Showing papers by "AT&T Labs published in 2019"

Joint Task Offloading and Resource Allocation for Multi-Server Mobile-Edge Computing Networks

Abstract: Mobile-edge computing (MEC) is an emerging paradigm that provides a capillary distribution of cloud computing capabilities to the edge of the wireless access network, enabling rich services and applications in close proximity to the end users. In this paper, an MEC enabled multi-cell wireless network is considered where each base station (BS) is equipped with a MEC server that assists mobile users in executing computation-intensive tasks via task offloading. The problem of joint task offloading and resource allocation is studied in order to maximize the users’ task offloading gains, which is measured by a weighted sum of reductions in task completion time and energy consumption. The considered problem is formulated as a mixed integer nonlinear program (MINLP) that involves jointly optimizing the task offloading decision, uplink transmission power of mobile users, and computing resource allocation at the MEC servers. Due to the combinatorial nature of this problem, solving for optimal solution is difficult and impractical for a large-scale network. To overcome this drawback, we propose to decompose the original problem into a resource allocation (RA) problem with fixed task offloading decision and a task offloading (TO) problem that optimizes the optimal-value function corresponding to the RA problem. We address the RA problem using convex and quasi-convex optimization techniques, and propose a novel heuristic algorithm to the TO problem that achieves a suboptimal solution in polynomial time. Simulation results show that our algorithm performs closely to the optimal solution and that it significantly improves the users’ offloading utility over traditional approaches.

Fine-grained queue measurement in the data plane

Abstract: Short-lived surges in traffic can cause periods of high queue utilization, leading to packet loss and delay. To diagnose and alleviate performance problems, networks need support for real-time, fine-grained queue measurement. By identifying the flows that contribute significantly to queue build-up directly in the data plane, switches can make targeted decisions to mark, drop, or reroute these flows in real time. However, collecting fine-grained queue statistics is challenging even with modern programmable switch hardware, due to limited memory and processing resources in the data plane. We present ConQuest, a compact data structure that identifies the flows making a significant contribution to the queue. ConQuest operates entirely in the data plane, while working within the hardware constraints of programmable switches. Additionally, we show how to measure queues in legacy devices through link tapping and an off-path switch running ConQuest. Simulations show that ConQuest can identify contributing flows with 90% precision on a 1 ms timescale, using less than 65 KB of memory. Experiments with our Barefoot Tofino prototype show that ConQuest-enabled active queue management reduces flow-completion time.

Toward Practical Volumetric Video Streaming on Commodity Smartphones

Abstract: Volumetric videos offer six degree-of-freedom (DoF) as well as 3D rendering, making them highly immersive, interactive, and expressive. In this paper, we design Nebula, a practical and resource-efficient volumetric video streaming system for commodity mobile devices. Our design leverages edge computing to reduce the computation burden on mobile clients. We also introduce various optimizations to lower the perceived "motion-to-photon" delay, to dynamically adapt to the fluctuating network bandwidth, and to reduce the system's resource consumption while maintaining a high QoE.

Answering range queries under local differential privacy

Abstract: Counting the fraction of a population having an input within a specified interval i.e. a range query, is a fundamental data analysis primitive. Range queries can also be used to compute other core statistics such as quantiles, and to build prediction models. However, frequently the data is subject to privacy concerns when it is drawn from individuals, and relates for example to their financial, health, religious or political status. In this paper, we introduce and analyze methods to support range queries under the local variant of differential privacy [23], an emerging standard for privacy-preserving data analysis.The local model requires that each user releases a noisy view of her private data under a privacy guarantee. While many works address the problem of range queries in the trusted aggregator setting, this problem has not been addressed specifically under untrusted aggregation (local DP) model even though many primitives have been developed recently for estimating a discrete distribution. We describe and analyze two classes of approaches for range queries, based on hierarchical histograms and the Haar wavelet transform. We show that both have strong theoretical accuracy guarantees on variance. In practice, both methods are fast and require minimal computation and communication resources. Our experiments show that the wavelet approach is most accurate in high privacy settings, while the hierarchical approach dominates for weaker privacy requirements.

