The limited availability of high-quality computer networking data, and the privacy risks of sharing what does exist, has prompted development of ML-based methods for generating synthetic network data that mimics real communication between networked devices. The viability of these models hinges on both the quality of their output and how well they preserve private information encoded in their training data. Prior work has sought to address this by training models with differential privacy (DP). However, how this choice affects the actual privacy of the training data, and subsequently the quality of the generated output, is not well understood. In this work, we analyze the relationship between privacy and quality in generative network data models. Using the success of membership inference attacks (MIAs) as the metric for privacy, we observe that whether DP mitigates MIAs depends heavily on model architecture and representation of network data used for training. In particular, we empirically find that some approaches to generating synthetic network data train models that heavily skew towards either overgeneralizing or undergeneralizing to their training data, resulting in poor or inconsistent MIA performance. In these cases, using DP does not yield substantive improvements in vulnerability to MIAs. As for the quality of generated data, we find that DP synthetic network data can retain statistical similarity to real data even under strict privacy budgets, and that downstream models (e.g., classifiers, regressors) trained on this data tend to achieve at least as good accuracy as models trained on non-DP data. These results suggest that DP, depending on the model, offers protection against MIAs without degrading the utility of the generated output, and in some cases, improves utility.

Autonomous mobility systems increasingly operate in environments shared with animals, from urban pets to wildlife.However, their design has largely focused on human interaction, with limited understanding of how non-human species perceive, respond to, or are affected by these systems.Motivated by research in Animal-Computer Interaction (ACI) and more-than-human design, this study investigates animal interactions with autonomous mobility through a multi-method approach combining a scoping review (45 articles), online ethnography (39 YouTube videos and 11 Reddit discussions), and expert interviews (8 participants).Our analysis surfaces five key areas of concern: Physical Impact (e.g., collisions, failures to detect), Behavioural Effects (e.g., avoidance, stress), Accessibility Concerns (particularly for service animals), Ethics and Regulations, and Urban Disturbance.We conclude with design and policy directions aimed at supporting multispecies coexistence in the age of autonomous systems.This work underscores the importance of incorporating non-human perspectives to ensure safer, more inclusive futures for all species.

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.

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.

While great strides have been taken in advancing the field of Human-Robot Interaction (HRI), challenges abound in understanding and improving how Autonomous Vehicles (AVs) will interact with and within society. Through this paper, the authors attempt to paint the picture of challenges unique to the study and advancement of interfaces between AVs and vulnerable road users (VRUs). In turn, these gaps in research highlight the opportunities for academia, industry, and public policy to collaborate and advance the state of the art of AV-VRU interaction, and the need for a dedicated forum for sharing insights across these various sectors.

The three decade struggle to ensure Internet data confidentiality---a key aspect of communications privacy---is finally behind us. Encryption is fast, secure, and standard in all browsers, modern transports, and major protocols. Yet it has long seemed that network privacy is not unified by core principles but a grab bag of techniques and ideas applied to an equally wide range of applications, contexts, layers of infrastructure, and software stacks.

Machine learning is leveraged for many network traffic analysis tasks in security, from application identification to intrusion detection. Yet, the aspects of the machine learning pipeline that ultimately determine the performance of the model---feature selection and representation, model selection, and parameter tuning---remain manual and painstaking. This paper presents a method to automate many aspects of traffic analysis, making it easier to apply machine learning techniques to a wider variety of traffic analysis tasks. We introduce nPrint, a tool that generates a unified packet representation that is amenable for representation learning and model training. We integrate nPrint with automated machine learning (AutoML), resulting in nPrintML, a public system that largely eliminates feature extraction and model tuning for a wide variety of traffic analysis tasks. We have evaluated nPrintML on eight separate traffic analysis tasks and released nPrint, nPrintML and the corresponding datasets from our evaluation to enable future work to extend these methods.

Recent concerns over the privacy implications of the Domain Name System (DNS) have led to encrypting DNS queries and responses through protocols like DNS-over-HTTPS (DoH) and DNS-over-TLS (DoT). Although the trend towards encryption is a positive development, the accompanying centralization of the DNS has fomented tussles involving ISPs, browser and device vendors, content delivery networks, and users. This paper articulates several current DNS tussles and offers principles to guide system design and implementation such that all stakeholders in the space could participate. We argue that refactoring name resolution in a stub resolver that is separate from devices and applications can preserve the benefits of encrypted DNS while satisfying other architectural desiderata, including performance, resilience, and privacy.

