Post on 13-Aug-2020
2019c9� $ Ê Æ Æ � 123ò 13Ï
Sept., 2019 Operations Research Transactions Vol.23 No.3
DOI: 10.15960/j.cnki.issn.1007-6093.2019.03.004
Ã�Ï&XÚ�O¥�ü�`z¯KÚ�'`z�{∗
4æ¹1,†
Á� Ã�Ï&XÚ�O¥�Nõ¯K�ï��`z¯K. ��¡, ù`z¯K~~ä
kpÝ���55, ���¹eJu¦); ,��¡, §�qkg��AÏ(�, ~XÛ¹
�à5!�©5�. |^`z��{(ܯK�AÏ(�¦)Ú?nÃ�Ï&XÚ�O¯K
´Cc5Æâ.ïÄ�9:. �©:?ØÃ�Ï&XÚ�O¥�ü�`z¯KÚ�'`z
�{, �)õ^rZ6&�����OKe�éÜDÑ/�ÂÅå¤/�OÚõÑ\õÑÑ
(Multi-Input Multi-Output, MIMO) uÿ¯K, Ì�0�y�`zEâ(ܯK�AÏ(
�3¦)Ú?nþãü�¯K��#?Ð.
'�c ��½tµ, Cþ�O�`z, õ^rZ6&�, éÜDÑ/�ÂÅå¤/�O,
O�E,5, ;tµ, MIMO uÿ, Ã�Ï&XÚ
¥ã©aÒ O221, TN92
2010 êÆ©aÒ 65K05, 90B18, 90C22, 90C30, 90C90
Two optimization problems in wireless communicationsystem design and related optimization methods∗
LIU Yafeng1,†
Abstract Many problems arising from wireless communication system design canbe formulated as optimization problems. On the one hand, these optimization problemsare often non-convex and highly nonlinear and thus are difficult to solve; on the otherhand, these problems have their own special structures such as (hidden) convexity andseparability. Recently applying mathematical optimization methods to solve/deal withthese problems while judiciously taking care of their special structures is a hot researchtopic. This (survey) paper aims to introduce two optimization problems in wirelesscommunication system design, max-min fairness linear transceiver design problem andMIMO detection problem, and related optimization methods. This paper will focus on theabove two problems and overview recent advances of applying mathematical optimizationtechniques to solve/deal with them by exploiting their special structures.
Keywords semi-definite relaxation, alternating optimization, multiuser interferencechannel, transceiver beamforming design, computational complexity, tight relaxation,MIMO detection, wireless communication system
ÂvFÏ: 2019-03-01
*Ä7�8:I[g,�ÆÄ7 (Nos. 11688101, 11671419, 11631013, 11571221),�®½g,Ä7:;� (No. L172020), I[uU��Æ�êâú�ÑÖ²��M#A^«��8 (No. 2016-999999-65-01-
000696-01)
1. ¥I�Æ�êÆ�XÚ�ÆïÄ�, O�êÆ��Æó§O�ïĤ, �Æ�ó§O�I[:¢�¿,�® 100190; State Key Laboratory of Scientific and Engineering Computing, Institute of Computa-
tional Mathematics and Scientific/Engineering Computing, Academy of Mathematics and Systems Science,
Chinese Academy of Sciences, Beijing 100190, China
† Ï&�ö E-mail: yafliu@lsec.cc.ac.cn
48 4æ¹ 23ò
Chinese Library Classification O221, TN92
2010 Mathematics Subject Classification 65K05, 90B18, 90C22, 90C30, 90C90
0 Ú ó
�`z[1-3]?Øûü¯K��ZÀJ, �EϦ�Z)�O��{, ïÄùO��
{�nØ5�9¢SO�Ly. �`z3I�!²L!7K!ó§!�Ï!+n!êâ©
Û!ÅìÆS!<ó�U�Nõ+�kX2��A^. Ã�Ï&Eâ[4-6]Ì�|^Ã�>
^Å5DÑ&E.y�¿Âe�Ã�Ï&©u 19V 90c�¿�|<ê�Z (Guglielmo
Marconi)��X�>^Å¢�,¦¤õ/|^>^Å¢yÃ�êâDÑ. 1948c, �à
(Claude Elwood Shannon) uLK�“A Mathematical Theory of Communication”�Ø
©, ïáÏ&�êÆnØ[7]. A�c5, <�Øä²{X#�Ã�Ï&�{��)ÚC
�, ¿�ÉXõ«õ��Ã�Ï&ÑÖ. Ã�Ï&Eâ�uÐ4�/Uõ<��Ñ1!
