Effective congestion control is an issue strongly impacting basic features demanded from modern network environment as reliability, high and stable throughput, and low delays. These characteristics define the quality of communication channels. Optimizing network nodes configuration for only one of mentioned features, can exacerbate other parameters. This paper focuses on avoiding and alleviating network congestions using multi-objective optimization for gain setting of used controllers. Unlike in other presented approaches, in this case the non-stationary, discrete, dynamical model is discussed. The significant advantage of this approach is in the better reflection of the real environment conditions, where the transmission delay is floating. As the further development of the control strategy, the controller with the memory of previous steps have been deployed. Such control strategy mitigates the unfavorable impact of extended delays. Both proposed control strategies tune the presented model of communication channel to alleviate the results of sudden, unexpected network state changes. It is obtained by maximization of available bandwidth usage combined with minimization of buffer utilization. This supports avoiding undesirable congestion effects like packet dropping, retransmissions, high delay, and low network throughput.
Multi-Objective Control Optimization for Greenhouse Environment Using Evolutionary Algorithms