Range expansion (RE), when applied with inter-cell interference coordination (ICIC) may have significant benefits for improving the throughput and fairness of macrocell/picocell networks. In this paper, first, benefits of RE in the uplink and downlink are reviewed, emphasizing its definitive merit for uplink communications. It is shown that while RE also improves the fairness in the downlink, it may seriously degrade the total system capacity if no ICIC is utilized. Then, a novel cell-selection technique that capitalizes on subframe-blanking at the macrocell is introduced for improving the downlink capacity and fairness. System level simulation results show that the proposed cellselection method adaptively achieves a good balance between sum capacity and fairness, and can be used in place of biasbased cell-selection.

We propose multiple antenna transmission methods for fast fading channel conditions. Proposed methods allocate alternative multiple antenna transmission modes depending on the SNR, modulation order and Doppler frequency to increase reliability, i.e., decrease bit-error-rate (BER). A major difference of our approach from previous works is to allow different transmission modes during a single frame by considering the channel variations within the frame for high mobility scenarios. First method effectively makes use of available channel knowledge by allocating beamforming (BF) as long as channel knowledge is not outdated and switches to space frequency block coding (SFBC) afterwards. Approximate BER expressions for BF and SFBC that are functions of SNR, modulation order, Doppler frequency and initial channel knowledge are derived to be used as a decision metric for mode allocation throughout the frame. Second method adapts rate and transmission mode across symbols in a frame by considering that average channel power is monotonically decreasing as a function of time within the frame when multiuser diversity is exploited. First method performs as good as the better of BF and SFBC over all SNR values. Second method provides additional performance gain over the first method due to more efficient use of better channel conditions.

In wireless networks, average system through-put can be improved by scheduling transmissions according to instantaneous signal-to-noise ratio (SNR) values of the users assuming that channel fading is fast enough to maintain multiuser diversity. If the fading is slow, opportunistic multiple antenna approaches can introduce artificial fading in the system. These schemes require significant amount of SNR feedback. A novel channel estimation method for opportunistic beamforming is proposed based only on SNR feedback. More importantly, the effectiveness of the proposed channel estimation algorithm is verified using our multiple antenna testbed. The experimental results show that the channel estimation based beamforming can attain true beamforming gains.