Mobile VR on edge cloud: a latency-driven design

Abstract: In this paper we design and implement MEC-VR, a mobile VR system that uses a Mobile Edge Cloud (MEC) to deliver high quality VR content to today's mobile devices using 4G/LTE cellular networks. Our main contribution is in realizing a low latency control loop that streams VR scenes containing only the user's Field of View (FoV) and a latency-adaptive margin area around the FoV. This allows the clients to render locally at a high refresh rate to accommodate and compensate for the head movements before the next motion update arrives. Compared with prior approaches, our MEC-VR design requires no viewpoint prediction, supports dynamic and live VR content, and adapts to the real-world latency experienced in cellular networks between the MEC and mobile devices. We implement a prototype of MEC-VR and evaluate its performance on a MEC node connected to an LTE testbed. We demonstrate that MEC-VR can effectively stream live VR content up to 8K resolution over 4G/LTE networks and achieve more than 80% of bandwidth savings.

COSTA: Cost-aware Service Caching and Task Offloading Assignment in Mobile-Edge Computing

Abstract: This paper considers a Mobile-Edge Computing (MEC) enabled wireless network where the MEC-enabled Base Station (MBSs) can host application services and execute computation tasks corresponding to these services when they are offloaded from resource-constrained mobile users. We aim at addressing the joint problem of service caching—the provisioning of application services and their related libraries/database at the MBSs—and task-offloading assignment in a densely-deployed network where each user can exploit the degrees of freedom in offloading different portions of its computation task to multiple nearby MBSs. Firstly, an offloading cost model is introduced to capture the user energy consumption, the service caching cost, and the cloud usage cost. The underlying problem is then formulated as a Mixed-Integer Linear Programming (MILP) problem, which is shown to be NP-hard. Given the intractability of the problem, we exploit local-search techniques to design a polynomial-time iterative algorithm, named COSTA. We prove that COSTA produces a locally optimal solution with cost of at most a constant approximation ratio compared to the optimum. Trace-driven simulations using the workload records from a Google cluster show that COSTA can significantly reduce the offloading cost over competing schemes while achieving a very small optimality gap.

Internet of Things Security - Multilayered Method For End to End Data Communications Over Cellular Networks

Abstract: The aim of this paper is to put forth a multilayered method for securing data transport from a cellular connected Internet of Things device to a host through a cellular network. This method employs many interlocking security elements – described in this paper – that when implemented in their totality provide a highly secure connectivity solution.

Multi-Tier Caching Analysis in CDN-Based Over-the-Top Video Streaming Systems

Abstract: Internet video traffic has been rapidly increasing and is further expected to increase with the emerging 5G applications, such as higher definition videos, the IoT, and augmented/virtual reality applications. As end users consume video in massive amounts and in an increasing number of ways, the content distribution network (CDN) should be efficiently managed to improve the system efficiency. The streaming service can include multiple caching tiers, at the distributed servers and the edge routers, and efficient content management at these locations affects the quality of experience (QoE) of the end users. In this paper, we propose a model for video streaming systems, typically composed of a centralized origin server, several CDN sites, and edge-caches located closer to the end user. We comprehensively consider different systems design factors, including the limited caching space at the CDN sites, allocation of CDN for a video request, choice of different ports (or paths) from the CDN and the central storage, bandwidth allocation, the edge-cache capacity, and the caching policy. We focus on minimizing a performance metric, stall duration tail probability (SDTP), and present a novel and efficient algorithm accounting for the multiple design flexibilities. The theoretical bounds with respect to the SDTP metric are also analyzed and presented. The implementation of a virtualized cloud system managed by Openstack demonstrates that the proposed algorithms can significantly improve the SDTP metric compared with the baseline strategies.

Freedom: Fast Recovery Enhanced VR Delivery Over Mobile Networks

Abstract: In this paper we design and implement Freedom, a mobile VR system that deliver high quality VR content on today's mobile devices using 4G/LTE cellular networks. Compared to existing state-of-the-art, Freedom does not rely on any video frame pre- rendering or viewpoint prediction. We send a latency-adaptive VAM frame that contains pixels around the FoV. This allows the clients to render locally at a high refresh rate of 60 Hz to accommodate and compensate for the user's head movements before the next server update arrives. We demonstrate that Freedom is the first system in the world that can support dynamic and live 8K resolution VR content, while adapting to the real-world latency variations experienced in cellular networks. Compared to streaming the whole 360° panoramic VR content, we show that Freedom achieves up to 80% bandwidth savings. Finally, we provide detailed end to end latency measurements of actual VR systems by running extensive experiments in a private LTE testbed using a Mobile Edge Cloud (MEC).

