The Study of Service Discovery Schemes in Wireless Sensor and Mobile Ad-hoc Networks

• Main Authors: Shyr-Kuen Chen, 陳石坤
• Other Authors: 王丕中
• Format: Others
• Language: en_US
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/sykgus

博士 === 國立中興大學 === 資訊科學與工程學系 === 99 === Two types of ad-hoc routing protocols are proposed in the current work to address the issue of packet delivery: topology based and position based (or geographic based) routing protocols. In topology-based routing protocols, each node broadcasts control messages to build up routes directed to itself. It maintains a routing table to record routes to the other nodes in the mobile ad-hoc network (MANET). When searching for services on networks, the system often depends on the broadcast or multicast mechanism to acquire information, which usually results in a large overhead. In contrast to topology-based routing protocols, position-based routing protocols are considered widely as a potentially scalable routing solution because they do not need to maintain routing tables. However, the challenge involved is the determination by the source node of the position of the destination before sending a packet. The simplest method for retrieving a destination node''s position is flooding. However, the flooding-based approach usually leads to heavy signaling traffic and power consumption for the MANET. Hence, the present dissertation designs and proposes the Design and Implementation of a Services Discovery Scheme in Wireless Sensors and Mobile Ad-hoc Networks (SDS). Three issues on SDS needed to be considered: (1) construction of a minimum-cost tree for routing protocols, (2) development of a reliable transmission protocol, and (3) construction of efficient distributed location services with shortcut schemes for position-based routing. To address topology-based routing protocol problems and to reduce the amount of request/reply packets, we propose an anycasting scheme for ad-hoc wireless networks. In this scheme, an anycast tree is established, and control-gates are used to reduce the control overhead. When a client node sends out a service request message to the anycast tree, the nearest or best server responds to the client. The other responses of service providers are discarded by the control-gate. Therefore, the anycasting scheme can reduce the control overhead and is suitable for large-scale ad-hoc wireless networks. It also supports any K service; it can be used for backup or fault tolerance. Our proposed anycast scheme can reduce the control overhead. However, a major challenge involved is the unstable forwarding path in wireless networks. Therefore, we present a recovery point (RP) and rerouting scheme to increase the reliability of packet transmission. The retransmit packet work is handed over to the nearest RP node by the receiver, and the sender does not need to retransmit the lost packets. The transmission path is rebuilt from the last node before the failure link. Therefore, the latency of path recovery is shorter than that for unicast-based approaches that must rebuild their path from the source node. In the current literature, position-based routing protocols are regarded to have better scalability and lower control overhead than topology-based routing protocols. Location services are the most critical part of position-based routing protocols, so we present two multi-home-region schemes. First is the Distributed Virtual Home Region with Spatial Awareness (DVHR-SA) scheme, which aims to improve the performance of location service. This scheme selects different update and query procedures adaptively according to the location of the source node. Second is the Efficient Location Forwarding with Shortcuts (ELFS) scheme, which uses the idea of shortcut to decrease the frequency of global updates and has packet forwarding with direction awareness. This scheme aims to reduce the path lengths of messages and to improve the cost of transmitting location information. Finally, experimental results are shown as follows. First, the anycast scheme can reduce the number of transmitted packets regardless of the number of service providers, service requests, and the required service providers in one service instance. Second, a reliable scheme also has the capability of fast rerouting. Therefore, a broken path can be recovered in a short latency, and the reliability of the transmitted packets can be ensured. Finally, the DVHR-SA and ELFS schemes shorten the lengths of the update, query, and reply paths. Both schemes also reduce the overall network message overhead.