Sudeept Bhatnagar

ReInForM: reliable information forwarding using multiple paths in sensor networks

Sensor networks are meant for sensing and disseminating information about the environment they sense. The criticality of a sensed phenomenon determines its importance to the end user. Hence, data dissemination in a sensor network should be information aware. Such information awareness is essential firstly to disseminate critical information more reliably and secondly to consume network resources proportional to the criticality of information. In this paper, we describe a protocol called RelnForM to deliver packets at desired reliability at a proportionate communication cost. RelnForm sends multiple copies of each packet along multiple paths from source to sink, such that data is delivered at the desired reliability. It uses the concept of dynamic packet state in context of sensor networks, to control the number of paths required for the desired reliability, and does so using only local knowledge of channel error rates and topology. We show that for uniform unit disk graphs, the number of edge-disjoint paths between nodes is equal to the average node degree with very high probability. RelnForm utilizes this property in its randomized forwarding mechanism which results in use of all possible paths and efficient load balancing.

Information assurance in sensor networks

Sensor networks are deployed to monitor the surroundings and keep the end-user informed about the events witnessed. Different types of events have different levels of importance for the user. Information Assurance is an ability to disseminate different information at different assurance levels to the end-user. The assurance level is determined by the criticality of the sensed phenomenon. Thus, information assurance capability allows a sensor network to deliver critical information with high assurance albeit potentially at a higher cost, while saving energy by delivering less important information at a lower assurance level.In this paper, we look at the problem of efficient information assurance in sensor networks when the assurance level of information is defined as the probability of information delivery (desired reliability) to the sink. We propose a new scheme for information delivery at a desired reliability using hop-by-hop broadcast. We show how the wireless broadcast can be utilized to increase the packet delivery rate at each hop and attain a desired reliability at minimal cost. Finally, we derive the optimal strategy for allocation of desired reliabilities at each hop in order to attain any given desired end-to-end reliability.

Multi-resolution state retrieval in sensor networks

Large-scale dense sensor networks require mechanisms to extract topology information that can be used for various aspects of sensor network management. It is critical for any topology discovery algorithm in dense networks not only to adhere to the resource constraints of bandwidth and energy but also to provide several views of the network. Due to factors of density, redundancy and failures it may not be possible or practical to get a complete view of the topology. We describe a distributed parameterized algorithm for Sensor Topology Retrieval at Multiple Resolutions (STREAM), which makes a tradeoff between topology details and resources expended. The algorithm retrieves network state at multiple resolutions at a proportionate communication cost. We also define various classes of topology queries and show how the parameters in the algorithm can be used to support queries specific to sensor networks. We show that topology determined at different resolutions is sufficient for approximating different network properties. We also show that STREAM can be used for general-purpose multi-resolution information retrieval in sensor networks.

Fast replication in content distribution overlays

We present SPIDER-a system for fast replication or distribution of large content from a single source to multiple sites interconnected over Internet or via a private network. In order to exploit spatial diversity of the underlying network, SPIDER uses an overlay structure composed of dedicated transit nodes (TNs). The data transport mechanism in SPIDER leverages this overlay structure to provide a coordinated approach that minimizes the maximum time to replicate to all destination sites (the make span of content replication). In order to achieve this objective, SPIDER employs two orthogonal components: a) creation of multiple dynamic distribution trees using the transit nodes b) end-to-end reliable data transport with flow control on these trees by chaining point-to-point TCPs. We further present simulations based results to quantify benefits of tree construction algorithms in random topologies. We evaluate the real implementation of the SPIDER in Planet Lab and observe a 2-6 times speed up compared to different existing schemes.

STREAM: Sensor Topology Retrieval at Multiple Resolutions

Large-scale sensor networks need energy-efficient mechanisms to extract topology for various aspects of sensor network management. Some network properties can be inferred from a relatively low-resolution representation of topology. Different topology resolutions suffice for different management applications to perform at a desired level. In these cases, it is an overkill to retrieve the entire topology of large-scale networks particularly because sensor nodes are energy constrained. In this paper, we describe a distributed parameterized algorithm for Sensor Topology Retrieval at Multiple Resolutions (STREAM), which makes a tradeoff between topology details and resources expended. We also define various classes of topology queries and rules for optimal parameter selection to support these queries.

