Fixed Relay Nodes Development Strategy In The Sparse Opportunistic Networks

In a wide range of opportunistic networks, due to the little communication between nodes, the network is in a divided state. In order to improve the performance of a wide range of opportunistic network, it is an option to place fixed relay nodes. Since fixed relay nodes are additional resources, too many fixed relay nodes increase the cost of network constructing, therefore, the location of fixed relay nodes is crucial.Firstly, the paper proposes a multi-objective optimization of energy effectiveness based for fixed relay nodes deployment strategies and a network revenue-energy consumption model, in which the performance of effectiveness is a selected metric. As for the limit of the greedy algorithms, the paper proposes to use genetic algorithms to solve fixed relay nodes deployment by improving the selection operator, and use the selected fixed relay nodes in the sparse opportunistic network to improve the average ability of the unit energy consumption transmitting the message. Secondly, for the particularity of buffer space of fixed relay nodes, the paper designs a cache management strategy of fixed relay nodes which associates with messages. The strategy interacts the message queue information and the contact information of nodes in order to reduce numbers of redundant messages in the cache of fixed relay nodes.Finally, we use ONE simulator to verify the performance of proposed fixed relay nodes deployment strategy and the cache management policy, then comparing them with the existing fixed relay nodes deployment and the cache management strategy. Experiments show that the proposed fixed relay node deployment strategy in this paper can improve the successful rate of message transmission by about 2%, improve the average forward-energy efficiency of fixed relay nodes by about 0.005 and the proposed cache management strategy can improve the successful rate of message transmission by about 3%, decrease the average end-to-end latency by about 40s, decrease network overhead by about 10.