Wireless sensor networks (WSN) have recently received an increasing interest. They are now expected to be deployed for long periods of time, thus requiring software updates. Updating the software code automatically on a huge number of sensors is a tremendous task, as ''by hand'' updates can obviously not be considered, especially when all participating sensors are embedded on mobile entities. In this paper, we investigate an approach to automatically update software in mobile sensor-based application when no localization mechanism is available. We leverage the peer-to-peer cooperation paradigm to achieve a good trade-off between reliability and scalability of code propagation. More specifically, we present the design and evaluation of GCP ({\emph Gossip-based Code Propagation}), a distributed software update algorithm for mobile wireless sensor networks. GCP relies on two different mechanisms (piggy-backing and forwarding control) to improve significantly the load balance without sacrificing on the propagation speed. We compare GCP against traditional dissemination approaches. Simulation results based on both synthetic and realistic workloads show that GCP achieves a good convergence speed while balancing the load evenly between sensors.
GCP: Mise à jour épidémique de logiciels pour les réseaux de capteurs mobiles, larges échelles Résumé : GCP est un systeme de mise à jour de code automatique pour réseaux de capteurs mobiles, utilisant le concept de diffusion épidémique développés dans le cadre de réseaux filaires. Ce rapport présente la conception et l'évaluation de GCP (Gossip-based Code Propagation), protocole proposé dans le cadre de ces travaux de recherches.
Celui-ci est évalué par comparaison avec des algorithmes traditionnels de dissemination de données. Les résultats de simulations sont fondés à la fois sur des traces réelles et générées, permettant de montrer l’efficacité de GCP tant dans la vitesse de propagation que dans l’équilibrage des charges sur le réseau.
Recently, compact devices, called micro-electro mechanical systems (MEMS), have appeared. Such devices combine small size, low cost, adaptability, low power consumption, large scale and self-organization. Equipped with wireless communication capability, such appliances (called mote node or sensor 1 ) together form a wireless sensor network (WSN). Due to their tiny size, sensors possess slim resources in term of memory, CPU, energy, etc. [1,16].
The increasing interest in WSNs is fundamentally due to their reliability, accuracy, flexibility, cost effectiveness and ease of deployment characteristics. Such WSNs can be deployed for monitoring purposes for example. For example we can cite Ecosystem monitoring, Military (battlefield surveillance, enemy tracking, . . . ), Biomedical and health monitoring (cancer detector, artificial retina, organ monitor, . . . ), Home (childhood education, smart home/office environment, . . . ).
Sensors may be deployed both in static and dynamic environments. They are usually deployed for a long period of time, during which the software may require updates. While efficient solutions to software update may be deployed in fixed WSNs, this is far more complex when sensors are embedded on mobile entities such as people. In this paper, we consider this latter setting, namely a WSN deployed over a group of people.
Given the potentially large number of participating sensors in WSN and their limited resources, it is crucial to use fully decentralized solutions and to balance the load as evenly as possible between participating sensors. In that context, we investigate the use of the P2P communication paradigm which turns out to be a relevant candidate in this context. Considering similarities between these two systems, Section 2 investigates the relevance of adapting the P2P paradigm to mobile WSN. Using epidemic-based dissemination, we introduce a greedy protocol (GCP: Gossip-based Code Propagation) balancing the dissemination load without increasing diffusion time. GCP relies on piggy-backing to save up energy and forwarding control to balance the load among the nodes.
Code propagation (or reprogramming service) has a lot in common with broadcast and data dissemination [5,8] with an additional main-constraint. In broadcast, each message sent before a node arrival can be ignored by this node. In software update protocols, each new node in the network has to be informed of the existence of the software’s latest version as soon as possible to be operational.
This paper is organized as follows. Section 2 presents the P2P paradigm and more specifically the epidemic principle. Section 3 introduces the GCP algorithm and alternative approaches. We compared GCP with classical approaches by simulation and depict the results in Section 4. Finally, Section 5 introduces a short state of art of the different domains cited above before concluding in Section 6.
Classical peer-to-peer systems are composed of millions of Personal Computers connected together by a wired network and as opposed to sensor networks are 1).
In this section we promote the idea that sensor networks and peer-to-peer systems are similar enough so that P2P solutions can be seriously considered in the context of WSNs.
On one hand, sensors and personal computers have incomparable resources: PCs have large resources in terms of CPU, storage while sensors are very limited; sensors have strong energy constraints, limiting their capability to communicate. On the other hand, due to the node’s resources compared to the size of the system, no entity is able to manage the entire network in both systems. Every application designed for both networks requires a strong cooperation between entities to be able to manage the network and to take advantage of the entire network. Peer to peer solutions heavily rely on such collaboration. The peer to peer communication paradigm has been clearly identified as a key to scalability in wired systems. In a P2P system, each node may act both as a client and a server, and knows only few other nodes. Each node is logically connected to a subset of participating nodes forming a logical overlay over the physical network. With this local knowledge, the resource aggre
This content is AI-processed based on open access ArXiv data.