapidly in the last few years and providers of video on-demand, such as Netflix or YouTube, are increasing the number of users even more quickly. The majority of these companies implement their services using huge Content Delivery Networks that are as much powerful as expensive, e.g. Amazon and Akamai. In this paper we propose a hybrid CDN/P2P solution that aims at reducing the infrastructural costs exploiting local caching and P2P while guaranteeing an optimal quality of service. The proposed architecture uses a classic CDN complemented by a geographically distributed layer where P2P can be activated exploiting network, content awareness and locality. The performance of the proposed solution is evaluated by means of a prototype implementation that has been deployed using the PlanetLab network and the Amazon AWS cloud services. Our findings show that the proposed approach provides adaptive, flexible, scalable and content centric service to the end users while significantly reducing the infrastructural costs.