Search

Position-based routing has been proven to be a scalable and efficient solution for packet routing in mobile ad hoc networks (MANETs) by utilizing location information of mobile nodes. The location service provides geographic locations for all nodes and is therefore critical to position-based routing. In general, the control overhead in a position-based routing protocol is mainly dominated by location updates. In this paper, we propose a location service called Modified Grid Location Service (MGLS), which employs a binary grid partitioning scheme to reduce the control overhead associated with the location management and supports large scale ad hoc networks. We then use a theoretical model to analyze both MGLS and GLS. Both theoretical analysis and simulation results show that MGLS can reduce the location update overhead in location services. [ABSTRACT FROM AUTHOR]

The broadcast operation in ad hoc wireless networks is essential, but expensive in terms of power consumption. In DP (dominant pruning) algorithms, the number of forward nodes is treated as a criterion to measure the consumed power. However, this principle is not suitable anymore if the power-adaptive characteristic is supported and each node can detect the strength of a received packet and tune its transmission power level. Thus, in this paper, we propose a modified version of the DP algorithm with the objective being to minimize the total power consumption of forward nodes. Comparing with original DP algorithms, the sender in our algorithm not only chooses its forward nodes, but also designates the transmission ranges of its forward nodes. As a result, the designated transmission ranges not only maintain the same coverage as DP algorithms, but also reduce the total power consumption. We propose a new forward nodes selection process where the weight of a node is dependent on the incremental transmission cost rather than the effective degree. Although the resultant number of forward nodes in the proposed approach is often greater than DP algorithms, the power consumption shown in simulations is less than DP algorithms. [ABSTRACT FROM AUTHOR]

We presents a power-aware routing protocol called Multi-Route Transmission Routing that utilizes multiple routes to transmit the data traffic simultaneously and leads to a balanced energy consumption. The proposed routing approach can extend the system lifetime of the network. A new routing decision index is also proposed for the route selection mechanism, which takes both the shortest-path and the maximum system lifetime into consideration and dynamically adjusts its weight between them according to the energy usage of the network. Experiment results show that, the proposed routing protocol provides a higher performance than other well-known power-aware routing protocols in terms of the energy-efficiency. [ABSTRACT FROM AUTHOR]

This paper presents a problem independent weighting strategy for weighted support vector machines (SVMs). SVMs can be applied with a weighting to each training vector to reflect the importance of different classes or training samples. Weightings are often assigned to the two classes inversely proportional to the sample count of each class, or according to a priori knowledge. Such a strategy can be applied to skewed data sets to balance the importance, error contribution and cost between the two classes. In this paper we propose a strategy to give each training pattern a weighting according to their distances to the classifier. The strategy regards the importance of the training patterns to the training process but not the importance of the data to the problem, thus it is suitable for general SVM applications. Experiments show that the performance of the proposed method is competitive to standard SVM while the training processes are even sped up. [ABSTRACT FROM AUTHOR]

Support vector machines like other classification approaches aim to learn the decision surface from the input points for classification problems or regression problems. In many applications, each input points may be associated with different weightings to reflect their relative strengths to conform to the decision surface. In our previous research, we applied a fuzzy membership to each input point and reformulate the support vector machines to be fuzzy support vector machines (FSVMs) such that different input points can make different contributions to the learning of the decision surface. [ABSTRACT FROM AUTHOR]