Evolutionary computation with spatial receding horizon control to minimize network coding resources.

Area of Science:

• Computer Science
• Network Engineering
• Optimization Theory

Background:

• Minimizing network coding resources is challenging due to the NP-hard nature and massive scale of real-world networks.
• Genetic algorithms (GAs) show promise for NP-hard problems but struggle with scalability in large-scale systems.
• Existing network partitioning methods are limited in their ability to leverage global network topology information.

Purpose of the Study:

• To propose a novel spatial receding horizon control (SRHC) strategy for effective network partitioning.
• To design an efficient genetic algorithm (GA) tailored for the network coding problem (NCP) within partitioned networks.
• To generalize network partitioning by introducing an N-step-wide SRHC approach.

Main Methods:

• Developed a generalized N-step-wide spatial receding horizon control (SRHC) for network partitioning.
• Designed an efficient genetic algorithm (GA) to address the network coding problem (NCP) using the SRHC strategy.
• Compared the proposed SRHC method against traditional one-step-wide partitioning techniques.

Main Results:

• The proposed SRHC strategy effectively partitions large-scale networks, overcoming scalability issues of traditional GAs.
• The efficient GA, combined with SRHC, demonstrates superior performance in solving the network coding problem (NCP).
• Experimental results validate the advantages of the SRHC and GA approach for NCP optimization.

Conclusions:

• The novel SRHC strategy provides a generalized and effective network partitioning technology for large-scale systems.
• The developed GA, integrated with SRHC, offers an efficient solution for the NP-hard network coding problem (NCP).
• The SRHC approach shows potential for extension to other large-scale complex optimization problems in computer science and engineering.