Vassil Kriakov

GPS-Free node localization in mobile wireless sensor networks

An important problem in mobile ad-hoc wireless sensor networks is the localization of individual nodes, i.e., each nodes awareness of its position relative to the network. In this paper, we introduce a variant of this problem (directional localization) where each node must be aware of both its position and orientation relative to the network. This variant is especially relevant for the applications in which mobile nodes in a sensor network are required to move in a collaborative manner. Using global positioning systems for localization in large scale sensor networks is not cost effective and may be impractical in enclosed spaces. On the other hand, a set of pre-existing anchors with globally known positions may not always be available. To address these issues, in this work we propose an algorithm for directional node localization based on relative motion of neighboring nodes in an ad-hoc sensor network without an infrastructure of global positioning systems (GPS), anchor points, or even mobile seeds with known locations. Through simulation studies, we demonstrate that our algorithm scales well for large numbers of nodes and provides convergent localization over time, even with errors introduced by motion actuators and distance measurements. Furthermore, based on our localization algorithm, we introduce mechanisms to preserve network formation during directed mobility in mobile sensor networks. Our simulations confirm that, in a number of realistic scenarios, our algorithm provides for a mobile sensor network that is stable over time irrespective of speed, while using only constant storage per neighbor.

Management of highly dynamic multidimensional data in a cluster of workstations

Due to the proliferation and widespread use of mobile devices and satellite based sensors there has been increased interest in storing and managing spatio-temporal and sensory data. It has been recognized that centralized and monolithic index structures are not scalable enough to address the highly dynamic nature (high update rates) and the unpredictable access patterns in such datasets. In this paper, we propose an adaptive networked index method designed to address the above challenges. Our method not only facilitates fast query and update response times via dynamic data partitioning but is also able to self-tune highly loaded sites. Our contributions consist of techniques that offer dynamic load balancing of computing sites, non-disruptive on-the-fly addition/removal of storing sites, distributed collaborative decision making for the self-administering of the manager, and statistics-based data reorganization. These features are incorporated into a distributed software layer prototype used to evaluate the design choices made. Our experimentation compares the performance of a baseline configuration with our multi-site system, examines the attained speed-up as a function of the sites participating, investigates the effect of data reorganization on query/update response times, asserts the effectiveness of our proposed dynamic load balancing method, and examines the behavior of the system under diverse types of multi-dimensional data.

Managing cohort movement of mobile sensors via GPS-free and compass-free node localization

A critical problem in mobile ad hoc wireless sensor networks is each nodes awareness of its position relative to the network. This problem is known as localization. In this paper, we introduce a variant of this problem, directional localization, where each node must be aware of both its position and orientation relative to its neighbors. Directional localization is relevant for applications that require uniform area coverage and coherent movement. Using global positioning systems for localization in large scale sensor networks may be impractical in enclosed spaces, and might not be cost effective. In addition, a set of pre-existing anchors with globally known positions may not always be available. In this context, we propose two distributed algorithms based on directional localization that facilitate the collaborative movement of nodes in a sensor network without the need for global positioning systems, seed nodes or a pre-existing infrastructure such as anchors with known positions. Our first algorithm, GPS-free Directed Localization (GDL) assumes the availability of a simple digital compass on each sensor node. We relax this requirement in our second algorithm termed GPS- and Compass-free Directed Localization (GCDL). Through experimentation, we demonstrate that our algorithms scale well for large numbers of nodes and provide convergent localization over time, despite errors introduced by motion actuators and distance measurements. In addition, we introduce mechanisms to preserve swarm formation during directed sensor network mobility. Our simulations confirm that, in a number of realistic scenarios, our algorithms provide for a mobile sensor network that preserves its formation over time, irrespective of speed and distance traveled. We also present our method to organize the sensor nodes in a polygonal geometric shape of our choice even in noisy environments, and investigate the possible uses of this approach in search-and-rescue type of missions.

Spatio-temporal data services in a shared-nothing environment

Recently, there has been a proliferation of applications that produce spatiotemporal data that has to be processed, stored and queried efficiently. These applications necessitate the execution of millions of updates in order to keep the underlying database up-to-date. Consequently, there is a need for spatiotemporal data management systems that are able to support such update intensive operations. Moreover, these systems should offer users the capability to examine present as well as past (historical) data versions in an on-line fashion. We propose a system that exploits the inherent parallelism of a shared-nothing computing environment for storing and indexing the spatiotemporal data. We describe our proposed system architecture, data organization, and outline techniques for ensuring robustness and scalability under excessive query loads and high update rates.

