Institut for Forretningsudvikling og Teknologi

Ramjee Prasad

A novel buffer management for delayed delivery traffic in future networks

Publikation: Bidrag til bog/antologi/rapport/proceedingKonferencebidrag i proceedingsForskningpeer review

  • Sri Hanuma Chitti, Danmark
  • Neeli R. Prasad, Danmark
  • Knut Øvsthus, Ukendt
  • Ramjee Prasad

A lot of attention has been on developing the next generation networks to deliver the traffic in intelligent and scalable manner. With the development of Internet of Things (IoT), new and dynamic methods to access data are surfacing in mobile networks. One such approach is to send or receive data at a scheduled time where the users predefine the tasks. For example, the customer may wish to download/upload the data before a particular time and in such instance, the network intelligence will schedule the task at appropriate time. This paper considers the model of delayed traffic delivery and proposes an innovative buffer management algorithm to deal with it. By applying the proposed algorithm in TCP environment, significant amount of savings in packet loss and higher link utilization is observed. The results are compared with other queuing algorithms like drop-tail, random early detection (RED) and they promises the objective. It is expected that by deploying it in bigger scale, the overall network link utilization can be increased. In addition, some directions of future work are presented to embrace the scheme together with effective frameworks.

OriginalsprogEngelsk
TitelWireless Communications, Vehicular Technology, Information Theory and Aerospace & Electronic Systems (VITAE), 2014 4th International Conference on
Antal sider5
ForlagIEEE Signal Processing Society
Udgivelsesår1 jan. 2014
Artikelnummer6934470
ISBN (trykt)9781479946266
DOI
StatusUdgivet - 1 jan. 2014
Eksternt udgivetJa
BegivenhedWireless VITAE2014 - Aalborg, Danmark
Varighed: 11 maj 201414 maj 2014

Konference

KonferenceWireless VITAE2014
LandDanmark
ByAalborg
Periode11/05/201414/05/2014

Se relationer på Aarhus Universitet Citationsformater

ID: 171390107