Andrei Pruteanu

NetDetect: Neighborhood Discovery in Wireless Networks Using Adaptive Beacons

It is generally foreseen that the number of wirelessly connected networking devices will increase in the next decades, leading to a rise in the number of applications involving large-scale networks. A major building block for enabling self-* system properties in ad-hoc scenarios is the run-time discovery of neighboring devices and somewhat equivalently, the estimation of the local node density. This problem has been studied extensively before, mainly in the context of fully-connected, synchronized networks. In this paper, we propose a novel adaptive and decentralized solution, the NetDetect algorithm, to the problem of discovering neighbors in a dynamic wireless network. The main difference with existing state of the art is that we target dynamic scenarios, i.e., multihop mesh networks involving mobile devices. The algorithm exploits the beaconing communication mechanism, dynamically adapting the beacon rate of the devices in the network based on local estimates of neighbor densities. We evaluate NetDetect on a variety of networks with increasing levels of dynamics: fully-connected networks, static and mobile multi-hop mesh networks. Results show that NetDetect performs well in all considered scenarios, maintaining a high rate of neighbor discoveries and good estimate of the neighborhood density even in very dynamic situations. More importantly, the proposed solution is adaptive, tracking changes in the local environment of the nodes without any additional algorithmic reconfiguration. Comparison with existing approaches shows that the proposed scheme is efficient from both convergence time and energy perspectives.

ChurnDetect: a gossip-based churn estimator for large-scale dynamic networks

With the ever increasing scale of dynamic wireless networks (such as MANETs, WSNs, VANETs, etc.), there is a growing need for performing aggregate computations, such as online detection of network churn, via distributed, robust and scalable algorithms. In this paper we introduce the ChurnDetect algorithm, a novel solution to the distributed churn estimation problem. Our solution consists in a gossiping-based algorithm, which incorporates a periodic reset mechanism (introduced as DiffusionReset). The main difference with existing state-of-the-art is that ChurnDetect does not require nodes to advertise their departure from the network nor to detect neighbors leaving the network. In our solution, all the nodes are interacting with each other wirelessly, by using a gossip-alike approach, thus keeping the message complexity to a minimum. We only use easy accessible information (i.e., about new nodes joining the network) rather than presuming knowledge on nodes leaving the system since that is highly unfeasible for most distributed applications. We provide convergence proofs for ChurnDetect, and present a number of results based on simulations and implementation on our local testbed. We characterize the performance of the algorithm, showcasing its distributed light-weight characteristics. The analysis leads to the conclusion that ChurnDetect is an attractive alternative to existing work on online churn estimation for dynamic wireless networks.

FailDetect: Gossip-Based Failure Estimator for Large-Scale Dynamic Networks

Ubiquitous and wirelessly connected devices are the present status quo in terms of networks around us. With the ever increase of scale, there comes also the problem of various communication failures. They are usually caused by hardware, software, or any other medium access contention. For the case of mobile networks, path uncertainty comes also into picture due to node mobility. All this leads to low quality of service and reduced user experience. The main contribution of the paper is the introduction of a novel distributed algorithm called FailDetect for the statistical estimation of the average packet loss in large-scale wireless distributed systems. It is based on a gossip mechanism, with the adding of periodic resets of the exchanged values. FailDetect is a fully- distributed scheme that does not presume time synchronization among the reset intervals for various nodes. A model and an evaluation by means of simulation and experiments show that FailDetect succeeds in evaluating the average packet loss of the network, while exhibiting low message-complexity.

ASH: Tackling Node Mobility in Large-Scale Networks

With an increased adoption of technologies like wireless sensor networks by real-world applications, dynamic network topologies are becoming the rule rather than the exception. Node mobility, however, introduces a range of problems (communication interference, path uncertainty, low quality of service and information loss, etc.) that are not handled well by periodically refreshing state information, as algorithms designed for static networks typically do. The main contribution of this paper is the introduction of a novel mechanism (called ASH) for the creation of a quasi-static overlay on top of a mobile topology. It is powered by simple, local interactions between nodes and exhibits self-healing and self-organization capabilities with respect to failures and node mobility. We show that the overlay mechanism works without assumptions about position, orientation, speed, motion correlation, and trajectory prediction of the nodes. A preliminary evaluation by means of simulation shows that ASH succeeds in tackling node mobility, while consuming only minimal resources.