Ï&Ú©zDÂ��ª,�<�Jø=�B$�ÑÖ.Ã�Ï&Eâ�uÐ�´�cI
[ÔÑuÐOy��|¤Ü©. �©:?ØÃ�Ï&XÚ�O¥�`z¯KÚ`z
�{, Ì�0�y�`zEâ(ܯK�AÏ(�3Ã�Ï&XÚ5U`z�ü�Y~
9Ù�#?Ð.
Ã�Ï&XÚ�O¥�Nõ¯K8�(.Ñ´`z¯K, �$^`z�{\±¦),
'XÏ&�ä�ÿÀ(��O!&Ò�?)èL§!�`] ���. 8c,�éuÏ&
Eâ3y¢)¹¥�%ÇuÐ,Ï&XÚ`z�êÆnØ��{w��é¢�,3,�
¡®²¤�K�ÙuÐÚA^�'�Ï�. Ã�Ï&XÚ�O¥�éõ¯K~~�ï�
��kAÏ(���à��5�å`z¯K.��¡,ù`z¯K~~äkpÝ���
55,���¹eJu¦);,��¡,§�qkg��AÏ(�,~XÛ¹�à5!�©
5!DÕ5�. |^`z��{(ܯK�AÏ(�¦)Ú?nÃ�Ï&XÚ�O¯K,
C��c²þzcÑkù�¡�ó�¼ IEEE &Ò?nƬ½ö IEEE Ï&Ƭ�ZØ
©ø. .Ͷ`z;[!\<��[�Æ��¬!&Ò?nÚÃ�Ï&º?Ïr IEEE
Transactions on Signal Processing c?Ì?Û���Ç�Ì�ïÄ��´`z�{��
O!©Û9Ù3Ã�Ï&Ú&Ò?n¥�A^.Û���Ç3þã����´a�ïÄ
¤J, l �¼� 2004!2009!2011 c IEEE &Ò?nƬ�ZØ©ø, 2015 c IEEE
Signal Processing Magazine �ZØ©ø, 2011 cî³&Ò?nƬ�ZØ©ø, Ú 2011
cISÏ&�¬�ZØ©ø. `z�{3Ã�Ï&XÚ�O¥��5dd����.
�©X0�y�`zEâ(ܯK�AÏ(�3Ã�Ï&XÚ5U`z�ü�Y
~9Ù�#?Ð. ·�ò31 1 !0�õÑ\õÑÑ (Multi-Input Multi-Output, MIMO)
Z6&�¥����OKe�éÜDÑ/�ÂÅå¤/�O¯K9Ù�#ïÄ?Ð; 31
2!0� MIMOuÿ (MIMO Detection)¯K9Ù�#ïÄ?Ð.�©ýul`z��
Ý"ÀÃ�Ï&XÚ�O,kXeü�ý:.�©�1��ý:´'uÃ�Ï&XÚ
�O¥`z¯KO�E,5��x[8]. O�E,5�±�x¤�įK���J´§Ý,
´)ûÚ©Û¢S¯K�'�Ú½. �©�1��ý:´�â¯K�AÏ(��Ok
���{?nÚ¦)�A�Ã�Ï&XÚ`z�O¯K, ¿é�'�{?1nØ©Û�.
'uÃ�Ï&XÚ`z�O�nã©Ù, �ë�©z [9-16].
