| Summary: | 碩士 === 國立東華大學 === 資訊工程學系 === 94 === In mobile ad-hoc networks, communication between two mobile nodes is achieved by the help of other relay nodes. Considering the limited resources of a node in ad-hoc networks, the node will not voluntarily relay the packets from other nodes.
In this thesis, we adopt a probabilistic micropayment strategy for packet pricing in ad-hoc networks. Our scheme can reduce the cost in billing process than the previous packet pricing schemes in which the relay nodes send back the requesting packets for charge to the backend billing server each time. The proposed scheme using probabilistic strategy does not send back the requesting packet except that the node checks a random value less than a probability P. Finally, we successfully reduce the number of charging requests to the billing server. By performing the probability check for a given packet size and setting different probabilities for different nodes and time slots, we may more fairly charge the user. The different P for different router means the different prices of transmission paths. This is reasonable, e.g., the optical fiber provides high QOS and it deserve the high price. The different P for different time slots is to show the concern of rush hours. For some real-time services, such as streaming audio and video services, even in rush hours we do not want the congestion occurred. Taking the packet size, transmission path and rush hours into account, we herein design three types of packet pricing scheme.
From experimental results, our proposed scheme really reduces the overheads of a billing system in ad-hoc networks. Moreover, our pricing system reduces the billing process of a relay node. This can save the resource of a relay node and finally improve the transmission efficiency in ad-hoc networks.
|
|---|
