Search

The performance of vehicular ad hoc networks (VANETs) is subject to frequent changes due to node mobility so that nonstationary/transient network behavior always exists. In addition, safety applications in VANETs are very sensitive to the real-time performance of delay and packet delivery ratio (PDR). Although there are extensive studies on the steady-state (SS) performance analysis of vehicular networks, little work exists on evaluating their time-varying behavior. In this paper, we develop a performance model to efficiently estimate the dynamic behavior of VANETs. In this paper, vehicle's transmission queue is modeled using fluid-flow (FF)-based differential equations, which are solved using numerical methods, whereas the network hearing topology is modeled by a time-varying adjacency matrix that can be determined from stochastic models, measurements, or discrete event simulations. Numerical results illustrate that our performance model is able to promptly respond to the ongoing dynamic conditions of VANETs and provide reasonably accurate performance results. [ABSTRACT FROM AUTHOR]

In this paper, we develop a performance modeling technique for analyzing the time-varying network layer queuing behavior of multihop wireless networks with constant bit rate (CBR) traffic. Our approach is a hybrid of a time-varying adjacency matrix and a fluid flow queuing network model. The mobile network topology is modeled using a time-varying adjacency matrix, whereas node queues are modeled using fluid-flow-based differential equations that are solved using numerical methods. Numerical and simulation experiments show that this new approach can provide reasonably accurate results. Moreover, when compared with the computation time required in a standard discrete event simulator, the fluid-flow-based model is shown to be a more scalable tool. Finally, an illustrative example of our modeling technique application is given to show its capability of capturing the time-varying network performance as a function of traffic load, node mobility, and wireless link quality. [ABSTRACT FROM PUBLISHER]

This letter addresses the problem of wireless indoor positioning and proposes the concept of personal mobility map (PMM). As a graph describing an individual's typical mobility pattern over a physical or abstract space, PMM can be used as the building block for crowd-sourcing based simultaneous localization and mapping (SLAM). PMM can also enable new kinds of location-based services (LBSs). To construct a PMM, we conduct intra-sequence clustering on the sequences of Wi-Fi received signal strength (RSS) measurements to cluster the measurements into different location points (LPs), followed by personal inter-sequence LP assembling, which aligns these LPs with the concept of the Smith-Waterman algorithm originally used for protein sequencing. Experiments with Wi-Fi RSS measurements collected in the university campus prove that the PMM performs well in indoor mapping and path estimation. [ABSTRACT FROM PUBLISHER]