ÎÒ�½: �©^��çNi1L«��þ, ��çNi1L«Ý. �½Ý A,
3Ï Ã�Ï&XÚ�O¥�ü�`z¯KÚ�'`z�{ 49
AT L«§�=�, A† L«§��Ý=�, A−1 L«§�_, Ai,j L«§�1 (i, j) ��
�. �q�ÎÒ�·^u�þ. �½�þ x = [x1, x2, · · · , xK ]T, ‖x‖p ,
∑Kk=1 |xk|p L«
§� p �ê, Ù¥ p ∈ {2,∞}; Diag(x) L«d�þ x )¤�é�Ý. �½ (·��ê
�) HermitianÝ AÚ B, A � 0L« A´����½Ý; A •BL« AB �,,=∑i
∑j Ai,jBj,i; A⊗B L«§�� Kronecker ¦È. ��, �©^ e L«·��ê�� 1
�þ, ^ I L«·��ê�ü Ý, ^ 0 L«·��ê�� 0 ݽ�þ.
1 ����OKeéÜDÑ/�ÂÅå¤/�O
1.1 `z�.
õ^rZ6&�Xã 1 ¤«. 3õ^rZ6&�¥, kõ�DÑàÚõ��Âà. z
�DÑàÚ�ÂàCkØÓêþ�U�.z�DÑàF"�éA��Âàux&Ò.AO
/,�DÑà 1 (TX1)��Âà 1 (RX1)ux&Ò�Ó�, TX1 ��Ù¦��Âà�5
Z6;�L5,� RX1 � TX1 DÑ&Ò�Ó�,��Â�Ù¦DÑàDÑ�&Ò,ùé
u RX1 5`�´Z6. ù=´õ^rZ6&�.
ã1 õ^rZ6&�«¿ã
b�XÚ¥k K �^r, ·�^ K = {1, 2, · · · ,K} L«XÚ¥¤k^r�8Ü. �
Äü1ÅÏ&XÚ,=1 k �DÑà3�����ux��&Ò sk ∈ C�éA��Âàk. ·�^ Hkj ∈ CMk×Nj L«1 j �DÑà�1 k ��Âà�&�Ý, Ù¥ Mk Ú
Nk ©OL«1 k ��ÂàÚ1 k �DÑàCk�U��ê, @o�Âà k �Â�&Ò
�
yk = Hkkvksk +∑j 6=k
Hkjvjsj + zk,
Ù¥ vk ∈ CNk×1 L«DÑà k ¦^�DÑÅå¤/�þ, zk ∈ CMk×1 ´\5pdxD
(, Ñl��©Ù CN (0, σ2kI). - uk ∈ CMk×1 L«�Âà k ¦^��ÂÅå¤/�þ,
50 4æ¹ 23ò
K�Âà k ²L�5?n±����&Ò�
sk = u†kyk = u†kHkkvksk +∑j 6=k
u†kHkjvjsj + u†kzk.
òZ6w¤D(, @o1 k �^r�&ZD' (Signal-to-Interference-plus-Noise-Ratio,
SINR) �L«�
SINRk =|u†kHkkvk|2
σ2k‖uk‖2 +
∑j 6=k
|u†kHkjvj |2, k ∈ K. (1.1)
�âXÚ¥DÑàÚ�ÂàCk�U��ê,�òõ^rZ6&��Xe©a:�X
Ú¥z�DÑàÚ�ÂàÑCkõ�U��, = Mk > 2, Nk > 2 (k ∈ K) �, d&�¡
� MIMO Z6&�, Ù SINR L�ªXª (1.1) ¤«; �XÚ¥z�DÑàCkõ�U
�, z��ÂàCk 1 �U�, = Mk = 1, Nk > 2 (k ∈ K) �, d&�¡�õÑ\üÑÑ
(Multi-Input Single-Output, MISO) Z6&�.
3MIMOÚMISOZ6&�¥,z�DÑà�õÇ�å�L«� ‖vk‖2 6 pk, k ∈ K,Ù¥ pk L«1 k �DÑà�õÇþ�. �XÚ¥z��ÂàCkõ�U�, z�DÑà
Ck 1 �U��, = Mk > 2, Nk = 1 (k ∈ K) �, d&��üÑ\õÑÑ (Single-Input
Multi-Output, SIMO) Z6&�.
