Multi-hop 802.11 wireless networks

Abstract:
In this talk, I will discuss the challenges associated with creating a high performance datapath consisting of multiple wLAN hops. Current data rates supported within a single wLAN cell range upto 108 Mbps. In this talk, I will first provide an overview of the 802.11 wireless LAN protocols followed by a discussion of multi-hop wireless networks. The next part of the talk will consist of work-in-progress on increasing data throughput : (a) an architecture for efficiently forwarding packets between successive hops in such multi-hop networks and (b) extending the 802.11 RTS/CTS MAC to enable neighboring transmitters (receivers) to operate simultaneously under certain conditions.

Algorithms for Computing QoS Paths with Restoration

Abstract:
There is a growing interest among service providers to offer their customers new revenue-generating services with Quality of Service (QoS) guarantees e.g., Virtual Private Networks (VPNs). This is facilitated by current efforts to provide resource reservations and explicit path routing, e.g., Multi- Protocol Label Switching (MPLS). A key requirement for such services is that they also be resilient to failures. This goal can be achieved by provisioning primary and restoration paths that satisfy the QoS constraints.
We undertake a comprehensive study of problems related to finding QoS Paths with Restoration. We consider both bottleneck QoS constraints, such as bandwidth, and additive QoS constraints, such as delay and jitter. This is the first study to provide a rigorous solution, with proven guaranties, to the combined problem of computing QoS paths with restoration. It turns out that the widely used approach of disjoint primary and restoration paths is not an optimal strategy. Hence, the proposed algorithms construct a restoration topology, i.e., a set of bridges, each bridge protecting a portion of the primary QoS path. This approach guaranties to find a restoration topology with low cost when one exists.
Joint work with Y. Bejerano, Y. Breitbart, A. Orda, R. Rastogi

Turbo Codes

Abstract:
In this talk, I would present a brief overview of what channel codes are and how the recently discovered channel codes achieve theoretical limits on transmission over a wide variety of channel. I will introduce several new channel codes, most notably turbo codes. The decoding of these new channel codes would be explained as a belief propagation algorithm on a bipartite graph with one set of vertices representing variables and another set of vertices representing the constraints. If the time permits, I will show connections of these with the Tornado codes which are proposed by Digital Fountain to make faster downloads of large files.

An important result of Network Calculus

Abstract:
Network Calculus is a set of recent developments, based on min-plus and max-plus algebra, which provide a deep insight into flow problems encountered in networking. After a short refresher on the basics of network calculus, I review a few of these problems (such as the greedy shaper, the window flow controller, the optimal multimedia smoother) and I show how a simple residuation theorem can be successfully applied in a quite straightforward manner to obtain important results in the context of computer networks (Joint work with J-Y Le Boudec, LCA-ISC-I&C, EPFL).

Photo-to-Grandma Problem: Compression Meets the Network
Complete Talk

Abstract:
A large fraction of the information that flows across today's networks is useful even in a degraded condition. Examples include speech, audio, still images and video. When this information is subject to packet losses or retransmission is impossible due to real-time constraints, superior performance with respect to total transmitted rate, distortion, and delay may sometimes be achieved by adding redundancy to the bit stream rather than repeating lost packets. In multiple description coding, the data is broken into several streams with some redundancy among the streams. When all the streams are received, one can guarantee low distortion at the expense of having a slightly higher bit rate than a system designed purely for compression. On the other hand, when only some of the streams are received, the quality of the reconstruction degrades gracefully, which is very unlikely to happen with a system designed purely for compression.

IP Traceback

Abstract:
The design of the IP protocol makes it difficult to reliably identify the originator of an IP packet. This anonymity can be exploited by an attacker, who wants hide the true origin of the attack by sending attack packets with spoofed IP addresses to a victim. The goal of an IP traceback system is to identify the source of packets sent across the network in an attempt to make attackers accountable. In this talk, I will discuss techniques to enable IP traceback focusing on the following two papers: (1) Practical Network Support for IP Traceback by S. Savage et al. appeared in ACM SIGCOMM 2000 and (2) Hash-based IP Traceback by A. Snoeren et al. appeared in ACM SIGCOMM 2001.

CDN Measurement Studies

Abstract:
In this talk I will present some background on content distribution network architectures followed by a review of some CDN performance measurement results published recently. I will focus, in particular, on a paper from the 2001 SIGCOMM Internet Measurement Workshop, "On the Use and Performance of Content Distribution Networks" which compares relative performance across several commercial CDNs. I will also discuss some unpublished measurement efforts related to CDNs that our department has conducted in collaboration with the IGS e-Technology Center.

A Scalable Content-Addressable Network, Sylvia Ratnasamy (University of California at Berkeley and ACIRI), Paul Francis (Tahoe Networks), Mark Handley (ACIRI), Richard Karp (U.C. Berkeley and ACIRI), Scott Shenker (ACIRI)

Abstract:
Hash tables -- which map ``keys'' onto ``values'' -- are an essential building block in modern software systems. We believe a similar functionality would be equally valuable to large distributed systems. In this paper, we introduce the concept of a Content-Addressable Network (CAN) as a distributed infrastructure that provides hash table-like functionality on Internet-like scales. The CAN is scalable, fault-tolerant and completely self-organizing, and we demonstrate its scalability, robustness and low-latency properties through simulation.