Adaptive geolocation based interference control for hierarchical cellular network with femtocells

This thesis presents adaptive interference control methods to mitigate undesirable interference effects from femtocells to macrocell users in hierarchical cellular networks. The study in this thesis begins by quantifying the deterioration in performance experienced by macrocell users on the downlink in a simulated 3G/CDMA environment. Our baseline results show that the median deterioration in signal-to-interference plus noise ratio (SINR) observed for the macrocell users may be up to 10dB and the outage probability increases by large extend. In the next part of study, we propose interference mitigation schemes - ‘Proximity Based Iterative’ (PBI) scheme and ‘Adaptive Interference Scaling’ (AIS) scheme to adjust femtocell power to reduce femtocell interference effect on macrocell users. We show that previously studied mechanisms like the load-spillage, utility based power adaptation usually require relatively high system overhead due to over-the-air signalling for estimation of interference. Proposed PBI and AIS schemes avoid such over-the-air signalling and make use of geo-location information and backhaul signalling for the femtocell interference estimation. These schemes achieve power re-distribution by scaling power uniformly across femtocells, while allowing the network operators to set desired target data rates. Results from simulations show that the PBI and AIS schemes are able to increase the number of macrocell users achieving chosen target data rates by up to 158% when compared with the value when femtocell transmission power is at maximum. However, in case of the PBI scheme, results shows that 25% of femtocell users may receive rates below the target rate. The AIS scheme provides an improvement over the PBI scheme by adjusting femtocell power according to the interference contribution by each femtocell. Thus, AIS achieves better performance and only up to 12:2% of femtocell users receive rates below the target rate. This study concludes with parametric evaluation of system throughput as a function of both macrocell and femtocell user densities. Qualitative results are provided to support the conclusion.M.S.Includes bibliographical referencesby Shweta Sagar