��y^r�m�ú²5, ·��Ä MIMO Z6&�¥��� SINR ��z (�
du���Ç��z) ¯K:
max{u,v}
G(u,v) , mink∈K{SINRk(uk,v)}
s.t. ‖uk‖2 = 1, ‖vk‖2 6 pk, k ∈ K,(1.2)
Ù¥ SINRk(uk,v) dª (1.1) �Ñ. 5¿�1 k �^r� SINR L�ª (1.1) Ø�6u
{uj}j 6=k , ¤±·�^ SINRk(uk,v) L«1 k �^r� SINR. Ú\9ÏCþ SINR =
mink∈K {SINRk} , þã�� SINR ��z¯K (1.2) �±�d/��
max{u,v}
SINR
s.t. SINR 6 SINRk, ‖uk‖2 = 1, ‖vk‖2 6 pk, k ∈ K.(1.3)
1.2 �#?Ð
�z��ÂàCk 1 �U� (Mk = 1, k ∈ K) �, = MISO Z6&�: ©z [17] J
ÑÏL��z_&Z'�\�Ú5%C¯K (1.2), ¿y²�±ÏLÀ�·��\�Xê
¼�¯K (1.2) ��`), dó�¼� 2007 c IEEE &Ò?nƬ�ZØ©ø; ©z
[18,19] JÑõ�ª�m�{�¦�¯K (1.2) ��Û�`).
�z�DÑàCk 1 �U� (Nk = 1, k ∈ K) �, = SIMO Z6&�, ·�Äg��
Xeõ�ª�m�)�(J, ù�(JÌ�´du¯K¥Û¹�à5[20]. duÙy²´
�E5�, ¤±Ó���ѯK (1.2) �DÑàk 1 �U���õ�ª�m�{; �
�©z [21].
½n 1.1 �z�DÑàCk 1 �U� (Nk = 1, k ∈ K) �, �� SINR ��z¯K
(1.2) �3õ�ª�mS¦).
3Ï Ã�Ï&XÚ�O¥�ü�`z¯KÚ�'`z�{ 51
Ó�, ·��Äg��DÑàÚ�ÂàÑCkõ�U��¯K (1.2) �E,5(
J[23]: =�z�DÑàCk�U�ê�u½�u 2(3), z��ÂàCk�U�ê�u½
�u 3(2) �, �ä�½ SINR 8I��15¯K´r NP-J�. 'uDÑàÚ�ÂàC
k�U�êÑ�u 2 �¯K (1.2) �E,5, ù�¯K��] �û. �C, ·�£�
ù�¯K; ��©z[22].
½n 1.2 �½¯K (1.3) ¥ SINR 8I ζ. �z�DÑà/�ÂàCk�U�ê�
u½�u 2 �, �ä�½ ζ ��15¯K´r NP-J�. Ïd, �`z¯K (1.2) �´r
NP-J�.
L 1 ¥o( MIMO Z6&��� SINR ��z¯K (1.2) 3DÑàÚ�ÂàCk
ØÓU�ê��O�E,5. lL 1 ¥·���, ¯K (1.2) �E,5�XDÑàÚ�Â
àCk�U�ê8�Cz Cz. ��5ù, O\U�ê8�±JpÏ&XÚ�5U, �
���A`z¯K�¦)�5(J.
L 1 ����OKeéÜDÑ/�ÂÅå¤/�O¯K (1.2) �O�E,5hhhhhhhhhhhhhhhh�Âà
DÑàNk = 1 Nk > 2
Mk = 1 õ�ª�) õ�ª�)[18,19]
Mk > 2 õ�ª�)[21] r NP-J[22]
·��JÑü�Cþ�O�`z�{¦)�� SINR ��z¯K (1.2): 1���
{´°(Ì��Iþ,�{ (Exact Cyclic Coordinate Ascent Algorithm, ECCAA), z�Ú
S�°(¦)f¯K;1���{´�°(Ì��Iþ,�{ (Inexact Cyclic Coordinate
Ascent Algorithm, ICCAA), Ù¥kf¯K�°(¦). Cþ�O�`z�{�Ä�g
�´Äk·�/ò¯K�Cþ©¬, ,� (°(½ö�°() �O/3��¬�m?1`
z/S�, zg`z/S���#Ù¥�¬Cþ, Ù¦¬Cþ�½ØC.
