Parallel genetic algorithm with elite and diverse cores for solving the minimum connected dominating set problem in wireless networks topology control

Finding an efficient communication structure among wireless network access points and wireless sensor nodes is essential in optimizing energy consumption and minimizing broadcast latency. Wireless networks Can control their nodes for efficient resource utilization via Topology Control. A topology control based on obtaining Minimum Connected Dominating Set (MCDS) is an efficient approach for constructing wireless network virtual backbone. A virtual backbone reduces energy consumption, reduce communication interference, reduce routing latency, and increase the bandwidth. We propose a new parallel genetic algorithm with elite and diverse cores for constructing wireless network virtual backbone based on finding MCDS of wireless nodes to be used in wireless network topology control. There are predefined number parallel workers, an elite core and a diverse core. All parallel components run genetic operators, and the elite core selects elite solutions among processed sub-population. On the other hand, diverse core looks for new solutions upon receiving elite solution in addition to received processed sub-population. Experimental results show that this algorithm gives better results compared to other methods, particularly for high dimension graph, which is the case in wireless sensor networks. Actually, using parallelism and featured elite and diverse search could help the proposed method to achieve 100% better results compared to its predecessor versions of sequential genetic algorithms. In addition to that, the algorithm is very stable as each result match the average result.