LIME: understanding commercial 360° live video streaming services

Abstract: Personalized live video streaming is an increasingly popular technology that allows a broadcaster to share videos in real time with worldwide viewers. Compared to video-on-demand (VOD) streaming, experimenting with personalized live video streaming is harder due to its intrinsic live nature, the need for worldwide viewers, and a more complex data collection pipeline. In this paper, we make several contributions to both experimenting with and understanding today's commercial live video streaming services. First, we develop LIME (Live video MEasurement platform), a generic and holistic system allowing researchers to conduct crowd-sourced measurements on both commercial and experimental live streaming platforms. Second, we use LIME to perform, to the best of our knowledge, a first study of personalized 360° live video streaming on two commercial platforms, YouTube and Facebook. During a 7-day study, we have collected a dataset from 548 paid Amazon Mechanical Turk viewers from 35 countries who have watched more than 4,000 minutes of 360° live videos. Using this unique dataset, we characterize 360° live video streaming performance in the wild. Third, we conduct controlled experiments through LIME to shed light on how to make 360° live streaming (more) adaptive in the presence of challenging network conditions.

Minimizing flowtime in a flowshop scheduling problem with a biased random-key genetic algorithm

Abstract: In this paper, we advance the state of the art for solving the Permutation Flowshop Scheduling Problem with total flowtime minimization. For this purpose, we propose a Biased Random-Key Genetic Algorithm (BRKGA) introducing on it a new feature called shaking. With the shaking, instead to full reset the population to escape from local optima, the shaking procedure perturbs all individuals from the elite set and resets the remaining population. We compare results for the standard and the shaking BRKGA with results from the Iterated Greedy Search, the Iterated Local Search, and a commercial mixed integer programming solver, in 120 traditional instances. For all algorithms, we use warm start solutions produced by the state-of-the-art Beam-Search procedure. Computational experiments show the efficiency of proposed BRKGA, in addition to identify lower and upper bounds, as well as some optimal values, among the solutions.

Quality-aware strategies for optimizing ABR video streaming QoE and reducing data usage

Abstract: Streaming videos over cellular networks is highly challenging. Since cellular data is a relatively scarce resource, many video and network providers offer options for users to exercise control over the amount of data consumed by video streaming. Our study shows that existing data saving practices for Adaptive Bitrate (ABR) videos are suboptimal: they often lead to highly variable video quality and do not make the most effective use of the network bandwidth. We identify underlying causes for this and propose two novel approaches to achieve better tradeoffs between video quality and data usage. The first approach is Chunk-Based Filtering (CBF), which can be retrofitted to any existing ABR scheme. The second approach is QUality-Aware Data-efficient streaming (QUAD), a holistic rate adaptation algorithm that is designed ground up. We implement and integrate our solutions into two video player platforms (dash.js and ExoPlayer), and conduct thorough evaluations over emulated/commercial cellular networks using real videos. Our evaluations demonstrate that compared to the state of the art, the two proposed schemes achieve consistent video quality that is much closer to the user-specified target, lead to far more efficient data usage, and incur lower stalls.

Interpreting deep learning models for entity resolution: an experience report using LIME

Abstract: Entity Resolution (ER) seeks to understand which records refer to the same entity (e.g., matching products sold on multiple websites). The sheer number of ways humans represent and misrepresent information about real-world entities makes ER a challenging problem. Deep Learning (DL) has provided impressive results in the field of natural language processing, thus recent works started exploring DL approaches to the ER problem, with encouraging results. However, we are still far from understanding why and when these approaches work in the ER setting. We are developing a methodology, Mojito, to produce explainable interpretations of the output of DL models for the ER task. Our methodology is based on LIME, a popular tool for producing prediction explanations for generic classification tasks. In this paper we report our first experiences in interpreting recent DL models for the ER task. Our results demonstrate the importance of explanations in the DL space, and suggest that, when assessing performance of DL algorithms for ER, accuracy alone may not be sufficient to demonstrate generality and reproducibility in a production environment.