Automatic discovery of algorithms for multi-agent systems

Automatic algorithm generation for large-scale distributed systems is one of the holy grails of artificial intelligence and agent-based modeling. It has direct applicability in future engineered (embedded) systems, such as mesh networks of sensors and actuators where there is a high need to harness their capabilities via algorithms that have good scalability characteristics. NetLogo has been extensively used as a teaching and research tool by computer scientists, for example for exploring distributed algorithms. Inventing such an algorithm usually involves a tedious reasoning process for each individual idea. In this paper, we report preliminary results in our effort to push the boundary of the discovery process even further, by replacing the classical approach with a guided search strategy that makes use of genetic programming targeting the NetLogo simulator. The effort moves from a manual model implementation to an automated discovery process. The only activity that is required is the implementation of primitives and the configuration of the tool-chain. In this paper, we explore the capabilities of our framework by re-inventing five well-known distributed algorithms.

GDE: a distributed gradient-based algorithm for distance estimation in large-scale networks

Today, wireless networks are connecting most of the devices around us. The scale of these systems demands for novel techniques to maintain availability for various services such as routing, localization, context detection etc. Distance estimation is one of their most important building blocks. The majority of current algorithms, presumes knowledge about node position via systems such as GPS. While for some application scenarios this approach is feasible, for a lot of cases it suffers from frequent unavailability and high costs in terms of energy consumption. The main contribution of this paper is the introduction of a novel distributed algorithm called GDE, for the estimation of distances in large-scale wireless networks. GDE is a mechanism which estimates distances between nodes based solely on local interactions. The evaluation by means of simulations shows that GDE succeeds in estimating the distance between nodes in both static and mobile scenarios with considerably high accuracy, even under the influence of different kinds of environment parameters, such as node density, node speed, spatial node distribution, multicast percentage, etc.

ASH: Tackling node mobility in large-scale networks

With the increased adoption of technologies like wireless sensor networks by real-world applications, dynamic network topologies are becoming the rule rather than the exception. Node mobility, however, introduces a range of problems (communication interference, path uncertainty, low quality of service and information loss, etc.) that are not handled well by periodically refreshing state information, as algorithms designed for static networks typically do. To address specifically this problem, the main contribution of this paper is the introduction of a novel mechanism (called ASH) for the creation of a quasi-static overlay on top of a mobile topology. It is powered by simple, local interactions between nodes and exhibits self-healing and self-organization capabilities with respect to failures and node mobility. We show that the overlay mechanism works without assumptions about position, orientation, speed, motion correlation, and trajectory prediction of the nodes. A preliminary evaluation by means of simulation shows that ASH succeeds in tackling node mobility, while consuming only minimal resources.

