Catálogo de publicaciones - libros

Many applications in wireless sensor networks require sensor nodes to obtain their absolute or relative positions. Although various localization algorithms have been proposed recently, most of them require nodes to be equipped with range measurement hardware to obtain distance information. In this paper, an area localization method based on Support Vector Machines (SVM) for mobile nodes in wireless sensor networks is presented. Area localization is introduced as an evaluation metric. The area localization procedure contains two phases. Firstly, the RF-based method is used to determine whether the nodes have moved, which only utilizes the value change of RSSI value rather than range measurement. Secondly, connectivity information and SVM algorithm are used for area localization of mobile nodes. The area localization is introduced to trade off the accuracy and precision. And area localization, as a new metric, is used to evaluate our method. The simulation experiments achieve good results.

Most existing mutual exclusion algorithms for mobile ad hoc networks (MANETs) adopt a token-based approach. In traditional wired networks, timeout-based mechanisms are commonly used to detect token losses. However, in MANETs, it is difficult to set a proper timeout value due to the network dynamics. In this paper, we propose a dual-token-based mutual exclusion algorithm, which can tolerate token losses without using timeout. Two tokens are concurrently circulated in the system to monitor each other by using sequence numbers. If one token is lost, the other token can detect the loss and regenerate a new token. Simulations have been carried out to evaluate the effectiveness and performance of the proposed algorithm in comparison with the timeout-based approach. The results show that the timeout-based algorithm may falsely claim the loss of a token, thus cannot guarantee the correctness of mutual exclusion algorithms. On the contrary, our proposed algorithm can avoid false detection of token losses and satisfy all the correctness requirements of mutual exclusion, though it costs a bit more messages and longer time.

Most of the current RSS (Received Signal Strength)-based localization algorithms in Wireless Sensor Networks (WSNs) rely on isotropic radio propagation model to infer the distance between a pair of nodes from received signal strength, which however has been proved to be quite unreliable in recent research work. Performance analysis is important to evaluate the applicability of a localization algorithm, however little work has been done on evaluating the existing localization algorithms in simulated realistic settings. This paper firstly presents the motivation and detailed implementation of a proposed Link-State Based Annulus (LSBA) localization algorithm, and then gives a panorama of performance comparison among LSBA and other four localization algorithms in terms of estimation error, convergence speed, computational complexity and communication cost in the simulated realistic environment. Simulation results show that LSBA achieves the best tradeoff among all the four metrics in WSNs with moderate number of anchors, and has good adaptability to irregular node deployment as well.

In wireless sensor network applications, effective clustering algorithm can reduce energy consumption, which can increase network scalability and lifetime. In most of traditional clustering methods, clusters do not form until their cluster heads are selected and these algorithms are usually running under the single data transmission mode. A clustering strategy for WSNs based on Fuzzy Cluster Mean (FCM) is proposed in this paper. It is a new clustering mechanism of generating the clusters first and then selecting the cluster head(CH). This strategy based on FCM has good characteristics of clustering quick, reducing energy consumption and being applied in different data transmission modes. The correctness and feasibility is validated in simulations. It is shown that energy consumption is better than the similar clustering algorithms. When different selection modes of CH are selected in this strategy, the clustering strategy based on FCM shows the best efficiency in two data transmission modes.

Mobile ad hoc networks must operate independent of fixed or centralized network management infrastructure, the communication between nodes rely on the node’s cooperation, so nodes’ selfish behavior to save the battery cost for their own communication cannot be neglect. However, general cooperation enforcement mechanisms do not consider node battery, so hotspot node issue may exist; even badly lead to network separation. This paper proposes an Energy Fairness Cooperation Enforcement Mechanism (EFCEM) in ad hoc networks. Theoretical analysis and numerical simulation results show that the EFCEM protocol, comparing with the traditional reputation mechanisms, can prolong networks existence time and improve network throughput, but at the cost of a little more packets delay.

Coverage is one of the most important issues in sensor networks. For random sensor deployment, one would like to know whether the deployed sensors can cover the whole field sufficiently. In this paper, we attempt to find a set of critical points in a sensor field so as to answer the yes/no question of complete coverage only by checking whether these points are covered, which is based on a new model of fuzzy information coverage. The higher order Voronoi diagram is analyzed and a sufficient condition is provided for a sensor field which is completely fuzzy information covered by a given set of sensors. The corresponding algorithm in (KN+NlogN) runtime is also presented.

RFID, as an emerging technology, has very huge potential in today’s social and business developments. Security and Privacy are one of the important issues in the design of practical RFID protocols. In this paper, we focus on RFID authentication protocol. RFID mutual authentication is used to ensure that only an authorized RFID reader can access to the data of RFID tag while the RFID tag is confirmed that it releases data to the authenticated RFID reader. This paper will propose an anonymous mutual authentication protocol for RFID tag and reader. RFID tag is anonymous to RFID reader so that privacy can be preserved. In addition, mutual authentication does not need to rely on a back-end database.

Providing easy access to the Internet is a prerequisite for successful deployment of Mobile Ad-hoc Networks (MANET). This paper proposes an Identifiers Separating and Mapping Scheme (ISMS) based Internet access solution for MANET, called ISMS-MANET. Firstly, ISMS, which is a candidate for the future Internet architecture, is briefly described. Then, how the Ad hoc On-Demand Distance Vector (AODV) routing protocol can be used for connecting a MANET to the ISMS based Internet is discussed. In ISMS-MANET, the Access Router (AR) implements AODV protocol and works like a normal MANET node. The node in the MANET relies on the Route Reply message of AODV to find the AR. A prototype system is implemented. Performance analysis shows that ISMS-MANET has much faster handover and lower packet processing delay than the Mobile-IP based solution.

In sensor networks, queries need to be jointly designed, in order to minimize the power consumption and maximize the lifetime. Data reduction techniques can be employed to decrease the size of data to be transferred in the network, and therefore save energy of sensor nodes. This paper presents a novel method for optimizing multi-query in sensor networks. Our approach is, by using packet combination techniques, to reduce the data size of multiple simultaneous queries, so that the energy for data transmission can be saved to the best extent. A delay item is specified together with the query by the application. Then an optimal query plan can be obtained by studying the best time of sending local data to sink that can lead to the minimum cost. Algorithm is described in detail. Performance analysis is performed to validate the effectiveness of the proposed method.

It is common that a small group of autonomous mobile units requires location information of each other, but it is often not applicable to build infrastructure for location service. Hence, mobile units use RF signals to determine their relative positions within the group. Several techniques have been proposed for localization, but none of them consider both units mobility and anchor unavailability. In this paper we develop a propagation scheme for spreading the signal strength information through the network, and filtering techniques are employed to process the noisy signal. We use Floyd-Warshall algorithm to generate pairwise signal distance of each pair of units. Then Multidimensional Scaling technique is used to generate relative position from pairwise distances. Due to anchor unavailability, relative positions are adjusted by certain rules to reflect the continuous mobility. We verify our methods in MICAz platform. Experimental results show that we can obtain smooth relative positions under mobility and manage moving patterns of mobile units.