Linear Precoding and Power Allocation Optimization with Partial CSI under per BS Power Constraints for Cooperative MIMO TDD Systems

以部分通道狀態信息和基地台功率消耗限制下針對合作式多輸入多輸出分時系統設計的線性預編碼和功率分配最佳化

余書硯(Shu-Yan Yu)

指導教授 ： 馬席彬

國立清華大學/電機資訊學院/電機工程學系/碩士(2011年)

摘要

In downlink transmission, multicell cooperation is a powerful method to mitigate intercell interference, and transmit diversity gain is also improved through cooperation. When one of the BS suffers from large scale fading like shadow fading and fails to transmit the signal to users, the cooperative base stations (BSs) can substitute for transmitting the signal to those users. Obviously the cooperative BSs provide spatial diversity in the cooperative systems. It is especially good to improve the bit error rate (BER) performance when BSs suffer from large scale fading. However, those cooperative BSs need to share all the channel state information (CSI) in the downlink channel, and it would be a heavy loading of the system.
Now, we propose a method to lower that loading for two cooperative BSs in multiuser multiple input and multiple output (MIMO) downlink communication systems. The proposed
method only shares one real value to represent the complex channel matrix for each user and optimizes the power allocation through that value. The loading to share CSI is reduced to 1/2NrNt. BER of traditional case 1 is 10e−2 when signal to noise ratio (SNR) = 0 dB without cooperation, but BER of the proposed system can be reduced to 10e−4.5 when SNR = 0 dB through cooperation. The proposed system has the better BER performance than traditional systems without cooperation, so it could save more transmitted power in the same BER performance
and reach higher cell coverage.
In simulation results, two BSs in the proposed system and compared systems use the same MIMO precoding algorithm to maximize the effective channel diagonal elements or directly transmit the signal without precoding. The differences between these systems are cooperation conditions and power constraints. To enhance the BER performance, we use maximum likelihood detection (MLD) at the receivers in the proposed system, and the compared systems equitably use MLD at the receivers.
Simulations are based on the MIMO Rayleigh fading channel model with white Gaussian noise. The elements in the channel matrix are assumed i.i.d. complex Gaussian random variables with zero mean and variances are equal to 0.5 per dimension. From the simulation results, we can see that the proposed system is especially good in BER performance. The proposed system has 4dB SNR gain than conventional cooperative case 2 under the per BS power constraints when BER = 10e−3.

關鍵字

合作式通訊

並列摘要

在下行通道傳輸中，多小區合作是一種很有利的方去去減輕小區間的訊號干擾，透過合作傳送分集增益也可得到改善。當其中有一個基地台，因為大範圍衰減而無法傳送訊號給使用者時，合作的基地台就能代替傳輸訊號給這些使用者。這些合作的基地台顯然地在合作式系統裡提供了空間的多樣性。當基地台受到大範圍衰減影響時，這對位元錯誤率會有特別好的改善。然而這些合作的基地台必須分享彼此間的下行通道狀態訊息，這可能會是一個系統裡沉重的負荷。
現在我們提出了一個方法去降低這個負荷在兩個合作的基地台多個使用者多輸入多輸出下行傳輸系統。我們提出的方法，對每個使用者只需要去分享一個實值去代替整個複數通道矩陣，並且根據這個值去作功率分配最佳化。分享通道狀態訊息的負荷降低了2NrNt倍。當信噪比是0分貝時，沒有合作的traditional case 1位元錯誤率是10e-2，但當信噪比是0分貝時，透過合作提出的系統的位元錯誤率是10e-4.5。提出的系統在位元錯誤率的性能比傳統沒有合作的系統更好，所以這可以節省更多傳輸功率和小區覆蓋範圍。
在模擬中，兩個基地台在提出的系統裡和比較的系統裡，使用相同的多輸入多輸出預編碼演算法去極大有效通道矩陣中的對角線值，或是不使用預編碼直接傳送訊號。這些系統間的差異在合作的情況以及功率限制的條件。為了增進位元錯誤率的性能，在提出的系統裡，我們在接收端使用最大似然檢測法，比較的系統也平等地在接收端使用最大似然檢測法。
模擬是基於多輸入多輸出瑞利衰弱信道加上高斯雜訊。在矩陣通道裡的元素是相互獨立並具有相同分配的複數高斯隨機值，每個維度的平均值是0且變異數是0.5。從模擬結果中，提出的系統在位元錯誤率的性能特別好。當位元錯誤率是10e-3，在基地台功率限制下提出的系統比起conventional cooperative case 2有4分貝的信噪比增益。

參考文獻

[1] J.-C. Guey and L. Larsson, “Modeling and evaluation of MIMO systems exploiting

[2] S. Leach, “Singular value decomposition – a primer,” Unpublished Manuscript, Department

[4] Y. Jiang,W. Hager, and J. Li, “The geometric mean decomposition,” Linear Algebra and

application to precoder design for successive-cancellation detection,” IEEE Trans. Inform.

[6] W. Liu, S. Ng, and L. Hanzo, “Multicell cooperation based SVD assisted multi-user