Using Session Modeling to Estimate HTTP-Based Video QoE Metrics From Encrypted Network Traffic

Abstract: Understanding the user-perceived quality of experience (QoE) of HTTP-based video has become critical for content providers, distributors, and network operators. For network operators, monitoring QoE is challenging due to lack of access to video streaming applications, user devices, or servers. Thus, network operators need to rely on the network traffic to infer key metrics that influence video QoE. Furthermore, with content providers increasingly encrypting the network traffic, the task of QoE inference from passive measurements has become even more challenging. In this paper, we present a methodology called eMIMIC that uses passive network measurements to estimate key video QoE metrics for encrypted HTTP-based adaptive streaming (HAS) sessions. eMIMIC uses packet headers from network traffic to model an HAS session and estimate video QoE metrics, such as average bitrate and re-buffering ratio. We evaluate our methodology using network traces from a variety of realistic conditions and ground truth collected using a lab testbed for video sessions from three popular services, two video on demand (VoD) and one Live. eMIMIC estimates re-buffering ratio within 1% point of ground truth for up to 75% sessions in VoD (80% in Live) and average bitrate with error under 100 Kb/s for up to 80% sessions in VoD (70% in Live). We also compare eMIMIC with recently proposed machine learning-based QoE estimation methodology. We show that eMIMIC can predict average bitrate with 2.8%–3.2% higher accuracy and re-buffering ratio with 9.8%–24.8% higher accuracy without requiring any training on ground truth QoE metrics. Finally, we show that eMIMIC can estimate real-time QoE metrics with at least 89.6% accuracy in identifying buffer occupancy state and at least 85.7% accuracy in identifying average bitrate class of recently downloaded chunks.

Empowering video players in cellular: throughput prediction from radio network measurements

Abstract: Today's HTTP adaptive streaming applications are designed to provide high levels of Quality of Experience (QoE) across a wide range of network conditions. The adaptation logic in these applications typically needs an estimate of the future network bandwidth for quality decisions. This estimation, however, is challenging in cellular networks because of the inherent variability of bandwidth and latency due to factors like signal fading, variable load, and user mobility. In this paper, we exploit machine learning (ML) techniques on a range of radio channel metrics and throughput measurements from a commercial cellular network to improve the estimation accuracy and hence, streaming quality. We propose a novel summarization approach for input raw data samples. This approach reduces the 90th percentile of absolute prediction error from 54% to 13%. We evaluate our prediction engine in a trace-driven controlled lab environment using a popular Android video player (ExoPlayer) running on a stock mobile device and also validate it in the commercial cellular network. Our results show that the three tested adaptation algorithms register improvement across all QoE metrics when using prediction, with stall reduction up to 85% and bitrate switching reduction up to 40%, while maintaining or improving video quality. Finally, prediction improves the video QoE score by up to 33%.

Mobile Immersive Computing: Research Challenges and the Road Ahead

Abstract: By its name, immersive computing creates a sense of immersion by blurring the boundary between the digital and the physical worlds. Typical immersive computing technologies include VR, 360-degree videos, AR, MR, and so on. This article discusses and highlights several research challenges when consumers experience immersive media content on mobile devices. It focuses on the recent developments for improving network efficiency and quality of user experience for mobile immersive applications. It then briefly introduces related standardization activities. Finally, it points out several opportunities and directions for further innovation, such as edge-accelerated AR applications.

Analyzing viewport prediction under different VR interactions

Abstract: In this paper, we study the problem of predicting a user's viewport movement in a networked VR system (i.e., predicting which direction the viewer will look at shortly). This critical knowledge will guide the VR system through making judicious content fetching decisions, leading to efficient network bandwidth utilization (e.g., up to 35% on LTE networks as demonstrated by our previous work) and improved Quality of Experience (QoE). For this study, we collect viewport trajectory traces from 275 users who have watched popular 360° panoramic videos for a total duration of 156 hours. Leveraging our unique datasets, we compare viewport movement patterns of different interaction modes: wearing a head-mounted device, tilting a smartphone, and dragging the mouse on a PC. We then apply diverse machine learning algorithms - from simple regression to sophisticated deep learning that leverages crowd-sourced data - to analyze the performance of viewport prediction. We find that the deep learning approach is robust for all interaction modes and yields supreme performance, especially when the viewport is more challenging to predict, e.g., for a longer prediction window, or with a more dynamic movement. Overall, our analysis provides key insights on how to intelligently perform viewport prediction in networked VR systems.