Distributed admission control to support guaranteed services in core-stateless networks

The core-stateless service architecture alleviates the scalability problems of the integrated service framework while maintaining its guaranteed service semantics. The admission control methods proposed to support core-stateless guaranteed services have significant drawbacks: We propose a scalable and robust distributed admission control architecture to support core-stateless guaranteed services. Our architecture maintains high network utilization while ensuring that resources are not over-allocated. In our architecture, admission control is performed at the ingress edge routers of a request on an edge-to-edge path basis. A token-passing mechanism is used as the resource management framework. The token helps in dynamic and fair division of bandwidth and allows completely distributed resource allocation on a link unless it is close to saturation. The edge routers co-operate to provide fault tolerance effectively acting as a resilient overlay network. Our admission control framework can support statistical guarantees and diffserv architectures premium service as well. The resource management part of our architecture is well-suited to aid QoS routing algorithms. Analytical and simulation results are presented to show the effectiveness of our architecture.

An Interference-Aware Channel Assignment Scheme for Wireless Mesh Networks

Multichannel communication in a wireless mesh network with routers having multiple radio interfaces significantly enhances the network capacity. Efficient channel assignment and routing is critical for realization of optimal throughput in such networks. In this paper, we investigate the problem of finding the largest number of links that can be activated simultaneously in a wireless mesh network subject to interference, radio and connectivity constraints. Our goal is to activate all such links and we present an interference aware channel assignment algorithm that realizes this goal. We show that the Link Interference Graph created by utilizing a frequently used interference model gives rise to a special class of graphs, known as overlapping double-disk (ODD) graphs. We prove that the Maximum Independent Set computation problem is NP-complete for this special class of graphs. We provide a Polynomial Time Approximation Scheme (PTAS) for computation of the Maximum Independent Set of an ODD graph. We use this PTAS to develop a channel assignment algorithm for a multiradio multichannel Wireless Mesh Network. We evaluate the performance of our channel assignment algorithm by comparing it with the optimal solution obtained by solving an integer linear program. Experimental results demonstrate that our channel assignment algorithm produces near optimal solution in almost all instances of the problem.

Multi-path adaptive optical burst forwarding

Reducing burst contention is an inherently challenging problem in optical burst switched (OBS) networks. We present a new proactive approach for tackling this problem; the approach is based on adaptive use of multiple paths between edge nodes. The proposed adaptive framework is entirely edge-assisted, distributed, and scalable; it can be employed with any OBS control architecture. The framework is based on (1) dynamically finding, among multiple paths for forwarding bursts, the optimal path that yields a global low burst drop probability throughout the network and (2) pipelined burst delivery for seamless path shifting and also to avoid out-of-order packet delivery. Simulation studies on various network and load scenarios demonstrate a significant reduction of burst contention compared to the existing single path non-adaptive case.

Creating multipoint-to-point LSPs for traffic engineering

MPLS has been proposed as an efficient mechanism for traffic engineering in networks to account for and utilize network resources properly. A fundamental problem in MPLS is to reduce the label space usage by label switched paths (LSPs), while meeting the requirements of the flows traversing the network. Use of multipoint-to-point LSP trees has been proposed as a technique to reduce label space usage. However, various aspects with respect to the creation and maintenance of multipoint-to-point LSPs have not gained sufficient attention. We address the problem of merging a given set of LSPs into a minimum number of multipoint-to-point trees. We show that the problem of optimal merging is NP-complete and propose a heuristic for merging. The effectiveness of the algorithm in both offline and online cases is shown by analysis and simulation on real world scenarios.

A Fast Content-Based Data Distribution Infrastructure

We present Sieve – an infrastructure for fast content-based data distribution to interested users. The ability of Sieve to filter and forward high-bandwidth data streams stems from its distributed pipelined architecture. The complex message filtering task is broken-up into a sequence of light-weight filtering components resulting in high end-to-end throughput. Furthermore, since each component is assigned to a node based on its resource constraints, the queue buildup inside the nodes is minimal resulting in low end-to-end latency. Our experimental results based on real system implementation show that Sieve can sustain a throughput of more than 5000 messages per second for 100000 subscriptions with predicates of 10 attributes.