Abstract In this paper, we study the performance of encrypted DNS protocols and conventional DNS from thousands of home networks in the United States, over one month in 2020. We perform these measurements from the homes of 2,693 participating panelists in the Federal Communications Commission’s (FCC) Measuring Broadband America program. We found that clients do not have to trade DNS performance for privacy. For certain resolvers, DoT was able to perform faster than DNS in median response times, even as latency increased. We also found significant variation in DoH performance across recursive resolvers. Based on these results, we recommend that DNS clients (e.g., web browsers) should periodically conduct simple latency and response time measurements to determine which protocol and resolver a client should use. No single DNS protocol nor resolver performed the best for all clients.

Nearly every service on the Internet relies on the Domain Name System (DNS), which translates a human-readable name to an IP address before two endpoints can communicate. Today, DNS traffic is unencrypted, leaving users vulnerable to eavesdropping and tampering. Past work has demonstrated that DNS queries can reveal a user’s browsing history and even what smart devices they are using at home. In response to these privacy concerns, two new protocols have been proposed: DNS-over-HTTPS (DoH) and DNS-over-TLS (DoT). Instead of sending DNS queries and responses in the clear, DoH and DoT establish encrypted connections between users and resolvers. By doing so, these protocols provide privacy and security guarantees that traditional DNS (Do53) lacks.

Today's smartphones are equipped with both Wi-Fi and cellular interfaces, creating usage opportunities for protocols such as Multi-path TCP (MPTCP), which enable devices to use more than one interface concurrently. One of the biggest hurdles in implementing MPTCP is the heterogeneity in performance characteristics that exists across multiple interfaces. This makes the selection of primary interface of paramount importance, as this interface is also used for DNS resolution. In this work, we explore performance and IP reachability over real world networks. Our findings indicate that widespread MPTCP deployment faces significant obstacles. In particular, we perform controlled and real world experiments over multiple paths with differing loss rates and round trip latencies to assess the effect of primary path selection, and the range of issues that arise from selecting the under-performing path. Using results from our experiments, we show how heterogeneous paths can adversely affect MPTCP performance, especially when one path is lossy.

Abstract Virtually every Internet communication typically involves a Domain Name System (DNS) lookup for the destination server that the client wants to communicate with. Operators of DNS recursive resolvers—the machines that receive a client’s query for a domain name and resolve it to a corresponding IP address—can learn significant information about client activity. Past work, for example, indicates that DNS queries reveal information ranging from web browsing activity to the types of devices that a user has in their home. Recognizing the privacy vulnerabilities associated with DNS queries, various third parties have created alternate DNS services that obscure a user’s DNS queries from his or her Internet service provider. Yet, these systems merely transfer trust to a different third party. We argue that no single party ought to be able to associate DNS queries with a client IP address that issues those queries. To this end, we present Oblivious DNS (ODNS), which introduces an additional layer of obfuscation between clients and their queries. To do so, ODNS uses its own authoritative namespace; the authoritative servers for the ODNS namespace act as recursive resolvers for the DNS queries that they receive, but they never see the IP addresses for the clients that initiated these queries. We present an initial deployment of ODNS; our experiments show that ODNS introduces minimal performance overhead, both for individual queries and for web page loads. We design ODNS to be compatible with existing DNS protocols and infrastructure, and we are actively working on an open standard with the IETF.

While cellular networks connect over 3.7 billion people worldwide, their availability and quality is not uniform across regions. Under-provisioned and overloaded networks, as are common in rural or post-disaster areas, lead to poor network performance and a poor-quality user experience. To address this problem, we propose HybridCell: a system that leverages locally-owned small-scale cellular networks to augment the operation of overloaded commercial networks. HybridCell is the first system to allow a user with their existing SIM card and mobile phone to seamlessly switch between commercial and local networks in order to maintain continuous connectivity. HybridCell accomplishes this by identifying poorly-performing networks and taking action to provide seamless cellular connectivity to end users. Using traces from commercial cellular networks collected during our visit to the Za'atari refugee camp in Jordan, we demonstrate HybridCell's capability to detect and act upon commercial network overload, offering an alternate communication channel during times of congestion. We show that even in scenarios where provider networks deny calls due to overload, HybridCell is able to accommodate users and facilitate local calling.