EF-Dedup: Enabling Collaborative Data Deduplication at the Network Edge

Abstract: The advent of IoT and edge computing will lead to massive amounts of data that need to be collected and transmitted to online storage systems. To address this problem, we push data deduplication to the network edge. Specifically, we propose a new technique for collaborative edge-facilitated deduplication (EF-dedup), wherein we partition the resource-constrained edge nodes into disjoint clusters, maintain a deduplication index structure for each cluster using a distributed key-value store and perform decentralized deduplication within those clusters. This is a challenging partitioning problem that addresses a novel tradeoff: edge nodes with highly correlated data may not always be within the same edge cloud, with non-trivial network cost among them. We address this challenge by first formulating an optimization problem to partition the edge nodes, considering both the data similarities across the nodes and the inter-node network cost. We prove that the problem is NP-Hard, provide bounded heuristics to solve it and build a prototype EF-dedup system. Our experiments on EF-dedup, performed on edge nodes in AT&T research lab and a central cloud at AWS, demonstrate that EF-dedup achieves 38.3-118.5% better deduplication throughput than sole cloud-based techniques and achieves 43.4-60.2% lesser aggregate cost in terms of the network-storage tradeoff as compared to approaches that solely favor one over the other.

Informative Summarization of Numeric Data

Abstract: We consider the following data summarization problem. We are given a dataset including ordinal or numeric explanatory attributes and an outcome attribute. We want to produce a summary of how the explanatory attributes affect the outcome attribute. The summary must be human-interpretable, concise, and informative in the sense that it can accurately approximate the distribution of the outcome attribute. We propose a solution that addresses the fundamental challenge of this problem--handling large numeric domains--and we experimentally show the effectiveness and efficiency of our approach on real datasets.

Energy-Latency-Aware Task Offloading and Approximate Computing at the Mobile Edge

Abstract: Task offloading with Mobile-Edge Computing (MEC) is envisioned as a promising technique for prolonging battery lifetime and enhancing the computation capacity of mobile devices. In this paper, we consider a multi-user MEC system with a Base Station (BS) equipped with a computation server assisting mobile users in executing computation-intensive real-time tasks via offloading technique. We formulate the Energy-Latency-aware Task Offloading and Approximate Computing (ETORS) problem, which aims at optimizing the trade-off between energy consumption and application completion time. Due to the centralized and mixed-integer natures of this problem, it is very challenging to derive the optimal solution in practical time. This motivates us to employ the Dual-Decomposition Method (DDM) to decompose the original problem into three subproblems—namely the Task-Offloading Decision (TOD), the CPU Frequency Scaling (CFS), and the Quality of Computation Control (QoCC). Our approach consists of two iterative layers: in the outer layer, we adopt the duality technique to find the optimal value of Lagrangian multiplier associated prime problem; and in the inner layer, we formulate the subproblems that can be solved efficiently using convex optimization techniques. We show that the computation offloading selection depends not only on the computing workload of a task, but also on the maximum completion time of its immediate predecessors and on the clock frequency as well as on the transmission power of the mobile device. Simulation results coupled with real-time experiments on a small-scale MEC testbed show the effectiveness of our proposed resource allocation scheme and its advantages over existing approaches.

Multi-tier Caching Analysis in CDN-based Over-the-top Video Streaming Systems

Abstract: Internet video traffic has been been rapidly increasing and is further expected to increase with the emerging 5G applications such as higher definition videos, IoT and augmented/virtual reality applications. As end-users consume video in massive amounts and in an increasing number of ways, the content distribution network (CDN) should be efficiently managed to improve the system efficiency. The streaming service can include multiple caching tiers, at the distributed servers and the edge routers, and efficient content management at these locations affect the quality of experience (QoE) of the end users. In this paper, we propose a model for video streaming systems, typically composed of a centralized origin server, several CDN sites, and edge-caches located closer to the end user. We comprehensively consider different systems design factors including the limited caching space at the CDN sites, allocation of CDN for a video request, choice of different ports (or paths) from the CDN and the central storage, bandwidth allocation, the edge-cache capacity, and the caching policy. We focus on minimizing a performance metric, stall duration tail probability (SDTP), and present a novel and efficient algorithm accounting for the multiple design flexibilities. The theoretical bounds with respect to the SDTP metric are also analyzed and presented. The implementation on a virtualized cloud system managed by Openstack demonstrate that the proposed algorithms can significantly improve the SDTP metric, compared to the baseline strategies.