AO/, éu¯K(1.2), òÙCþ©�ü¬un = (un1 , · · · ,unK)Úvn = (vn1 , · · · ,vnK),
Ù¥ n > 0 L«S��I. �½ vn, Ï� SINRk =� uk �', ¤±�±ÏL¦) K �
�¯K
max{uk}
|u†kHkkvnk |2
σ2k‖uk‖2 +
∑j 6=k
|u†kHkjvnj |2
s.t. ‖uk‖2 = 1
(1.4)
��¯K (1.2) � vn �½��). ½Â
Mk(v) =
∑j∈K
Hkjvj (Hkjvj)†
+ σ2kI
−1 , k ∈ K,K¯K (1.4) ��`) unk ��5��þ�Ø� (Linear Minimum Mean Square Error,
LMMSE) �ÂÅå¤/�þ
unk = unk/‖unk‖, uk = Mk(vn)Hkkvnk . (1.5)
52 4æ¹ 23ò
�½ un, ¯K (1.2) 'uCþ v C�Xe MISO Z6&�¥�`z¯K:
max{v}
mink∈K
{| (unk )
†Hkkvk|2
σ2k +
∑j 6=k | (unk )
†Hkjvj |2
}s.t. ‖vk‖2 6 pk, k ∈ K.
(1.6)
dL 1 ��, ¯K (1.6) �±3õ�ª�mS¦��Û�`).
�{ 1.1 ECCAA
Ú 1 Щz: �½Ð©: v0 ÚÊ�OK ε. - n = 0.
Ú 2 O� un: �½ vn, O��`� LMMSE �ÂÅå¤/�þ un (�ª (1.5)).
Ú 3 O� vn+1: �½ un, ¦)¯K (1.6) ���`�DÑÅå¤/ vn+1.
Ú 4 ª�u�: �G(un,vn+1)−G(un,vn−1)
max {G(un,vn−1), 1}6 ε (1.7)
÷v�, ª��{; ÄK, � n := n+ 1, =Ú 2.
���¹e, =¦z�f¯KÑ°(¦), E,ØU�yCþ�O�{�Âñ5[23].
8I¼ê�1wÜ©±9�å¥Cþ��©5éu�{�Âñ5�~�. N´�E�
~y²,�8I¼ê�1wÜ©½ö�å¥�CþØ�©�,Cþ�O�{�)�:��
UÂñ���ÿÂ�:.¦+¯K (1.2)�8I¼ê´�1w��8I¼ê¥Cþ uÚ
v ÍÜ3�å, ·�|^Cþ uk �Ñy3 SINRk ¥�AÏ(�y²þã�O�`z
�{ ECCAA �)�:�Âñ�¯K (1.2) ��� KKT (Karush-Kuhn-Tucker) :[24].
½n 1.3 - ECCAA �{¥ ε = 0. ECCAA �)�:� {(un,vn)} ½ª�u¯K(1.2) ��� KKT :, ½§�?¿à: (u, v) ´¯K (1.2) � KKT :.
�?�ÚJp ECCAA �{�O�k�5, ·�3© [25] ¥JÑ���°(�
Cþ�O�`z�{¦)¯K (1.2), ÙzgS��°(/¦)'uCþ v �`z¯K.
äN/, © [25] JÑÏL¦)f¯K
max{v, θ}
θ
s.t.(unk )
†Hkkvk − θ√
σ2k +
∑j 6=k | (unk )
†Hkjvj |2
>√G2n, k ∈ K,
‖vk‖2 6 pk, k ∈ K,
(1.8)
�#Åå¤/�þ vn+1, Ù¥ G2n ´¯K (1.2) 3: (un,vn) ?�¼ê�. 5¿�¯K
(1.8) ´����I5y (Second Order Cone Program, SOCP) ¯K, ¤±�3õ�ª
�mS¦�Ù�Û�`). ��#DÑÅå¤/�þ vn+1, ECCAA I�¦)�X�
SOCP �15¯K, ICCAA �I�¦)�� SOCP ¯K, ¤±4�/ü$zg�#
Cþ vn+1 �E,5. � ECCAA �{aq, ·��y² ICCAA ��ÛÂñ5[25].
