المستخلص: |
load balancing problem remains a critical issue and has a high priority in ad hoc network especially after the growing demand for this network. Designing effective load balancing strategies for this network is known to be difficult because of the complexity of the structure of the network, due to the nature of the variety of the dynamic connection nodes in ad hoc network. This complexity in ad hoc network may be influenced on many factors like performance, energy efficiency, scalability, power consumption, network topology. This paper describes five proposed methods for solving load balancing problem by applying them on node, link or on both of the node and link at the same time. The proposed methods are designed for dynamic load balancing based on the fuzzy neural network characterization benefits. Each proposed method uses some metrics to find load balancing in ad hoc network like bandwidth of link (amount of data that the link accepted it at the time (capacity)), throughput of node (a moment of data that processing at the time), queue state of node (under full….. over full)….etc. Since some problems like routing can be depended on these metrics. Load balancing attempts to maximize network throughput by distributing the load between links (or nodes) of the network, for example the traffic may be moved from the overloaded nodes to other lightly loaded.
|