Speaking with Actions - Learning Customer Journey Behavior

Abstract: To provide intelligent care, effortless experience and promote customer loyalty, it is essential that companies understand customer behavior and predict customer needs. Customers “speak” to companies through a sequence of interactions across different care channels. Companies can benefit from listening to this speech. We use the term customer journey to refer to the aggregated sequence of interactions that a customer has with a company. Most existing research focuses on data visualization, descriptive analysis, and obtaining managerial hints from studying customer journeys. In contrast, the goal of this paper is to predict future customer interactions within a certain period based on omni-channel journey data. To this end, we introduce a new abstract concept called “action” to describe customers' daily behavior. Using LSTM and DNN, we propose a systematic two-step framework based on omni-channel care journey data and customer profile data. The framework enables us to perform “action embedding”, which learns vector representations of actions. Our framework predicts whether or not a customer will contact in the time period directly following the recent contacts. Comparing the performance on large-scale real datasets to other machine learning techniques such as logistic regression and random forest, our approach yields superior results. In addition, we further cluster the action embedding learned by our model and investigate the intrinsic properties of customer behavior.

Directional Frame Timing Synchronization in Wideband Millimeter-Wave Systems With Low-Resolution ADCs

Abstract: In this paper, we propose and evaluate a novel beamforming strategy for directional frame timing synchronization in wideband millimeter-wave (mmWave) systems operating with low-resolution analog-to-digital converters (ADCs). In the employed system model, we assume multiple radio frequency chains equipped at the base station to simultaneously form multiple synchronization beams in the analog domain. We formulate the corresponding directional frame timing synchronization problem as a max-min multicast beamforming problem under low-resolution quantization. We first show that the formulated problem cannot be effectively solved by conventional single-stream beamforming based approaches due to large quantization loss and limited beam codebook resolution. We then develop a new multi-beam probing based directional synchronization strategy, targeting at maximizing the minimum received synchronization signal-to-quantization-plus-noise ratio (SQNR) among all users. Leveraging a common synchronization signal structure design, the proposed approach synthesizes an effective composite beam from the simultaneously probed beams to better trade off the beamforming gain and the quantization distortion. Numerical results reveal that for wideband mmWave systems with low-resolution ADCs, the timing synchronization performance of our proposed method outperforms the existing approaches due to the improvement in the received synchronization SQNR.

Differentiated Latency in Data Center Networks with Erasure Coded Files Through Traffic Engineering

Abstract: This paper proposes an algorithm to minimize weighted service latency for different classes of tenants (or service classes) in a data center network where erasure-coded files are stored on distributed disks/racks and access requests are scattered across the network. Due to the limited bandwidth available at both top-of-the-rack and aggregation switches, and differentiated service requirements of the tenants, network bandwidth must be apportioned among different intra- and inter-rack data flows for different service classes in line with their traffic statistics. We formulate this problem as weighted queuing and employ a class of probabilistic request scheduling policies to derive a closed-form upper-bound of service latency for erasure-coded storage with arbitrary file access patterns and service time distributions. The result enables us to propose a joint weighted latency (over different service classes) optimization over three entangled “control knobs”: the bandwidth allocation at top-of-the-rack and aggregation switches for different service classes, dynamic scheduling of file requests, and the placement of encoded file chunks (i.e., data locality). The joint optimization is shown to be a mixed-integer problem. We develop an iterative algorithm which decouples and solves the joint optimization as 3 sub-problems, which are either convex or solvable via bipartite matching in polynomial time. The proposed algorithm is prototyped in an open-source, distributed file system, Tahoe, and evaluated on a cloud testbed with 16 separate physical hosts in an OpenStack cluster using Cisco switches. Experiments validate our theoretical latency analysis and show significant latency reduction for diverse file access patterns. The results provide valuable insights on designing low-latency data center networks with erasure coded storage.

