Spring 2010

March 31 - Fixing Anti-Virus

John Viega, Perimeter eSecurity

Anti-virus software (AV) is found on most Windows desktops (well over 90%). But many people are amazed at this, because conventional wisdom is that anti-virus technologies absolutely suck. Technical people will often claim that AV "doesn't work" and that it "causes stability problems". And almost everyone will claim that it "slows your machine down". While there is a lot of truth here, there are also a lot of misconceptions about how AV works, and why it fails.

In this talk, I'll give an overview of AV technology, the challenges the industry has faced, and the solutions that the industry has been exploring to address those issues, including virtualization, herd technology, application whitelisting, and reputation systems.

March 25 - Understanding Scam Victims: Seven Principles for Systems Security

Frank Stajano, University of Cambridge, UK

Frank Stajano and Paul Wilson

The success of many attacks on computer systems can be traced back to the security engineers not understanding the psychology of the system users they meant to protect. We examine a variety of scams and short cons that were investigated, documented and recreated for the BBC TV programme The Real Hustle and we extract from them some general principles about the recurring behavioural patterns of victims that hustlers have learnt to exploit. We argue that an understanding of these inherent human factors vulnerabilities, and the necessity to take them into account during design rather than naively shifting the blame onto the gullible users, is a fundamental paradigm shift for the security engineer which, if adopted, will lead to stronger and more resilient systems security.

To appear in Comms ACM.

March 24 - Peer-to-peer Anonymous Communication, or, Relying upon the Kindness of Strangers (with a Little Help from My Friends)

Nikita Borisov, University of Illinois at Urbana-Champaign

Anonymous communication can help reduce the ever-growing digital fingerprint that we are leaving when using online services by disassociating online actions from user identity. The Tor anonymizing network is used by a hundreds of thousands of users, from file sharers to diplomatic agents to citizens of repressive countries. The fundamental approach behind Tor is to use volunteer relays to forward traffic; however, the scarcity of such volunteers has resulted in significant performance bottlenecks for the users. Increasing the number of relays requires better incentives, but also an architectural change in how relays are located. I will focus my presentation on the latter problem.

The current Tor architecture of using a centralized directory cannot scale to much larger network sizes. A decentralized approach, where nodes use other nodes to locate relays is scalable, and several such designs have been previously proposed; however, they turned out to be vulnerable to attacks where malicious nodes misdirect the search for other nodes and/or use the information leaked by the search to anonymize connections. I will discuss ShadowWalker, a system for decentralized anonymous communication that is able to avoid these attacks by using a specialized redundant peer-to-peer topology. Time permitting, I will also discuss Grapevine, our approach to defending from Sybil attacks based on social networks and network-layer DHTs that is effective with minimal state per node.

March 10 - Tradeoffs in Private Search

Binh Vo, PhD student, Columbia University

Encrypted search - performing queries on protected data - has been explored in the past; however, its inherent inefficiency has raised questions of practicality. Here, we step back from the goal of achieving maximal privacy guarantees in an encrypted search scenario and consider efficiency as a priority for the practicality of a scheme. We propose a privacy framework for search that allows tuning and optimization of the tradeoffs between privacy and efficiency, taking into consideration the available resources in different settings. As an instantiation of the privacy framework, we introduce a tunable search system based on the SADS scheme, and provide detailed measurements demonstrating the tradeoffs of the constructed system. We also analyze other existing encrypted search schemes with respect to this framework. We further propose a protocol that addresses the challenge of document retrieval in a search setting with relaxed privacy requirements.

February 24 - The Impact of TLS on SIP Server Performance

Charles Shen, PhD student, Columbia University

We study the performance impact of using TLS as a transport protocol for SIP servers. We evaluate the cost of TLS experimentally using a testbed with OpenSIPS, OpenSSL, and Linux running on an Intel-based server. We analyze TLS costs using application, library, and kernel profiling, and use the profiles to illustrate when and how different costs are incurred, such as bulk data encryption, public key encryption, private key decryption, and MAC-based verification. We show that using TLS can reduce performance by up to a factor of nearly 20 compared to the typical case of SIP over UDP. The primary factor in determining performance is whether and how TLS connection establishment is performed, due to the heavy costs of RSA operations used for session negotiation. This depends both on how the SIP proxy is deployed (e.g., as an inbound or outbound proxy) and what TLS options are used (e.g., mutual authentication, session reuse). The cost of symmetric key operations such as AES or 3DES, in contrast, tends to be small. Network operators deploying SIP over TLS should attempt to maximize the persistence of secure connections, and will need to assess the server resources required. To aid them, we provide a measurement-driven cost model for use in provisioning SIP servers using TLS. Our cost model predicts performance within 15 percent on average.

February 17 - Toward Guarantees of Data Possession for Cloud Storage

Reza Curtmola, New Jersey Institute of Technology

Faced with cost and regulatory considerations, many companies are outsourcing the storage of their data to storage clouds managed by third parties. Outsourcing data storage achieves economies of scale for the management of storage and avoids the large initial investment to set up data centers. Recently, many such online archival systems have emerged from within the research and industrial communities.

In storage outsourcing, a client sends data to a server, which is required by contract to provide persistent archival of the data. Since the server is not trusted and may misbehave, the client typically retains a small piece of metadata which is used to verify the authenticity of the data upon its retrieval. The problem is that by the time data is retrieved, it might be already too late to recover lost or damaged data. Current systems lack a basic guarantee: Proving data possession upon a user's request (usually before data retrieval).

In this presentation, we introduce a model for remote data checking which allows a client that has stored data at an untrusted server to verify that the server possesses the original data. We present provably-secure schemes that have low (or even constant) overhead at the server and and minimize network communication by transmitting a small, constant, amount of data for every challenge/response. The constructs use novel homomorphic verification tags, which allow checking data possession without retrieving the data from the server and without having the server access the entire data. This revolutionizes the ability of users to outsource large data sets by providing a previously-unattainable degree of performance and scalability in verifying the health of external data repositories.

Finally, we extend our data possession framework from a single to multiple storage servers, under various models for data redundancy.

February 3 - What I Did During My Sabbatical: A Survey of VoIP Security Research and Vulnerabilities

Angelos Keromytis, Columbia University Professor

We'll look at the types of bugs that are found in VoIP products and contrast them with the type of research that is being conducted to address VoIP security problems.

January 27 - Network Trace Anonymization

Yingbo Song, PhD student, Columbia University

Good luck! I'm behind seven proxies. An overview of network trace anonymization: history, modern techniques, current directions, and future potential.

January 20 - Ceremony Design and Analysis

Carl Ellison

A ceremony is an extension of the concept of network protocol, including human beings and other physical entities alongside computer nodes. Well designed and analyzed network security protocols frequently fail in the field because of human mistakes or misuse - during setup, administration or normal use. Rather than complain about stupid human users, ceremony analysis takes the approach of including human behavior as part of the design and analysis. It models the entire process as a protocol so that existing protocol design and analysis tools can be applied.