'u MIMO Z6&�éÜDÑ/�ÂÅå¤/�OÚõÇ��?�Ú�ó�, �)
�\E,�|µ (~X MIMO 2Â&�!É��ä!�VóXÚ)Ú�O (~XÄÕg·
AÜ�!ÄÕ-¹!ÄÕ-^r'é) �, �ë� [26-39]; 'uCþ�O�`z�{¦)
Ã�Ï&XÚ�O¯K��õó�, �ë� [34,37,38,40-43];'u�O�`z�{�nã
©Ù9Ù3��+�¥�A^, �ë� [3,44,45].
3Ï Ã�Ï&XÚ�O¥�ü�`z¯KÚ�'`z�{ 53
2 MIMO uÿ
2.1 `z�.
MIMO uÿ¯K´y�êiÏ&¥�Ä�¯K��[46,47]. êÆþ, MIMO &��Ñ
\ÑÑ'X�L«�
r = Hx∗ + v, (2.1)
Ù¥ r ∈ Cm ��Âà�Â���þ; H ∈ Cm×n �E�&�Ý; x∗ ∈ Cn L«DÑ�ý¢&Ò�þ; v ∈ Cm �\5D(. XJ¦^ M -Phase-Shift Keying (M -PSK) �N��
ª, @o x∗ �z��©þ x∗i Aáu��k�8Ü, =
x∗i ∈{
exp (iθ) | θ =2jπ
M, j = 0, 1, · · · ,M − 1
}, i = 1, 2, · · · , n,
Ù¥ i�JÜü ÷v i2 = −1.b�&�&E H®�, MIMOuÿ¯KF"ÄuÂ��
&Ò r ¡EDÑ�&Ò x. êÆþ, MIMO uÿ¯K�ï��
minx∈Cn
‖Hx− r‖22s.t. |xi|2 = 1, arg (xi) ∈ A, i = 1, 2, · · · , n,
(2.2)
Ù¥ arg (·) L«Eê�Ë�, A = {0, 2π/M, · · · , 2(M − 1)π/M} .�ª (2.1) ¥�D( v Ñlpd©Ù�, þã4�z�. (2.2) �du��q,�
O�.. N´�y, MIMO uÿ¯K (2.2) ��1:�ê� Mn. ¢Sþ, MIMO uÿ¯
K (2.2) ´r NP-J�[48], ¤±Ø�3õ�ª�m��{U¦)�Ù�Û�`) (Ø�
P=NP).
2.2 �#?Ð
Äu��½tµ (SemiDefinite Relaxation, SDR) ��{´¦)¯K (2.2) �~k�
��{��. Äu SDR ��{Ø�äkõ�ª�m���E,Ý, �3¢S¥�Ðy
ÑéÐ�uÿØèÇ. Äu SDR ��{�Ä�g�´ò¯K (2.2) ktµ��� (E½
ö¢�) SDR, ,�ÏL��{ (�Åz��{½öéuª��{) ���¯K���
�1).
�Ù¥·�X0�'u MIMOuÿ¯K (2.2) SDR��#?Ð.�Qã�B,·
�Ú\Xe�PÒ: -
Q = H†H, c = −H†r;
- s = [s1, s2, · · · , sM ]T ∈ CM , Ù¥
sj = cos
(2(j − 1)π
M
)+ i sin
(2(j − 1)π
M
), j = 1, 2, · · · ,M ;
- sR = Re(s), sI = Im(s), Ù¥ÎÒ Re(·) Ú Im(·) ©OL«�AEÝ/E�þ/Eê
�¢ÜÚJÜ.
Ú\ n× n EÝ X = xx†, ¯K (2.2) ��dC/�
minx,X
Q •X + 2Re(c†x)
54 4æ¹ 23ò
s.t. Xi,i = 1, i = 1, · · · , n,arg (xi) ∈ A, i = 1, · · · , n,X = xx†,
Ù¥Cþ x ∈ Cn, X ∈ Cn×n. ¯K (2.2) �DÚ (��þ��) E SDR ´
minx,X
Q •X + 2Re(c†x)
s.t. Xi,i = 1, i = 1, · · · , n,
X � xx†,
(2.3)
Ù¥Cþ x ∈ Cn, X ∈ Cn×n. duÙé�AÏ(�, S:�{�3 O(n3.5
)�mS¦�
SDR (2.3)��`)[49].ê�¢�L²,�¥)è (Sphere Decoding)�{[50-52]�',Äu
SDR��{3ØèÇ5UÚ�mE,5þk�~Ð�²ï[49]. ¥)è�{´¦) MIMO
uÿ¯K (2.2) �Ͷ��{��. lêÆ�Ý, ¥)è�{´�«¿©|^¯K (2.2)
AÏ(��©|½.�{, ¤±Ù�±�y¦�¯K (2.2) ��Û�`), �´¥)è�
{���Ú²þE,ÝÑ´�ê�[48,53].
