AG Kommunikationstheorie


Energy based planning and dimensioning for multi-tier heterogeneous mobile networks


Due to burgeoning growth of wireless devices and exponential increase in mobile traffic demand, operating modern heterogeneous cellular network efficiently is a huge concern. On the other side, in recent years, due to advancement in computation power and storage technology as well as embracing emerging datascience techniques, telecom operators can effectively store and analyze huge databases like Call Detail Records etc. These databases give insights into various user and network related information such as type of traffic accessed, hotspot locations, traffic variation etc. Hence operators can potentially further optimize the network by harnessing full potential of information available from these big datas. Though there are plenty of information one can get, one interesting idea is, classification of users based on similarities or as communities and optimize the network accordingly. In this thesis, to get the insight into advantages of user classification, at first we model the heterogeneous wireless network with two community of users which differ in rate requirement using the stochastic geometry and evaluate the network in terms of downlink rate coverage. Then we analyse small cell sleep strategies and show that rate-based user classification in devising sleeping strategies provides better energy consumption and fair resource allocation compared to obliviously allocating maximum rate for all the users. Further we introduce the idea of user profile based offloading and evaluate the network.

zurück zur Terminübersicht