Modeling of automotive sensor networks based on statistical motion models

Dissertation for M.S.c degree

Tension: System

Abstract:

Automotive sensor networks include mobile nodes that measure the physical and chemical parameters of the environment during movement. Sensor nodes usually have limited energy, and this lifetime limitation has caused many researches by researchers on wireless sensor networks, but the topic of lifetime in automotive sensor networks is less important than other sensor networks.

Coverage in automotive sensor networks refers to how the space is physically observed. Therefore, dynamic coverage is coverage based on the movement model of nodes that several vehicles work together to cover the environment.

Dynamic coverage has been studied by several researchers, most of these researchers have used the Boolean measurement model to model the dynamic coverage of the environment. Although the Shadow fading and Elfes measurement models are more realistic models for modeling the dynamic coverage of the environment by vehicle sensor networks.

In the upcoming thesis, in addition to examining different motion models, we have chosen the Gauss Markov motion model as the main motion model for dynamic coverage modeling according to the three Boolean, Shadow fading, and Elfes measurement models. We do this according to the theoretical calculations of the positions of the nodes in each movement step and comparing it with the direct results of the Gauss Markov movement model simulation. And finally, we make a comparison between dynamic coverage with three measurement models: Boolean, Shadow fading and Elfes.

Keyword: dynamic coverage, Gauss Markov movement model, Boolean measurement model, Shadow fading measurement model, Elfes measurement model

Introduction:

The wireless sensor network includes a number of static or dynamic sensor nodes that measure the environment and process the data obtained from the environment according to the defined needs of the network. This network has attracted the attention of researchers from various perspectives. One of the investigated aspects is the problem of network coverage. In this thesis, we intend to implement dynamic coverage modeling of automotive sensor networks based on statistical movement models. For this purpose, we have progressed step by step from the introduction of the network to the final goal of this project, i.e. the modeling of the dynamic coverage of automotive sensor networks based on statistical motion models. In the first chapter, we have introduced the wireless sensor network and we have introduced this network in a brief and useful way. In the second chapter, we have described the problem of coverage in wireless sensor networks and examined different types of coverage of this network. In the third chapter, we have evaluated the concepts related to energy and motion in the topic of wireless sensor network coverage. In the fourth chapter, we have examined the dynamic coverage and the advantages and challenges facing this coverage. In the fifth chapter, considering the previous four chapters and choosing the Gauss Markov movement model, we have implemented the dynamic coverage modeling of automotive sensor networks based on the Gauss Markov statistical model theoretically, as well as the simulation of theoretical calculations and the direct simulation of the Gauss Markov movement model at five different speeds using three measurement models: Boolean, Shadow fading and Elfes.

Goal:

The main goal This thesis is modeling the dynamic coverage of automotive sensor networks based on statistical motion models. Apart from examining different movement models, the main movement model chosen for this is Gauss Markov movement model. In order to better investigate the dynamic coverage, we use three different measurement models and compare the results of the investigations.

In this chapter, we first introduce wireless networks, and then we introduce different types of these networks in terms of structured and unstructured, and we also examine how nodes are deployed and move in wireless networks, and we introduce the characteristics of wireless sensor networks, and then we describe the configuration of wireless sensor network nodes.

In this chapter, we will first introduce wireless networks, and then we will introduce different types of these networks in terms of structured and unstructured, and we will also examine the deployment and movement of nodes in wireless networks, introduce the characteristics of wireless sensor networks, and then describe the configuration of wireless sensor network nodes.

Types of wireless networks:

Wireless networks are divided into two main types of structured and unstructured networks.     Structured wireless networks have infrastructures for setting up the network. One of the types of structured wireless networks is the GSM network [1]. Unstructured wireless networks are known as ad hoc wireless networks [2]. Contingent networks are called instant or temporary networks that are created for a specific purpose. The main difference between ad hoc networks and the usual 802.11 wireless networks is that in ad hoc networks, a set of mobile or non-mobile wireless nodes without any central infrastructure, access point or base station are connected to each other to send wireless information in a certain interval. Wireless networks can be a combination of these two mentioned items, i.e. a part of the structured network and a part of the unstructured network can be implemented.

Information packets are sent in contingent wireless networks by route nodes that have been previously specified by one of the routing algorithms. The noteworthy point is that each node is connected only with nodes that are in its radio radius, which are called neighboring nodes.

Routing protocols are optimally designed based on channel parameters such as attenuation, multipath propagation, interference, and also depending on the network application. During the design of these protocols, no attention was paid to guaranteeing security in ad hoc networks, but in recent years, due to the sensitive applications of this network, such as in military operations, medical emergencies, or assemblies and conferences, where the need to provide security in these networks has become more evident, researchers have put forward various proposals to ensure security in the two areas of performance and validity.

Ad hoc wireless networks without a central core to control sending and receiving is data and the transport of information packets is carried out personally by the nodes of a specific and dedicated route. The topology of contingency networks is variable because network nodes can be mobile and change their place at any moment of time. Sensor nodes usually have limited energy and lifetime is one of the topics being studied by researchers, but the topic of energy in the vehicular sensor network is less important.

Coverage of Vehicular Sensor Networks (VSNs) exhibits how well an area is observed in a particular physical space. Therefore, dynamic coverage is defined as coverage based on mobility models of nodes. Multiple vehicular sensors are required to collaboratively complete the scanning task. Dynamic coverage has been studied by several authors. Most authors have used the Boolean sensing model for network coverage, however, more realistic models are the Shadow fading sensing model and the Elfes sensing model.

In this paper, we choose the Gauss Markov mobility model in order to calculate the dynamic coverage rate based on the Boolean, Shadow fading and Elfes sensing models. This is accomplished by calculating the distribution and position of nodes in each step. Finally, we compare the dynamic coverage based on those calculation results with the dynamic coverage based on the direct simulation results of Gauss Markov mobility. Moreover, we compare the dynamic coverage of Boolean, Shadow fading and Elfes.