No abstract

No abstract

No abstract

In this paper, we have introduced three proactive maintenance strategies for static wireless sensor networks (WSNs) using a limited number of mobile maintainer robots: the centralised proactive maintenance strategy denoted by CPMS, the fixed distributed proactive maintenance strategy denoted by FDPMS and the adaptive distributed proactive maintenance strategy denoted by ADPMS. The proposed maintenance strategies are based on a simple energy dissipation analytical model to estimate the occurrence times of the expected sensor failures in the network. Once identified, the anticipated failures are replaced by the available robots before they happen. To select the appropriate maintainer robot for each expected failure, CPMS opts for a centralised scheduling method based on the genetic algorithm fundamentals. However, the distributed maintenance strategies (FDPMS and ADPMS) use two different WSN area partitioning methods to share the sensor maintenance tasks among the available robots. Simulation results have shown that CPMS gives the minimal network dysfunction time representing the interruption service time induced by the detected faulted nodes. However, due to its significant signalling cost, we have remarked that CPMS can be deployed only in small scale WSNs. In large-scale ones, ADPMS has demonstrated its efficiency in terms of the network dysfunction time, the robot travelled distance and the introduced signalling cost. In particular cases, when sensor failures are uniformly distributed on the network map, FDPMS has given the best performances. INTRODUCTIONProviding a continuous service is the main requirement for many wireless sensor network (WSN) applications. To achieve this goal, the WSN must deploy a set of mechanisms to protect the network coverage from the eventual sensor failures. In literature, many approaches have been proposed to conserve the initial WSN quality of service (QoS) parameters and restore the coverage and the connectivity upon a sensor failure such as the use of mobile wireless sensors [1] and the exploiting of the nodes redundancy in the network [2]. However, in order to reduce the maintenance strategy deployment cost, Y. Mei et al.[3] proposes to use a small number of mobile robots to deal with the detected sensor failures in a static WSN. Indeed, in [3], Y. Mei and all have presented a set of algorithms to detect, report and handle the occurred sensor failures in the WSN. Thus, three approaches are introduced to coordinate the robot motions: (i) the centralised manager algorithm (CMA); (ii) the fixed distributed manager algorithm (FDMA); and (iii) the Dynamic Distributed Manager Algorithm (DDMA).By reacting after the sensor failures detection, the classical WSN maintenance approaches (CMA, FDMA and DDMA) can cause the service interruption during the failure recovery times. The network dysfunction time ratio induced by the faulted node cannot be tolerable by many real-time WSN application types such as surveillance and military applications [4]. In this paper, we introduce newer proactiv...

Distributed Computing and Networking International audience In this paper, we focus on applications having quantitative QoS (Quality of Service) requirements on their end-to-end response time (or jitter). We propose a solution allowing the coexistence of two types of quantitative QoS garantees, deterministic and probabilistic, while providing a high resource utilization. Our solution combines the advantages of the deterministic approach and the probabilistic one. The deterministic approach is based on a worst case analysis. The probabilistic approach uses a mathematical model to obtain the probability that the response time exceeds a given value. We assume that flows are scheduled according to non-preemptive FP/FIFO. The packet with the highest fixed priority is scheduled first. If two packets share the same priority, the packet arrived first is scheduled first. We make no particular assumption concerning the flow priority and the nature of the QoS guarantee requested by the flow. An admission control derived from these results is then proposed, allowing each flow to receive a quantitative QoS guarantee adapted to its QoS requirements. An example illustrates the merits of the coexistence of deterministic and probabilistic QoS guarantees.