Scalable and Robust Delivery Infrastructures

Flash Dissemination in Heterogeneous Networks

The goal of Flash Dissemination is rapid distribution (over the Internet) of varying amounts of data to a large number of recipients in as short a time period as possible. Given the unpredictability in its need, the unstable nature of networks/systems and the varying characteristics of delivered content, flash-dissemination has different concerns and constraints as compared to traditional broadcast or content delivery systems. First, the underlying protocols must be able to disseminate as fast (or faster) as current highly optimized content delivery systems under normal circumstances. In addition, such protocols must be highly fault-resilient under unstable conditions and adapt rapidly in a constantly changing environment. In this study, we have developed a minimum-state gossip-based protocol called CREW. We implement the protocol in a framework that we call RapID that provides key building blocks for P2P content delivery systems. Extensive experimental results over an emulated Internet testbed show that CREW is faster than current state-of-art dissemination systems (such as BitTorrent, Splitstream and Bullet), while maintaining the stateless fault-tolerance properties of traditional gossip protocols. Further, we have also investigated centralized, global-knowledge protocols for dissemination. These protocols are much faster than localized randomized protocols and may be useful in situations where network and systems are stable and dissemination may be repeated multiple times to the same set of recipients, thus amortizing the knowledge collection and dissemination plan generation costs.

In addition, we have developed a fast and highly reliable flash dissemination protocol called FaReCast. With the notion of path diversity supported by a mesh-like overlay structre having multiple-parents and multiple-children, FaReCast selectively trigger multidirectional backup multicasting to send a message to the nodes which may not get the message. FaReCast achives 100% reliability under very high (over 40%) overlay failure with low latency and reasonable overhead which are comparable to other protocols (SplitStream, MultipleTree-based, Epidemic, and PRM). FaReCast is implemented in JAVA and the distributed configuration manager, which manages the structure of the used overlay with large number of unreliable users, is also implemented by using the FreePastry substrate.

Publications:

FaReCast: Fast, Reliable Application Layer Multicast for Flash Dissemination,
Kyungbaek Kim, Sharad Mehrotra and Nalini Venkatasubramanian,
In proceedings of ACM/IFIP/USENIX 11th International Middleware Conference (Middleware 2010). November 29 - December 3, 2010, Bangalore, India

Heuristics for Flash-Dissemination in Heterogenous Networks
Mayur Deshpande; Nalini Venkatasubramanian; Sharad Mehrotra
International Conference on High Performance Computing (HiPC). 2006.

CREW: A Gossip-based Flash-Dissemination System
Mayur Deshpande ; Bo Xing ; Iosif Lazaridis ; Bijit Hore ; Nalini Venkatasubramanian ; Sharad Mehrotra ;
ICSDCS (International Conference on Distributed Computing Systems), 2006., 2006-07

Scalable, Flash Dissemination of Information in a Heterogeneous Network
Mayur Deshpande; Nalini Venkatasubramanian; Sharad Mehrotra;
Technical Report, 2005

SPINE: Scalable Approximate Global State Maintenance in Peer-to-Peer Networks
Mayur Deshpande; Haimin Lee; Chen Li; S. Mehrotra; Nalini Venkatasubramanian;
Technical Report, 2004

Information Dissemination in Heterogeneous Wireless Environments

Wireless networks (e.g., cellular, Wi-Fi) extend wireline networks in warning and notifying large number of people (both the public and first responders) in crisis situations. People with handheld devices (e.g., cell phones, PDAs) not only receive emergency alerts, but also share warnings and other related information between each other via ad hoc networks. In this work, we study fast, reliable, and efficient dissemination of application-generated data in heterogeneous wireless networks. We consider coverage (the percentage of intended recipients that receive the information), redundancy (the time it takes for people to receive the information) and energy consumption (on handheld devices) as the primary metrics. We develop efficient dissemination strategies that are not only fast and reliable, but also resilient to network congestion and recipients' mobility. We propose protocols that manage the dissemination of data with large size. We also investigate exploiting multiple radio interfaces, hybrid networks as well as mobility for faster dissemination

Publications:

Fast Emergency Information Dissemination over Cellular/Wi-Fi Combined Networks
Jinsu Wang; Bo Xing;
Technical Report, 2006-05

Bandwidth Aggregation for Real –Time Applications in Heterogenerous Wireless Networks
K. Chebrolu; R. Rao;
IEEE Transactions on Wireless Communications, 2005

Energy Efficient Transmission Scheduling for Delay Constrained Wireless Networks
P. Nuggehalli; V. Srinivasan; R. Rao;
IEEE Transactions on Wireless Communications, 2005

Scheduling on a Channel with Time-Varying Capacity
A. Dhamdhere; R. Rao;
Workshop on Cross-Layer Issues in the Design of Tactical Mobile Ad Hoc Wireless Networks: Integration of Communication and Networking Functions to Support Optimal Information Management, Washington, D, 2004-06

Improving Energy Saving in Wireless Systems by Using Dynamic Power Management
Carla-Fabiana Chiasserini; Ramesh Rao;
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 2, NO. 5, 2003-09