'�ª (2.2) Ú (2.3) ��, ª(2.3) ò�å X = xx† tµ��å X � xx† ¿¿K
¤k�Ë��å arg (xi) ∈ A, i = 1, 2, · · · , n. N´�y
X � xx† ⇐⇒[
1 x†
x X
]� 0.
�JpþãE SDR ��þ, ·�3© [54] ¥JÑ3ÙÄ:þV\k�Ë��[55,56]. ä
N/,�â x�½Â, n�¥ xi A��u�þ s¥�,��©þ. ÏLtµþã�|Ü
�å, ·���Xe\r�E SDR:
minx,X,t
Q •X + 2Re(c†x)
s.t. Xi,i = 1, i = 1, · · · , n,X � xx†,
x = St,
At = en, t > 0,
(2.4)
Ù¥Cþ x ∈ Cn, X ∈ Cn×n, t ∈ RnM ,
S = In ⊗ sT, A = In ⊗ eTM . (2.5)
3ª (2.4) ¥, t = [tT1 , tT2 , · · · , tTn]T Ú ti = [ti,1, ti,2, · · · , ti,M ]T ∈ RM .
©z [54] y²þãE SDR ÚXe�¢ SDR �d
miny,Y
Q •Y + 2cTy
s.t. Yi,i + Yn+i,n+i = 1, i = 1, 2, · · · , n,
Y � yyT,
(2.6)
3Ï Ã�Ï&XÚ�O¥�ü�`z¯KÚ�'`z�{ 55
Ù¥Cþ y ∈ R2n, Y ∈ R2n×2n,
Q =
[Re(Q) −Im(Q)
Im(Q) Re(Q)
], c =
[Re(c)
Im(c)
], y =
[Re(x)
Im(x)
]. (2.7)
Äuª (2.6), ©z [54] ?�ÚJÑ��\r�¢ SDR. ½Â 3× 3 �Ý
Yi =
1 yi yn+i
yi Yi,i Yi,n+i
yn+i Yn+i,i Yn+i,n+i
, i = 1, 2, · · · , n (2.8)
Ú
Pj =
1
Re(sj)
Im(sj)
[ 1 Re(sj) Im(sj) ] , j = 1, 2, · · · ,M. (2.9)
�⪠(2.7) ¥ y �½Â, n�¥ª (2.8) ¥½Â� Yi A��u Pj (j = 1, 2, · · · ,M) ¥
�,��, =
Yi ∈ {P1, P2, · · · ,PM} , i = 1, 2, · · · , n.
ÏLtµþã�|Ü�å±9¿K�P{��å,©z [54]JÑXe�'u MIMOu
ÿ¯K (2.2) � SDR:
miny,Y,t
Q •Y + 2cTy
s.t. Yi =
M∑j=1
ti,jPj , i = 1, 2, · · · , n,
At = en, t > 0,
Y � yyT,
(2.10)
Ù¥Cþ y ∈ R2n, Y ∈ R2n×2n, t ∈ RMn, Q Ú c 3ª (2.7) ¥�Ñ, Yi 3ª (2.8) ¥�
Ñ, Pj 3ª (2.9)¥�Ñ, A 3ª (2.5)¥�Ñ.du Yi �é¡5,�å Yi =∑Mj=1 ti,jPj
��d/L«�Xe 5 ��5�å:
yi = tTi sR, yn+i = tTi sI , Yi,i = sTRDiag(ti)sR,
Yn+i,n+i = sTI Diag(ti)sI , Yi,n+i = sTRDiag(ti)sI .
