Crow Search Algorithm (CSA) can be defined as one of the new swarm intelligence algorithms that has been developed lately, simulating the behavior of a crow in a storage place and the retrieval of the additional food when required. In the theory of the optimization, a crow represents a searcher, the surrounding environment represents the search space, and the random storage of food location represents a feasible solution. Amongst all the food locations, the one where the maximum amount of the food is stored is considered as the global optimum solution, and objective function represents the food amount. Through the simulation of crows’ intelligent behavior, the CSA attempts to find the optimum solutions to a variety of the problems that are related to the optimization. This study presents a new adaptive distributed algorithm of routing on CSA. Because the search space may be modified according to the size and kind of the network, the algorithm can be easily customized to the issue space. In contrast to population-based algorithms that have a broad and time-consuming search space. For ten networks of various sizes, the technique was used to solve the shortest path issue. And its capability for solving the problem of the routing in the switched networks is examined: detecting the shortest path in the process of a data packet transfer amongst the networks. The suggested method was compared with four common metaheuristic algorithms (which are: ACO, AHA, PSO and GA) on 10 datasets (integer, weighted, and not negative graphs) with a variety of the node sizes (10 - 297 nodes). The results have proven that the efficiency of the suggested methods is promising as well as competing with other approaches.
-norm Least Mean Square (