A network composed of high density femtocells coexisting with low density macrocells in an orthogonal frequency division multiple access (OFDMA) array is considered. Two approaches are investigated: (1) spectrum sharing where macrocells and femtocells can access the same subchannels, and (2) division of the spectrum where macrocells and femtocells access separate portions of the spectrum. In both approaches, the balance between the bandwidth used and the signal-interference ratio (SIR) is studied, since femtocells or macrocells access more subchannels. The number of subchannels used to maximize the average spectral efficiency of femtocells, subject to a minimum average spectral efficiency restriction for macrocells, is analyzed for both approaches, demonstrating that the performance with the division of the spectrum is better than the shared spectrum for The model considered.
Femtocellular networks, consisting of a conventional macro-cellular display and superposition of femtocells, forming a hierarchical cellular structure, constitute an attractive solution to improve the capacity and coverage of macrocells. However, inter- and intra-level interference in such systems can significantly reduce capacity and cause an unacceptably high level of interruption. This article addresses the problem of uplink interference in orthopedic division of multiple access frequencies (OFDMA) based on femtocell cells with partial cochanel deployment. First, we propose an interterritorial interference mitigation strategy without the power control of femtocell users forcing femto-interfering macrophot users to use only some dedicated subcarriers. Non-interfering macrocell users, on the other hand, can use dedicated subcarriers, or shared subcarriers that are also used by femtocell users. Subsequently, we propose subcarrier allocation schemes based on the auction algorithm for macrocell users and femtocell users, respectively, to independently mitigate intra-tier interference. The interference mitigation scheme proposed for femtocell networks offers a significant improvement in performance over existing methods, substantially reducing inter- and intra-level inferences in the system.