The Research On MIMO Ergodic Capacity With Co-Channel Interference

MIMO has attracted great attention for providing a higher spectral efficiency and due to that,it becomes one of the most important techniques in the 4G.Capacity is one of the most elementary problem which denotes the maximum transmit rate with an arbitrarly low error rate.For fast fading channels,the erogodic capacity is an important metric to measure the rate.In modern wireless communication systems,especially in cellular systems,transmitters in the different clusters transmit the signals in the same frequency band concurrently to ensure a high spectral efficiency,and it inevitably makes the communication suffered by the co-channel interference,which deteriorate the rate severely.So,the research on MIMO ergodic capacity with co-channel interference is significantly important.The thesis presented here lays the interests in the scenery of both single link and multi-links.In the scenery of single link.At first,further research of Ye's paper(2005) is done.Then the channel is assumed following Nakagami-m distribution and investigation on ergodic capacity under different channel coefficients via Monte-Carlo simulations is conducted.At last,the extent of approximation between 2 kinds of channel model is observed when the substitution is taken of Nakagami-m fading for Rician fading.The conclusion is drawn that in the scenery of single link,ergodic capacity increases as m increases and it goes down when mi increases.If ergodic capacity is considered as the metric,the simulation results show that the 2 kinds of channel approximate with each other very well when Rician factor K is small,but the difference between the 2 kinds of ergodic capacity is significant as K or m increases.The research on the sum ergodic capacity of several links which interfere with each other.In this part,further research based on Blum(2003) and Ye(2003)'s investigation is done,firstly.The convergence rate of Nash Equililbrium algorithm and Global Gradient Projection method is shown via Monte-Carlo simulations,from which conclusions can be drawn.At last,a bound based on Rong and Blum's investigation is given.The conclusion from the results of Monte-Carlo simulations is that the convergence rate of Nash equilibrium increases as SNR increases or INR decreases and vice versa.The convergence rate of Global Gradient Projection method is very complex,but it is certain that 50 times' iterative can make the convergence rate arrive at an acceptable level.At last,the simple new algorithm of transmitting presented here is shown to have a gain over the independent water-filling algorithm at high INRs.