Empirical evaluation of hash functions for multipoint measurements

摘要: A broad spectrum of network measurement applications demand passive multipoint measurements in which data from multiple observation points has to be correlated. Examples are the one-way delay or path that a packet takes through network. Nevertheless, due high rates and need for fine granular measurements, resource consumption can immense. Furthermore, depends on traffic network, usually is highly dynamic. Packet flow-selection methods provide solution reduce control measurements. In order apply such techniques selection processes synchronized. Hash-based deterministic based hash function computed selected parts content. This decision consistent throughout enables tracing between nodes. Because it introduce bias leads wrong estimation characteristics. this paper we define set quality criteria select investigate most suitable hash-based selection. We analyze 23 non-cryptographic 2 cryptographic functions. Experiments performed with real traces different networks. Based results recommend fast functions show low sample representative subset population.