Ravi Sundaram

Delay tolerant bulk data transfers on the internet

Many emerging scientific and industrial applications require transferring multiple terabytes of data on a daily basis. Examples include pushing scientific data from particle accelerators/colliders to laboratories around the world, synchronizing datacenters across continents, and replicating collections of high-definition videos from events taking place at different time-zones. A key property of all above applications is their ability to tolerate delivery delays ranging from a few hours to a few days. Such delay-tolerant bulk (DTB) data are currently being serviced mostly by the postal system using hard drives and DVDs, or by expensive dedicated networks. In this paper, we propose transmitting such data through commercial ISPs by taking advantage of already-paid-for off-peak bandwidth resulting from diurnal traffic patterns and percentile pricing. We show that between sender-receiver pairs with small time-zone difference, simple source scheduling policies are able to take advantage of most of the existing off-peak capacity. When the time-zone difference increases, taking advantage of the full capacity requires performing store-and-forward through intermediate storage nodes. We present an extensive evaluation of the two options based on traffic data from 200+ links of a large transit provider with points of presence (PoPs) at three continents. Our results indicate that there exists huge potential for performing multiterabyte transfers on a daily basis at little or no additional cost.

Spanning Trees---Short or Small

We study the problem of finding small trees. Classical network design problems are considered with the additional constraint that only a specified number k of nodes are required to be connected in the solution. A prototypical example is the kMST problem in which we require a tree of minimum weight spanning at least k nodes in an edge-weighted graph. We show that the kMST problem is NP-hard even for points in the Euclidean plane. We provide approximation algorithms with performance ratio 2v/ for the general edge-weighted case and O(k1/4) for the case of points in the plane. Polynomial-time exact solutions are also presented for the class of treewidth-bounded graphs, which includes trees, series-parallel graphs, and bounded bandwidth graphs, and for points on the boundary of a convex region in the Euclidean plane. We also investigate the problem of finding short trees and, more generally, that of finding networks with minimum diameter. A simple technique is used to provide a polynomiM-time solution for finding k-trees of minimum diameter. We identify easy and hard problems arising in finding short networks using a framework due to T. C. Hu.

SPREAD: Foiling Smart Jammers Using Multi-Layer Agility

In this paper, we address the problem of cross-layer denial of service attack in wireless data networks. We introduce SPREAD -a novel adaptive diversification approach to provide resiliency against such attacks. SPREAD relies on a mechanism-hopping technique, which can be seen as a multi-layer extension of the frequency-hopping technique. We apply a game-theoretic framework for modeling the interaction of the communicating nodes and the adversaries and analyze the proposed approach. We reason about the advantages of SPREAD against various types of jammers and demonstrate the effectiveness of our approach in the case of IEEE 802.11 protocol stack by studying the EIFS attack, periodical jamming and a Packet-Size Game. As an example, we show that mechanism-hopping over two instances of IEEE 802.11 can achieve several orders of magnitude gain in throughput over a single-instance network under the EIFS attack.

Universal approximations for TSP, Steiner tree, and set cover

We introduce a notion of universality in the context of optimization problems with partial information. Universality is a framework for dealing with uncertainty by guaranteeing a certain quality of goodness for all possible completions of the partial information set. Universal variants of optimization problems can be defined that are both natural and well-motivated. We consider universal versions of three classical problems: TSP, Steiner Tree and Set Cover.We present a polynomial-time algorithm to find a universal tour on a given metric space over n vertices such that for any subset of the vertices, the sub-tour induced by the subset is within O(log4n/log log n) of an optimal tour for the subset. Similarly, we show that given a metric space over n vertices and a root vertex, we can find a universal spanning tree such that for any subset of vertices containing the root, the sub-tree induced by the subset is within O(log4n/log log n) of an optimal Steiner tree for the subset. Our algorithms rely on a new notion of sparse partitions, that may be of independent interest. For the special case of doubling metrics, which includes both constant-dimensional Euclidean and growth-restricted metrics, our algorithms achieve an O(log n) upper bound. We complement our results for the universal Steiner tree problem with a lower bound of Ω(log n/log log n) that holds even for n vertices on the plane. We also show that a slight generalization of the universal Steiner Tree problem is coNP-hard and present nearly tight upper and lower bounds for a universal version of Set Cover.

Symmetric alternation captures BPP

Abstract. We introduce the natural class

Bounded budget connection (BBC) games or how to make friends and influence people, on a budget

{\bf\,S}^P_2

Bicriteria Network Design Problems

containing those languages that may be expressed in terms of two symmetric quantifiers. This class lies between

Treewidth of Circular-Arc Graphs

\Delta^P_2

Steganographic communication in ordered channels

and

Scheduler vulnerabilities and coordinated attacks in cloud computing

\Sigma^P_2\,\cap\,\Pi^P_2