The Design Of Limited Feedback Scheme For CoMP In LTE-advanced

LTE-Advanced (LTE-A) system is the subsequent evolution of long term evolution (LTE).Compared to3G, in addition to improving the spectral efficiency and the peak rate, the LTE-Asystem has proposed several key techniques. Among these key techniques, coordinated multi-point(CoMP) technology has drawn significant attention recently, due to the ability that improves thecell-edge users’ performances.To achieve cooperation among multiple wireless access points, it is necessary to feed backchannel state information (CSI). But this will bring tremendous feedback overhead. Therefore, inorder to improve the usefulness of CoMP technology, it’s important to propose a more realisticfeedback strategy that can reduce feedback overhead.This paper mainly investigates a rational strategy for feeding back CoMP channel information.The strategy can reduce feedback overhead as much as possible without affecting system’squantization accuracy.Firstly, after comparing the three existing3GPP’s feedback mechanisms, the paper selects to useimplicit feedback mechanism which is more suitable for CoMP and has less feedback overhead. Inaddition, based on the researching about limited feedback technology, the paper chooses theprecoding scheme that based on codebook.Secondly, taking the dynamic ranges of the cooperative cell and the mobility of users intoaccount, this paper does not use the joint codebook, but cites a new codebook structure--per-cellcodebook, which has more flexibility and compatibility compared to the previous joint codebook.Thirdly, the paper studies the phase ambiguity that caused by independent codeword selectionstrategy, then cites the real and imaginary segments feedback strategy. The feedback strategy canreduce feedback overhead without performance degradation.Finally, based on the real and imaginary segments feedback scheme, we propose an improvedscheme which adds weights concept to the allocation of feedback bits so that limited feedback bitscan be allocated rationally on importance. Simulation results show that the improved scheme underthe same quantization accuracy of the system can reduce the feedback overhead effectively.