Fall 2012 - CATT Graduate Research Seminar VIII


CATT Graduate Research Seminar VIII
Wednesday, Nov. 28th, 2012, 12:00 pm – 1:00 pm
Dibner Building LC400

Speaker 1: Ivan Lam                                                                [Advisor: Prof. Jonathan Chao]

“Automating Dynamic Provisioning of Tenant-specific Network Appliances in a Multi-tenant Cloud” (Download Presentation)

Enterprises heavily rely on middleboxes for a variety of functions, e.g., security, policy, and regulatory compliance. In a cloud environment, the ideal model would be a tenant requests for a set of functions and the cloud operator performs dynamic deployment and capacity adjustment based on the traffic volume. However, the current cloud platform has insufficient support for the above and relies on ad-hoc approaches that require human intervention. This talk discusses a framework for supporting elastic network services in a cloud computing environment.

Our proposed Hybrid Security Architecture (HSA) introduces a security layer that decouples the two parts of the problem: high-level network service specifications by the tenants, and the operator's provisioning, placement and traversal enforcement of middlebox instances that provide network services. Specifically, tenants specify virtual appliances as a sequence of virtually any middlebox types according to their needs. HSA provides an autonomous demand-aware VA provisioning platform

that automates the dynamic provisioning and placement of middlebox instances. Resilience and visibility of network services are provided by a distributed middlebox monitoring and traversal enforcement system. Failures are promptly detected and traffic quickly rerouted. Our automated framework with self-serviced specification allows the operator to support a wide range of network services for a large number of tenants concurrently and autonomously.

A prototype is implemented using a small-scale Linux-based testbed. We show that our framework is feasible with prompt response to sudden surges in demand, and failure detection and recovery. Through emulation we also show that our framework is scalable to large-scale deployment.

Speaker 2: Xiwang Yang                                              [Advisor: Prof. Yong Liu]

“Circle-based Recommendation in Online Social Networks” (Download Presentation)

Online social network information promises to increase recommendation accuracy beyond the capabilities of purely rating/feedback-driven recommender systems (RS). As to better serve users’ activities across different domains, many online social networks now support a new feature of “Friends Circles”, which refines the domain-oblivious “Friends” concept. RS should also benefit from domain-specific “Trust Circles”. Intuitively, a user may trust different subsets of friends regarding different domains. Unfortunately, in most existing multi-category rating datasets, a user’s social connections from all categories are mixed together. This paper presents an effort to develop circle-based RS. We focus on inferring category-specific social trust circles from available rating data combined with social network data. We outline several variants of weighting friends within circles based on their inferred expertise levels. Through experiments on publicly available data, we demonstrate that the proposed circle-based recommendation models can better utilize user’s social trust information, resulting in increased recommendation accuracy.