The Simulation And Comparison Of Bloom Filter And Its Improved Algorithms In The Distributed Environment

Distributed system developed rapidly in recent decades, their all the timethrough the computer network to share resources and data interaction. Especially inrecent years, as cloud computing, and mobile Internet penetration, as well as thedevelopment of streaming video online, in real situations a very important issue is thenode to locate and retrieve resources.As an effective hash, said Bloom filter (Bloom Filter) uses a bit vector that datacollection and use of Hash function to effectively support search. It is very efficientand quickly determine whether a given set of elements.In a distributed application environment, the Bloom filter in the resourcelocation, route lookup and other aspects can be well applied. In recent years, the basicprinciples of Bloom Filter for the existence of a variety of improvements, which willinclude the full portfolio-type, K type and split type and other improvements. Theimproved algorithm can effectively alleviate the dynamic elements in a distributedenvironment, set up the problem and effectively reduce the rate to find misnomer.The content of this thesis work is through programming simulation to verify theBloom Filter and its improved algorithm in a distributed environment, theeffectiveness and performance indicators, through the comparison test error rates andother indicators that we compare a variety of system implementation, and follow-upwork discussed.In this paper, K type for the Bloom Filter algorithm, split-type Bloom FilterAlgorithm and Combined total efficiency of the algorithm Bloom Filter algorithm isverified through the simulation process we verify the correctness of severalalgorithms, and these algorithms compared.Bloom Filter is a data set using bit vectors and using Hash function to find datato support effective way. It could very well determine whether a particular element ofa given set.Split Bloom Filter-based Bloom Filter is an improvement, it can better alleviatethe distributed environment, a collection of elements of the dynamic growth rate dueto increasing problems to find a misnomer. As a new development, K-type sub-assemblies misnomer Bloom Filter can rate,vector space and determine the time average of the three indicators have betterbalance. A hybrid type of algorithm has certain advantages.From this, we see that, Bloom Filter and a variety of improved data collection ismainly used for large-scale search to find. Determine the error in allowing certainidentification of tolerance conditions, can effectively reduce the Hash storage spaceand search time determine.In this article focused on distributed application environments, in particularsplit-type Bloom filter can effectively alleviate the growing problem of space nodes.Although the local node also brought greater time and space overhead.The world is growing, the network is also at an alarming rate in the developmentof shared network resources and the efficiency of network resource sharingrequirements are also increasing, which search performance have becomeincreasingly demanding, I believe Bloom Filter algorithms will continue to beimproved or there is a better search algorithm, an algorithm always need a lot oftesting and validation before being widely used Er Suanfa verify the accuracy of theYaoqiu will Yuelai Yue Gao, Suo Yi algorithm Moni will verify this Fangmian Yewith greater development.Process generated by the random function through the hash function mappingrandom value to the hash table, hash table and then look through the many ways totest and improve the Bloom Filter algorithm misnomer rate.Programming with interactive interface allows users to set the hash table, datacollection, and the number of hash function and so on, and then drawn through thesimulation, error rate, said display to the user. This flexibility to simulate theexplosive growth of data (the data set by the user to change the size).Try this another way, to increase K-bit vector space, and the combination of thed a Hash function. This combination of K sub-type Bloom Filter, can flexibly selectthe variable according to the actual situation, the slow steady growth in the data canbe many scenarios in the application. We can be summarized in the misnomer rate,Hash vector space and determine the time to find balance among the three.