Caretta Cartozo Cécile, De Los Rios Paolo
Institute of Theoretical Physics, Ecole Polytechnique Fédérale de Lausanne, CH-1015, Lausanne, Switzerland.
Phys Rev Lett. 2009 Jun 12;102(23):238703. doi: 10.1103/PhysRevLett.102.238703. Epub 2009 Jun 11.
Navigability of networks, that is, the ability to find any given destination vertex starting from any other vertex, is crucial to their usefulness. In 2000 Kleinberg showed that optimal navigability could be achieved in small world networks provided that a special recipe was used to establish long range connections, and that a greedy algorithm, that ensures that the destination will be reached, is used. Here we provide an exact solution for the asymptotic behavior of such a greedy algorithm as a function of the system's parameters. Our solution enables us to show that the original claim that only a very special construction is optimal can be relaxed depending on further constraints, such as, for example, cost minimization, that must be satisfied.
网络的可导航性,即从任何其他顶点出发找到任何给定目标顶点的能力,对其效用至关重要。2000年,克莱因伯格表明,只要使用一种特殊方法来建立长程连接,并使用一种能确保到达目标的贪婪算法,就能在小世界网络中实现最佳可导航性。在此,我们给出了这种贪婪算法的渐近行为作为系统参数函数的精确解。我们的解使我们能够表明,最初认为只有非常特殊的构造才是最优的这一观点,可以根据必须满足的进一步约束条件(例如成本最小化)而放宽。