关键词: Rapidio网络, 遗传算法, 服务质量, 适度函数, 算法编码

Abstract: Concerning the drawback of the existing Rapidio network routing management strategy, an optimized strategy based on improved Genetic Algorithm (GA) was proposed. By doing some improvements on the routing strategy, coding, mutation and crossover of conventional GA, the performance of the Quality of Service （QoS） objects was improved effectively. At the meantime, compared to the conventional GA, the improved GA increases convergence speed greatly. The proposed algorithm is suitable for embedded applications based on Rapidio network, and has strong application value.

Key words: Rapidio network, Genetic Algorithm (GA), Quality of Service (QoS), fitness function, algorithm coding