Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

We will review some of the major results in random graphs and some of the more challenging open problems. We will cover algorithmic and structural questions. We will touch on newer models, including those related to the WWW.

Random graphs

Border Gateway Protocol (BGP) is the protocol backing the core routing decisions on the Internet. It maintains a table of IP networks or 'prefixes' which designate network reachability among autonomous systems (AS). Point of concern in BGP is its lack of effective security measures which makes Internet vulnerable to different forms of attacks. Many solutions have been proposed till date to combat BGP security issues but not a single one is deployable in practical scenario. Any security proposal with optimal solution should offer adequate security functions, performance overhead and deployment cost. This paper critically analyzes the deployment issues of best three proposals considering trade-off between security functions and performance overhead.

Comparative analysis of

Matrix Factorization Techniques for Recommender Systems

This document provides updates to IEEE Std 802.16's MIB for the MAC, PHY and asso- ciated management procedures in order to accommodate recent extensions to the standard.

IEEE Standard for Local and Metropolitan Area Networks Part 16: Air Interface for Fixed Broadband Wireless Access Systems Draft Amendment: Management Information Base Extensions

The most widely used technique for IP geolocation consists in building a database to keep the mapping between IP blocks and a geographic location. Several databases are available and are frequently used by many services and web sites in the Internet. Contrary to widespread belief, geolocation databases are far from being as reliable as they claim. In this paper, we conduct a comparison of several current geolocation databases -both commercial and free- to have an insight of the limitations in their usability.First, the vast majority of entries in the databases refer only to a few popular countries (e.g., U.S.). This creates an imbalance in the representation of countries across the IP blocks of the databases. Second, these entries do not reflect the original allocation of IP blocks, nor BGP announcements. In addition, we quantify the accuracy of geolocation databases on a large European ISP based on ground truth information. This is the first study using a ground truth showing that the overly fine granularity of database entries makes their accuracy worse, not better. Geolocation databases can claim country-level accuracy, but certainly not city-level.

/pdf/ip-geolocation-databases-unreliable-242zm1b2ej.pdf

IP geolocation databases: unreliable?

Since the beginning of the nineties, the internet has undergone impressive growth. This growth can be appreciated in terms of the equipment, such as routers and links, that has been added, as well as in the numbers of users and the value of commerce that it supports. In parallel to this expansion, over the past decade the networking research community has shown a growing interest in discovering and analyzing the internet topology. Some researchers have developed tools for gathering network topology data while others have tried to understand and model the internet?s properties. These efforts have brought us to a crucial juncture for toplogy measurement infrastructures: while, previously, these were both small (in terms of number of measurement points) and monolithic, we are starting to see the deployment of large-scale distributed systems composed of hundreds or thousands of monitors. As we look forward to this next generation of systems, we take stock of what has been achieved so far. In this survey, we discuss past and current mechanisms for discovering the internet topology at various levels: the IP interface, the router, the AS, and the PoP level. In addition to discovery techniques, we provide insights into some of the wellknown properties of the internet topology.

/pdf/internet-topology-discovery-a-survey-59idaxqwit.pdf

Internet topology discovery: a survey

There is a growing interest in discovery of internet topology at the interface level. A new generation of highly distributed measurement systems is currently being deployed. Unfortunately, the research community has not examined the problem of how to perform such measurements efficiently and in a network-friendly manner. In this paper we make two contributions toward that end. First, we show that standard topology discovery methods (e.g., skitter) are quite inefficient, repeatedly probing the same interfaces. This is a concern, because when scaled up, such methods will generate so much traffic that they will begin to resemble DDoS attacks. We measure two kinds of redundancy in probing (intra- and inter-monitor) and show that both kinds are important. We show that straightforward approaches to addressing these two kinds of redundancy must take opposite tacks, and are thus fundamentally in conflict. Our second contribution is to propose and evaluate Doubletree, an algorithm that reduces both types of redundancy simultaneously on routers and end systems. The key ideas are to exploit the tree-like structure of routes to and from a single point in order to guide when to stop probing, and to probe each path by starting near its midpoint. Our results show that Doubletree can reduce both types of measurement load on the network dramatically, while permitting discovery of nearly the same set of nodes and links.

/pdf/efficient-algorithms-for-large-scale-topology-discovery-hs6tztiqyz.pdf

Efficient algorithms for large-scale topology discovery

Middleboxes such as firewalls, NAT, proxies, or Deep Packet Inspection play an increasingly important role in various types of IP networks, including enterprise and cellular networks. Recent studies have shed the light on their impact on real traffic and the complexity of managing them. Network operators and researchers have few tools to understand the impact of those boxes on any path. In this paper, we propose tracebox, an extension to the widely used traceroute tool, that is capable of detecting various types of middlebox interference over almost any path. tracebox sends IP packets containing TCP segments with different TTL values and analyses the packet encapsulated in the returned ICMP messages. Further, as recent routers quote, in the ICMP message, the entire IP packet that they received, tracebox is able to detect any modification performed by upstream middleboxes. In addition, tracebox can often pinpoint the network hop where the middlebox interference occurs. We evaluate tracebox with measurements performed on PlanetLab nodes. Our analysis reveals various types of middleboxes that were not expected on such an experimental testbed supposed to be connected to the Internet without any restriction.

/pdf/revealing-middlebox-interference-with-tracebox-4z2eh6pk8e.pdf

Revealing middlebox interference with tracebox

During the last decade, a new class of large-scale globally-distributed network services and applications have emerged. Those systems are flexible in the sense that they can select their communication path among a set of available ones. However, ceaselessly gathering network information such as latency to select a path is infeasible due to the large amount of measurement traffic it would generate. To overcome this issue, Network Coordinates Systems (NCS) have been proposed. An NCS allows hosts to predict latencies without performing direct measurements and, consequently, reduce the network resources consumption. During these last years, NCS opened new research fields in which the networking community has produced an impressive amount of work. We believe it is now time to stop and take stock of what has been achieved so far. In this paper, we survey the various NCS proposed as well as their intrinsic limits. In particular, we focus on security issues and solutions proposed to fix them. We also discuss potential future NCS developments, in particular how to use NCS for predicting bandwidth.

/pdf/a-survey-on-network-coordinates-systems-design-and-security-ve1m6akgau.pdf

Benoit Donnet

Papers

IP geolocation databases: unreliable?

Internet topology discovery: a survey

Efficient algorithms for large-scale topology discovery

Revealing middlebox interference with tracebox

A Survey on Network Coordinates Systems, Design, and Security