Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4948119 | Neurocomputing | 2017 | 26 Pages |
Abstract
Congestion in wireless sensor networks causes packet loss, throughput reduction and low energy efficiency. To address this challenge, a transmission rate control method is presented in this article. The strategy calculates buffer occupancy ratio and estimates the congestion degree of the downstream node. Then, it sends this information to the current node. The current node adjusts the transmission rate to tackle the problem of congestion, improving the network throughput by using multi-classification obtained via Support Vector Machines (SVMs). SVM parameters are tuned, using genetic algorithm. Simulations showed that in most cases, the results of the SVM network match the actual data in training and testing phases. Also, simulation results demonstrated that the proposed method not only decreases energy consumption, packet loss and end to end delay in networks, but it also significantly improves throughput and network lifetime under different traffic conditions, especially in heavy traffic areas.
Keywords
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Majid Gholipour, Abolfazl Toroghi Haghighat, Mohammad Reza Meybodi,