?1120339085?Submitted in total fulfilment of the requirements for the degree of Master in Computer TechnologyA User-Oriented Redundancy Elimination Strategy in WLANszhou xinshengSupervisorxue guangtaoS C S ES J T US, P.R.C? ? ? ? ?X? ? 13%?Twitter ? 50% ?Facebook ? 30%?100% - 200% ?WLAN ? WLAN ? WLAN? WLAN ? WLAN ? WLAN ? 10000 ? 24% ? i ? ?,?,?,? ii ?ABSTRACTA User-Oriented Redundancy Elimination Strategyin WLANsABSTRACTRecentyears,withthedevelopmentofmobileInternetandthespreadingofmobilesmart equipments(laptops, smartphones, tablets), WLAN is widely considered as thenext generation of networking and will replace wired network in the future. The dataproduced by smart equipments during daily life is becoming bigger and faster sincethe large volume of network files become increasing popular, such as videos, musicand pictures. At the same time, the global mobile network grows in an astonishingspeed, mobile Internet account for 13% of total networking data traffic. The trend oftransferring from traditional Internet service to mobile equipment are very clear now,forexample,morethan50%ofdatatrafficofTwittercomefrommobileequipmentsandsmartphonesaccountfor30%ofthetotalnetworktrafficinFacebook. Thedevelopmentof mobile network is beyond imagination, the overall network traffic grows at the ratefrom 100% to 200%.In the future, the development of mobile networking wont meet the growth ofmobile Internet traffic demand. As a supplement to mobile network, WLAN will playavitalroleinthemobileInternettimes. ButWLANwillmeetthesamechallengefacedby mobile network such as highly mobile services, high volume data transferring andmore intelligent smart services. We need more effective solutions to solve the problemraised by high volume traffic other than blindly deploy more WLAN equipment.Researcher has proved that there exists tremendous redundant traffic in wired net-work, but how to use this technology to eliminate redundant data in WLAN is still aquestion. How to reduce redundancy traffic and improve traffic efficiency have been a iii ?research hotspot. In this paper, we collected read traces from over 10000 WiFi usersfrom a company in Shanghai. Through intensive data analysis, we found that thereexists more than 24% redundancy traffic in WLAN. By digging the patterns of dataset, we proposed a user-oriented redundancy elimination strategy which can allocatedifferent cache size to different users adaptively. Simulation results shows that user-oriented redundancy elimination strategy can detected and eliminate redundant trafficeffectively.KEYWORDS: WLAN,smartequipments, redundancyelimination, cache allocation iv ?iABSTRACTiii?v?ix?ix?xi? ?11.1?. . . . . . . . . . . . . . . . . . . . . . . . . . . . .11.2?. . . . . . . . . . . . . . . . . . . . . . . . . . . . .31.3? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41.4? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51.5? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5? ?72.1?. . . . . . . . . . . . . . . . . . . . . . . . . . . . .72.1.1? . . . . . . . . . . . . . . . . . . . . .72.1.2?. . . . . . . . . . . . . . . . . . . . . .92.2? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102.2.1? . . . . . . . . . . . . . . . . . . . . . . . .112.2.2?. . . . . . . . . . . . . . . . . . . . . .13 v ?2.3?. . . . . . . . . . . . . . . . . . . . . . . . .152.4? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15? WLAN ?173.1?