| Summary: | 博士 === 國立中山大學 === 資訊工程學系研究所 === 95 === In wireless ad hoc networks, the resources such as bandwidth, power, computing ability, etc., for end users are limited by the environments and hardware. In addition, the mobility of each user is different. Furthermore, the varied wireless networks with different properties such as transmission rate, coverage, power and protocols make it rarely difficult to establish and maintain the communications through the multiple intermediate nodes which are mobile devices. This dissertation will focus on the following three issues:
1. Most homogeneous wireless networks only considering one or less resources of the mobile device can not select and distribute the routes. In addition, it can cause the load of the intermediate nodes to increase.
2. The various wireless networks have different characteristics of resources. The route selection and distribution algorithms proposed before in homogeneous wireless networks are not adapted for the heterogeneous wireless networks.
3. Most route selection and distribution algorithms are designed for specific type of wireless environments (such as only for proactive type or reactive type). For different environment and different mobility, its cost will be different. To allow an environment-aware based adaptive management for heterogeneous wireless networks with different characteristics of resources becomes an important issue.
In homogeneous wireless ad hoc networks, the resources of the mobile devices such as bandwidth, power, computing ability, etc., are limited and different by the environments and hardware. To evaluate the individual resource of each mobile device, the route distribution and selection algorithm considering the multiple resource variations in wireless ad hoc networks is proposed. First, the evaluation and management for individual resource of each mobile device such as bandwidth, power, and computing ability, are proposed. Second, according to the proposed algorithm, a mobile device can be estimated whether it supports the QoS or not. In addition, these resources evaluation results are transformed as the bottleneck resource information. At last, according to the bottleneck resource information of all the found routes, the route distributed through different intermediate nodes with most resources is selected for load distribution and balance.
Considering the wireless network infrastructure today including 3G wireless communication, IEEE 802.11 wireless local area network, and wireless ad hoc networks, the characteristics such as transmission rates, bandwidth, power consumption, and transmission range of these wireless structures are not the same. In other words, the routing selection algorithm for homogeneous wireless networks may not be flexible for another wireless network. In this thesis, these different wireless network structures are integrated into one heterogeneous wireless network. Individual characteristics such as transmission rates, bandwidth, power consumption, and transmission range of these wireless structures are evaluated. According to the adaptive resource evaluation and route selection algorithm for the heterogeneous wireless networks, the route with the lowest cost and most resources is selected based on the integrated information of each resource estimation and the bottleneck information of each route found.
Besides considering the characteristic of the heterogeneous wireless networks, the module (type) of the route selection and distribution algorithms also induces different costs and performance. The existing algorithms can be defined and divided into two typical types: proactive (table-driven) protocol and reactive (source initiated) protocol. The proactive type protocol pays the memory and periodical update (bandwidth consumption) costs for selecting the routing directly. In opposite, the reactive type protocol saves the costs but may increase the overhead for finding a routing path to the destination. To decrease the total costs of finding and maintaining routing paths, the environment-aware based adaptive management for heterogeneous wireless networks is proposed in this thesis. Corresponding to the movement of the users, the appropriate type of the routing algorithm should be selected for reducing the total costs of finding routing paths. In the adaptive management, we propose the type selection process to individually evaluate the cost of the two type protocols (proactive and reactive) via considering the protocols and the mobility of the users. According to the evaluation results, the appropriate protocol with less cost will be selected for finding and maintaining the routes in the heterogeneous wireless networks.
According to the simulation, the routing selection and distribution algorithm proposed in this dissertation can not only effectively distribute the routing paths but also balance the load within homogeneous or heterogeneous wireless networks. Corresponding to the mobility of the users, the suitable type of routing protocol can be selected. Furthermore, the cost for routing paths searching, selection and maintenance is decreased. The block rate of these routing paths is reduced. Overall, according to the simulation, this dissertation makes the major contributions as follows:
1. The individual resource evaluations and managements for the mobile devices in the homogeneous wireless networks are proposed. The routes which satisfy the QoS can be selected and distributed according to the proposed algorithm.
2. The various wireless networks with different are integrated into a heterogeneous wireless network. The route search, selection, and maintenance algorithm for the hybrid properties heterogeneous wireless networks is proposed.
3. The classification to separate and classify the route search and selection algorithms is proposed. Corresponding to the mobility of the users, the environment-aware based adaptive management for heterogeneous wireless networks is proposed to reduce the costs caused by route search, selection, and maintenance.
|
|---|