AViC: a cache for adaptive bitrate video

Abstract: Video dominates Internet traffic today. Users retrieve on-demand video from Content Delivery Networks (CDNs) which cache video chunks at front-ends. In this paper, we describe AViC, a caching algorithm that leverages properties of video delivery, such as request predictability and the presence of highly unpopular chunks. AViC's eviction policy exploits request predictability to estimate a chunk's future request time and evict the chunk with the furthest future request time. Its admission control policy uses a classifier to predict singletons --- chunks evicted before a second reference. Using real world CDN traces from a commercial video service, we show that AViC outperforms a range of algorithm including LRU, GDSF, AdaptSize and LHD. In particular LRU requires up to 3.5× the cache size to match AViC's performance. Further, AViC has low time complexity and has memory complexity comparable to GDSF.

Robust network design for IP/optical backbones

Abstract: Recently, Internet service providers (ISPs) have gained increased flexibility in how they configure their in-ground optical fiber into an IP network. This greater control has been made possible by improvements in optical switching technology, along with advances in software control. Traditionally, at network design time, each IP link was assigned a fixed optical path and bandwidth. Now modern colorless and directionless reconfigurable optical add/drop multiplexers (CD ROADMs) allow a remote controller to remap the IP topology to the optical underlay on the fly. Consequently, ISPs face new opportunities and challenges in the design and operation of their backbone networks [IEEE Commun. Mag.54, 129 (2016); presentation at the International Conference on Computing, Networking, and Communications, 2017; J. Opt. Commun. Netw.10, D52 (2018); Optical Fiber Communication Conference and Exposition (2018), paper Tu3H.2]. Specifically, ISPs must determine how best to design their networks to take advantage of new capabilities; they need an automated way to generate the least expensive network design that still delivers all offered traffic, even in the presence of equipment failures. This problem is difficult because of the physical constraints governing the placement of optical regenerators, a piece of optical equipment necessary to maintain an optical signal over long stretches of fiber. As a solution, we present an integer linear program (ILP) that does three specific things: It solves the equipment placement problem in network design; determines the optimal mapping of IP links to the optical infrastructure for any given failure scenario; and determines how best to route the offered traffic over the IP topology. To scale to larger networks, we also describe an efficient heuristic that finds nearly optimal network designs in a fraction of the time. Further, in our experiments our ILP offers cost savings of up to 29% compared to traditional network design techniques.

Measurements As First-class Artifacts

Abstract: The emergence of programmable switches has sparked a significant amount of work on new techniques to perform more powerful measurement tasks, for instance, to obtain fine-grained traffic and performance statistics. Previous work has focused on the efficiency of these measurements alone and has neglected flexibility, resulting in solutions that are hard to reuse or repurpose and that often overlap in functionality or goals.In this paper, we propose the use of a set of reusable primitive building blocks that can be composed to express measurement tasks in a concise and simple way. We describe the rationale for the design of our primitives, that we have named MAFIA (Measurements As FIrst-class Artifacts), and using several examples we illustrate how they can be combined to realize a comprehensive range of network measurement tasks. Writing MAFIA code does not require expert knowledge of low-level switch architecture details. Using a prototype implementation of MAFIA, we demonstrate the applicability of our approach and show that the use of our primitives results in compiled code that is comparable in size and resource usage with manually written specialized P4 code, and can be run in current hardware.

Height and Facet Extraction from LiDAR Point Cloud for Automatic Creation of 3D Building Models

Abstract: Three-dimensional models of buildings have a variety of applications, e.g., in urban planning, for making decision where to locate power lines, solar panels, cellular antennas, etc. Often, 3D models are created from a LiDAR point cloud, however, this presents three challenges. First, to generate maps at a nationwide scale or even for a large city, it is essential to effectively store and process the data. Second, there is a need to produce a compact representation of the result, to avoid representing each building as thousands of points. Third, it is often required to seamlessly integrate computed models with non-geospatial features of the geospatial entities. In this paper, we demonstrate an end-to-end automation of a large-scale 3D-model creation for buildings. The tool compacts the point cloud and allows to effortlessly integrate the results with information stored in a database. The main motivation for our tool is 5G network planning, where antenna locations require careful consideration, given that buildings and trees could obstruct or reflect high-frequency cellular transmissions.