Self-organizing fractional frequency reuse for femtocells using adaptive frequency hopping

In this paper, an adaptive and self-organizing spectrum allocation policy for ad hoc femtocells in two-tier networks based on usage of random frequency hopping is presented. As femto- and macrocells normally share the same frequency resources, the proposed approach is able to minimize InterCell Interference (ICI) in an adaptive manner. Each femto- and macrocell user chooses random carrier frequencies for transmission corresponding to a given probability density function (pdf). If the level of interference reaches a certain limit, which is typically the case for femtocell edge users, the probability density function of the carrier frequencies is changed in a way that it is quasi-orthogonal to the one of the macrocell users nearby. The changeover point between different probability density functions is realized in a self-organizing manner regarding the detected level of interference. Hence, the approach is able to adjust the trade-off between spectral efficiency and introduced interference during operation. We present analytical models for the Signal-to-Interference-and-Noise-Ratio (SINR), the Bit Error Ratio (BER) and the required transmission power of Orthogonal Frequency Division Multiple Access (OFDMA) femtocells applying adaptive random frequency hopping. The mobile radio channel is characterized by non-frequency selective Rayleigh fading and all results are verified by simulations implemented in MATLAB. In order to evaluate the improvements, the results are compared to systems using centralized orthogonal planning, systems using neither centralized nor distributed planning, and systems using static random frequency hopping. The comparison shows that adaptive random frequency hopping is capable of reducing the BER by a factor of 15 and the transmission power of femtocells by 17 dBm compared to systems where the macrocell user transmits in the same subbands as the femtocell user.