Adaptive neighborhood selection in peer-to-peer networks based on content similarity and reputation

To address the two most critical issues in P2P file-sharing systems: efficient information discovery and authentic data acquisition, we propose a Gnutella-like file-sharing protocol termed Adaptive Gnutella Protocol (AGP) that not only improves the querying efficiency in a P2P network but also enhances the quality of search results at the same time. The reputation scheme in the proposed AGP evaluates the credibility of peers based on their contributions to P2P services and subsequently clusters nodes together according to their reputation and shared content, essentially transforming the P2P overlay network into a topology with collaborative and reputed nodes as its core. By detecting malicious peers as well as free-riders and eventually pushing them to the edge of the overlay network, our AGP propagates search queries mainly within the core of the topology, accelerating the information discovery process. Furthermore, the clustering of nodes based on authentic and similar content in our AGP also improves the quality of search results. We have implemented the AGP with the PeerSim simulation engine and conducted thorough experiments on diverse network topologies and various mixtures of honest/dishonest nodes to demonstrate improvements in topology transformation, query efficiency, and search quality by our AGP.

Self-tuning management of update-intensive multidimensional data in clusters of workstations

Contemporary applications continuously modify large volumes of multidimensional data that must be accessed efficiently and, more importantly, must be updated in a timely manner. Single-server storage approaches are insufficient when managing such volumes of data, while the high frequency of data modification render classical indexing methods inefficient. To address these two problems we introduce a distributed storage manager for multidimensional data based on a Cluster-of-Workstations. The manager addresses the above challenges through a set of mechanisms that, through selective on-line data reorganization, collectively maintain a balanced load across a cluster of workstations. With the help of both a highly efficient and speedy self-tuning mechanism, based on a new data structure called stat-index, as well as a query aggregation and clustering algorithm, our storage manager attains short query response times even in the presence of massive modifications and highly skewed access patterns. Furthermore, we provide a data migration cost model used to determine the best data redistribution strategy. Through extensive experimentation with our prototype, we establish that our storage manager can sustain significant update rates with minimal overhead.

Adaptive Management of Communities Based on Peer Content and Reputation

Most P2P file-sharing systems are unable to create self- organizing communities of similar nodes that provide good services to their members. In this paper, we propose a Gnutella-like file-sharing protocol based on the premise that each peer only creates links with the best counterparts which the peer has discovered in the network. Termed adaptive Gnutella protocol (AGP), our proposal transforms the overlay topology based on a reputation scheme that evaluates the provided services and offers a mechanism that organizes trusted nodes with similar content. We have implemented the AGP protocol using the PeerSim engine and conducted experiments on diverse network topologies. Over time, the network topology improves as every peer locates counterparts with similar content and good reputation. Moreover, malicious nodes are pushed to the edge of the overlay network and are excluded from participating in the AGP search.

FlexFS: Transparent resilience for GRID storage resources

Existing GRID infrastructures rely on explicit user instructions in order to replicate files for the purposes of resiliency. This human-intensive process is inefficient, error prone and, more importantly, makes file replication in GRIDs a cumbersome task. To address this problem, we introduce FlexFS – a fully automated file-system framework that seamlessly plugs into existing GRID structures providing automated file replication and transparent-to-user resilience. FlexFS breaks apart files into blocks and injects resilient information into these blocks through the use of Forward Erasure Correction codes. FlexFS employs a number of methods that facilitate the automated storage and efficient retrieval of the blocks in order to provide I/O throughput similar to that of local hard disks, all in the face of ever-changing utilization and availability of the GRID resources. Compared to currently available GRID replication schemes, FlexFS attains 15% to 230% higher throughput, both for reading and writing files.

Effective mechanisms for detecting and thwarting Distributed

The emergence of applications producing continuous high-frequency data streams has brought forth a large body of research in the area of distributed stream processing. In presence of high volumes of data, efforts have primarily concentrated on providing approximate aggregate or top-k type results. Scalable solutions for providing answers to window join queries in distributed stream processing systems have received limited attention to date. We provide a solution for the window join in a distributed stream processing system which features reduced inter-node communications achieved through automatic throughput handling based on resource availability. Our approach is based on incrementally updated discrete Fourier transforms (DFTs). Furthermore, we provide formulae for computingDFT compression factors in order to achieve information reduction. We perform WAN-based prototype experiments to ascertain the viability and establish the effectiveness of our method. Our experimental results reveal that our method scales in terms of throughput and error rates, achieving sub-linear message complexity in domains that exhibit a geographic skew in the joining attributes.The emergence of applications producing continuous high-frequency data streams has brought forth a large body of research in the area of distributed stream processing. In presence of high volumes of data, efforts have primarily concentrated on providing approximate aggregate or top-k type results. Scalable solutions for providing answers to window join queries in distributed stream processing systems have received limited attention to date. We provide a solution for the window join in a distributed stream processing system which features reduced inter-node communications achieved through automatic throughput handling based on resource availability. Our approach is based on incrementally updated discrete Fourier transforms (DFTs). Furthermore, we provide formulae for computing DFT compression factors in order to achieve information reduction. We perform WAN-based prototype experiments to ascertain the viability and establish the effectiveness of our method. Our experimental results reveal that our method scales in terms of throughput and error rates, achieving sub-linear message complexity in domains that exhibit a geographic skew in the joining attributes.