Translated Abstract
Frequent disasters can cause serious loss of life and property. Disaster relief netowrk asims to improve the efficiency of the relief efforts. From the technical features point of view, the Mobile Ad Hoc networks (MANETs) is undoubtedly a good way to build the disaster relief network. The key question of applying the MANETs to this purpose is if ti has haigh enough scalability, as a disaster may require a large scale network. In addition, due to the different nature of different types of disasters relief networks with different features might be required.
This thesis focuses on the MANETs-based disaster relief network. Considering the characteristics of this type of network, we improved the mechanisms of network latency, throughput and scalability. Several aspects of wireless communication and networks are explored in the thesis, including location services, channel assignment, and routing design. The main contributions are listed as follows:
1) This thesis propose a method named Group mobility oriented history information based light location service (G-HLLS) to address the large update overhead problem in traditional location service. Considering the group mobility property in many MANETs applications, G-HLLS combines the location service with clustering scheme in order to meet the request of group mobility and reduce the storage cost. Using the simulation based on OPNET, the performance of G-HLLS is analyzed in terms of query success ratio, overhead and storage cost. The results show that G-HLLS has good scalability in MANETs with group mobility.
2) This thesis propose a Geographic Opportunistic Routing with Self-decision (SGOR) to solve the inefficiencies or failure problems of the routing protocol caused by frequent topology detection and location update in the high dynamic disaster relief networks. In stead of specifying the forwarding node, SGOR carries rules in the broadcasting message. The receiving node decides whether to forward the message according to the pre-set rules or its own geographic information independently, and calculates its own priority according to the greedy rule. Thus, the topology detection and location updates are avoided. This thesis also propose an ETT-based Semi fixed Geographic routing algorithm (ESGR) to solve the localized problem in the disaster relief networks. ESGR maps the 2-hops links to the virtual neighbors. It saves the intermediate node’s routing time, uses the Expected Transmission Time (ETT) as the fixed links preferred reference factors, and reduces the probability of retransmission by using high-quality links. The algorithm uses boundary node set mapping to boundary link set to cope with the localized problem, which can save the time of protocol switching and path finding. Simulation results show that our routing protocols outperforms the traditional reactive routing in terms of average delay and packet delivery rate.
3) Multi channel technology is an effective technology to improve the performance of large data transmission in the disaster relief network. But channel interference and resource contention problems greatly decrease the network throughput and seriously affect the network efficiency. Further, the ripple effect which caused by inappropriate channel switching will result in poor channel utilization and unstable networks. In this thesis, based on a searching tree structure, this thesis combine the impact of interference degree with the channel resource allocation to reduce the ripple effect and use of channel resources. Simulation results show that this mechanism effectively reduces the channel interference and improves the network throughput.
4) To overcome the problem of network performance degradation and throughput reduction caused by the link interference in multi-channel network, this thesis proposed a cross-layer routing algorithm based on a new channel bandwidth estimation method. With the direct representation of channel interference, i.e. channel idleness, our algorithm uses a more accurate channel idleness ratio estimation method as basis. Based on channel serviceable condition, channel interference are avoided and performance of network is improved.
Corresponding authors email
