Abstract: Based on the small world phenomenon and the characters of scale-free network in evolutions of complex network,a new optimization algorithm,knapsack-problem optimization algorithm based on complex network(KOABCN),is put forward.There are two main steps in proposed algorithm: Firstly,knapsack-problem optimization space,with nodes increasing and weighted nodes’ degree preferential attachments,is formed,the scale-free network and degree distribution of nodes are presented.Secondly,taking nodes’ degree distribution in optimization space as prior knowledge,four operators,clustering,small world effect,adjust and optimal,are proposed to optimize attachments between nodes based on clustering and small world effect.Using Markov chain,the algorithm is proved to be convergent.Compared with other algorithms,KOABCN is shown to be an effective strategy to solve the combination optimization problem,such as 0/1 knapsack problem.