Efficient Transmission Schemes And Resource Management For Interference Channels

In the multi-user network, communication activity is limited by the interferencecaused by each other. In order to improve the efficiency and quality in such aninterference environment, efficient management and scheduling of thecommunication resources become the focus of the research.In this dissertation, the author mainly investigated the management andscheduling of the resources (including time, frequency, space, power and code) in theinterference channel. Some main results are as follows,a) An improved rate-splitting scheme, so called distributed rate-splittingscheme (DRS), was proposed to approach the rate region boundary of the two-userGaussian interference channel. It is shown that the DRS scheme can be formulated asa non-cooperative game. Moreover, an iterative multiple waterlevels water-fillingalgorithm (IML-WFA) was developed to efficiently reach the Nash equilibrium of thegame. The existence of Nash equilibrium was established. The convergence of theIML-WFA was proved. Numerical examples show that the rate-tuples achieved by theIRS are very close to the boundary of the well-known HK regions.b) The author constructed a competitive game with an asymptotically optimalcompensation to approximate the optimization problem of sum-rate. Thecompensation can also be modeled as a sub-game whose Nash equilibrium can bereached by an iterative fashion. Then, a distributed power allocation algorithm callediterative multiple waterlevels water-filling algorithm (IML-WFA) was proposed toefficiently achieve the equilibrium of the game. The author also derived somesufficient conditions on the convergence of IML-WFA. Through simulation, theIML-WFA has a significant improvement of the performance over iterative waterfilling algorithm (IWFA) and achieves the close-to-optimal performance.c) The author considered a distributed interference avoidance problememploying frequency assignment in the Gaussian interference channel. The authordivided the common channel into several subchannels and each user chooses thesubchannel with less amount of interference from other users as the transmit channel.This mechanism can be modeled as a competitive game model, and a completelyautonomous distributed iterative algorithm called distributed interference avoidancealgorithm (DIA) is adopted to achieve the Nash equilibrium. Due to the self-optimum,DIA is a sub-optimal algorithm. Therefore, through introducing an optimalcompensation into the game, the author developed a compensation-based game model to approximate the optimal interference avoidance problem. Moreover, an optimalalgorithm called iterative optimal interference avoidance algorithm (IOIA) is proposedto reach the Nash equilibrium. The author analyzed the implementation complexitiesof the two algorithms and gave the proof on the convergence. The performance upperbound and lower bound were derived. The simulation results show that IOIA doesreach an optimal sum-rate and DIA performs a close-to optimal performance in theGaussian interference channel.d) The author put forward a novel joint time-frequency interferencealignment scheme in the frequency-selective Gaussian interference channel. Thefrequency-selective Gaussian interference channel is changed into a set of parallelGaussian interference channels. Then the interference alignment aligns theinterference not only in the time dimension but also in the frequency dimensionemploying a set of time-frequency alignment matrices. An algorithm for calculatingthe time-frequency alignment matrices was presented based on the convexoptimization. By the joint time-frequency interference alignment, the author canshorten the alignment time (i.e., decoding delay) while the performance of thealignment is not affected. Computer simulations have evaluated an excellentperformance of the proposed scheme at both decoding delay and alignmentperformance. The joint time-frequency interference alignment provides us with atradeoff between time extension and frequency extension.e) The author studied the joint power control, spatial and temporal resourceallocation scheme in the MIMO interference channel. At the transmitter, the authorintroduced a space-time pre-process matrix to process the transmit signal andde-correlate the receive signal by a linear detector matrix. The transmitters andreceivers were optimized to achieve a target signal-to-interference-plus-noise-ratio(SINR) with minimizing the total transmit power in the MIMO interference channel.By an energy-efficient perspective, the author presented an iterative algorithm torealize the proposed scheme. Simulation results show that our algorithm offerssavings in total transmit power over the distributed interference alignment algorithmwhile the performance is the same as that of the distributed interference alignment andour algorithm can achieve higher SINR than that of the distributed interferencealignment algorithm when given total transmit power constraint.