LossEstimate: Distributed failure estimation in wireless networks

The ongoing evolution of software-intensive distributed systems to ultra-large-scale (ULS) systems require innovative methods for building, running, and managing these systems. Component self-adaptation and self-configuration properties are thus becoming mandatory requirements in order to cope with application complexity. An increasing number of systems, such as video content distribution, make use of distributed feed-back mechanisms to build-up intelligent, robust and self-managing services. Technology wise, with the wide-spread usage of wireless communication interfaces on todays mobile devices, communication failures are an ever increasing nuisance in the design of distributed self-adaptive services and applications. Communication protocols designed for wired networks are not suited for this new class of networks (including mobile ad-hoc networks, wireless sensor networks, vehicular ad-hoc networks, etc.) due to the several orders of magnitude higher amount of communication failures. Although virtually every single existing communication protocol tries to deal with the various effects introduced by communication failures, almost all existing state of the art relies on previous knowledge about the amount of errors occurring at run time (information usually collected from previous deployments). A survey of current literature easily shows that, in contrast, applications that make use of distributed feedback mechanisms via online estimation of communication errors has received relatively small attention. In this paper we introduce a new distributed feedback mechanism, named LossEstimate, for runtime quantification of the global amount of communication failures present in a large-scale network. The new algorithm helps building self-adaptive services and has the advantage of being fully distributed - each node computes an estimate of the amount of errors using a gossip-alike approach. The algorithm is adaptive in the sense that it can follow changes in the mean value of the amount of communication failures over time. We focus our analysis on the impact of various network topologies, discussing the case of fully connected networks (relevant for the case of peer-to-peer networks), static multihop topologies (mapping on the case of wireless sensor networks) and mobile multihop networks (mapping on the case of mobile ad-hoc networks and vehicular ad-hoc networks). The results show that the algorithm performs well in all three scenarios, without requiring specific adaptations. Besides the lack of an alternative protocol, the gossip-alike characteristics make LossEstimate an attractive choice for building a distributed feedback mechanism via the online quantification of the amount of communication failures in large-scale networks, due to the fact that it exhibits a small communication overhead and has a small convergence time. It stands as an important building-block for engineering self-adaptive distributed applications and services, such as video streaming, by means of distributed feedback mechanisms.

Gossip-based density estimation in dynamic heterogeneous wireless sensor networks

The density estimation of diverse sensor types in a heterogeneous sensor network is an important service that can be used in clustering schemes, node redeployment and sleep scheduling strategies. Similar to any wireless sensor network service, energy efficiency is one of the main requirements. The service has to provide an updated estimation at each node. Network dynamics, especially node mobility, introduce new challenges. Moreover, churn makes the problem even more complicated. In this paper we introduce a new approach called Gossip based Density Estimation GDE for heterogeneous dynamic networks. The devised method is able to cope with node mobility and churn, as well as redeployment of new nodes. It is fully distributed and adaptive to network dynamics. We analyse the effect of mobility as well as increased scale in the number of clusters and the quantity of nodes. The simulation results support the idea that our algorithm has a fast convergence speed and provides more accurate estimation compared to similar approaches.

Gradient-Based Distance Estimation for Spatial Computers

Todays wireless networks are connecting more and more devices around us, leading to the birth of a new distributed computing platform, in the form of a spatial computer. The main difference with traditional computing models is that space and time become intertwined with computation, especially when scaling up the system. Computations performed by each element are now related to its spatial position. This property is the key ingredient when assuring the availability for various distributed networking services and applications. Computations become linked to the concept of space. Estimating distances between components (especially in dynamic networks characterized by the node mobility) thus becomes one of the most important building blocks for spatial computing. The majority of the algorithms that come from the MANET community presume knowledge about node position via systems such as GPS, or employ a one-time manual network topology configuration. While for some application scenarios this approach is feasible, for a lot of cases it suffers from frequent unavailability (e.g. indoors) and high costs in terms of energy consumption. Therefore, intense demand exists for a new kind of distance estimation algorithm using only simple local interactions, without knowledge of global information. The main contribution of the article is the introduction of a novel distributed algorithm, called gradient-based distance estimation (GDE), for the estimation of distances in networks characterized by mobility, specifically targeting the context of spatial computing. GDE is based on a gossiping mechanism to estimate distances between nodes with only local interactions. It significantly improves current state of the art by employing statistical analysis and making better use of the information available at each node.We analyze the parameters that should be considered by real applications, and present mathematical models to compensate their influence for distance estimation. Three spatial computing applications using GDE are presented: geographical cluster center detection, topological overlay shape construction and geographic routing. The simulation-based evaluation shows that GDE succeeds in estimating the distance between nodes in both static and mobile scenarios with considerably high accuracy for various simulations setups, such as varying node density, node speed or spatial node distribution.