The current paradigm of exclusive spectrum assignment and allocation is creating artificial spectrum scarcity that has a dramatic impact on network performance and user experience. Thus, governments, industry and academia have endeavored to create novel spectrum management mechanisms that allow multi-tiered access. A key component of such an approach is deep understanding of spectrum utilization in time, frequency and space. To address this challenge, we propose AirVIEW, a one-pass, unsupervised spectrum characterization approach for rapid transmitter detection with high tolerance to noise. AirVIEW autonomously learns its parameters and employs wavelet decomposition in order to amplify and reliably detect transmissions at a given time instant. We show that AirVIEW can robustly identify transmitters even when their power is only 5dBm above the noise floor. Furthermore, we demonstrate AirVIEW's ability to inform next-generation Dynamic Spectrum Access by characterizing essential transmitter properties in wideband spectrum measurements from 50MHz to 4.4GHz.

While the networking literature is replete with work on managing and operating networks---from the specifics of protocols to the design of management tools and architectures---there is comparatively little work on planning a network to be rolled out. In part this is because the task of network planning typically falls to carriers (for backbones) and cloud providers (for datacenters), which have the resources and the control to meet their specific needs. Here we consider network planning in situations that are quite different: resource poor and highly constrained.

Ubiquitous cellular coverage is often taken for granted, yet numerous people live outside, or at the fringes, of commercial cellular coverage. Further, natural disasters and human rights violations cause the displacement of millions of people annually worldwide, with many of these people relocating to shelters and camps in areas at or just beyond the margins of existing cellular infrastructure. In this work we design PhoneHome, a system prototype that extends existing cellular coverage to areas with no or damaged cellular infrastructure, or infrastructure that is otherwise poorly performing. We explore the feasibility of PhoneHome and address current limitations and future directions for independently operated, user-extensible cellular infrastructure.

Despite the appearance of uniform availability of mobile services, in many locales granular network analyses reveal the persistence of physical access divides. It stands to reason these divides, similar to those at larger scales, are also reflections of community-level social and economic divides.

While cellular networks connect over 3.7 billion people worldwide, their availability and quality is not uniform across regions. Under-provisioned and overloaded networks lead to poor network performance and an aggravated user experience. To address this problem we propose HybridCell: a system that leverages locally-owned small-scale cellular networks to augment the operation of overloaded commercial networks. HybridCell is the first system to allow a user with their existing SIM card and mobile phone to seamlessly switch between commercial and local networks in order to maintain continuous connectivity. Hybrid-Cell accomplishes this by identifying poorly-performing networks and taking action to provide seamless cellular connectivity to end users. Using traces collected from observing the cellular infrastructure during our visit to the Za'atari refugee camp in Jordan, we demonstrate HybridCell's capability to detect and act upon commercial network overload, offering an alternate communication channel during times of congestion. We show that even in scenarios where provider networks deny calls due to overload, HybridCell is able to accommodate users and facilitate local calling.

Cellular networks are often the first telecommunications infrastructure in developing regions. By studying cellular net- work traffic, researchers gain insight into how technologies can be used to access services critical to further development. In this work, we approach a cellular traffic dataset provided by Orange in Cote d'Ivoire with the goal of identifying distinctions between urban and rural use of cellular infrastructure. We report on a number of interesting differences between urban and rural usage of cellular infrastructure. For instance, 70% of calls that originate in rural areas occur within the vicinity of the same antenna, whereas the same is true for only 23% of calls with urban origin. We are compelled to conclude that development efforts for rural areas might be implemented differently from development efforts in urban areas based on divergent use of current cellular infrastructure.

Studies of user behavior in cellular networks have served as a knowledge base for development of critical applications and services catered to specific user needs. In this paper we examine community persistence in egocentric social graphs extracted from cellular network traces in the Cote d'Ivoire provided by Orange. The goal of our study is to inform mechanisms for improved dissemination of information by identifying subscribers or groups that can serve as information relays. We find that communities that persist in an egocentric network are independent of one another. Thus, multiple information relays can be selected from each independent community, to increase the probability that information will flow to the ego.