(2.11)
'u MIMOuÿ¯K (2.2)� SDR,����nدK´Ù3�o^�e´;�,
=3�o^�e¦)�A� SDR�±���¯K��`).'u M = 2� SDR;5�
(Jéõ, �'(J�ë�©z [49,57-60]. AO/, ©z [60] y²Xe(Ø.
½n 2.1 éu M = 2, XJª (2.1) ¥�ëê H Ú v ÷v
λmin
(Re(H†H)
)> ‖Re(H†v)‖∞, (2.12)
Ù¥ λmin (·) L«�'Ý���A��, KXe¢ SDR 'u¯K (2.2) ´;�:
minx,X
Re(Q) •X + 2Re(c)Tx
s.t. Xi,i = 1, i = 1, 2, · · · , n,X � xx†.
(2.13)
56 4æ¹ 23ò
'u M > 2 � SDR ;5©Û�(J3©z¥�~��. AO/, ©z [60] ¥JÑ
��úm¯K, =éu M > 2 ��/, ´Ä�3'u¯K (2.2) ;� SDR. ·��C3
©z [54] ¥£�ù�úm¯K, y²ª (2.4) Ú (2.10) ¥� SDR Ñ´;�. ùü�
SDR �´'u�� MIMO uÿ¯K (2.2) ®��@�3nØþk;5�y� SDR. ä
N(JXe.
½n 2.2 éu?¿� M > 3, XJª (2.1) ¥�ëê H Ú v ÷v
λmin
(H†H
)sin( πM
)>∥∥H†v∥∥∞ , (2.14)
Kª (2.4) Ú (2.10) ¥� SDR 'u¯K (2.2) Ñ´;�.
'uª (2.4)Ú (2.10)¥� SDR±9½n 2.2��?ØXe.Äk, � M = 2�,
(2.9) ¥½Â�Ýòz�
P0 =
1 1 0
1 1 0
0 0 0
Ú P1 =
1 −1 0
−1 1 0
0 0 0
.¤± (2.11) ¥��5�åòz�
Yi,i = 1, yi = ti,1 − ti,2, Ú yn+i = Yn+i,i = Yn+i,n+i = 0.
�ÑÝ Y ¥�"¬, N´�yª (2.10)¥� SDRòz�ª (2.13)¥� SDR. ¤±�
M = 2 �, SDR (2.10) 3^�ª (2.12) ¤á�'u¯K (2.2) �´;�.
Ùg, ª(2.12) Ú (2.14) ¥� (¿©) ^�ÑkéÐ�ó§ÚÔn)º. ±ª (2.14)
�~: (2.14) ¥�� λmin
(H†H
)L�&��Ð�, �&�Ý H ¥���O'��
�, λmin
(H†H
)��, ^� (2.14) �J÷v; (2.14) ¥�� sin (π/M) �x M XÛK
�¯K�JÝ, M ��, ��ü�ÎÒ�ål 2 sin (π/M) ��, ^� (2.14) �J÷v;
(2.14) ¥��∥∥H†v∥∥∞ �xD(�K�, D( v é&�Ý H ���mK��� (=∥∥H†v∥∥∞ ��),^� (2.14)�J÷v.{ ó�,�N���ê��!&���!D( (é
&�Ý��mK�)��, (2.14)¥�^��J÷v;�� (2.14)¥�^��N´÷v.
2g, ·�?Ø8c���©¥�Ñ� SDR �m±9�©z [61] ¥ SDR �'X.
N´�y, SDR (2.10)' SDR (2.4)r, SDR (2.4)' SDR (2.3)Ú (2.6)r, SDR (2.3)
Ú (2.6) �d. ©z [61] ¥JÑ� SDR ´ÄuXe�*: éu x∗ ¥�?Û x∗i Ñ�3
ti ∈ RM ¦�Ù¥�k��©þ� 1 Ù{©þÑ� 0 �÷v x∗i = tTi s. Äud, ©z [61]
ò¯K (2.2) �d��
mint
tTQt + 2cTt
s.t. At = en, t > 0,
t ∈ {0, 1}Mn,
Ù¥
Q = STQS, c = STc, S =
(Re(S)
Im(S)
). (2.15)
3Ï Ã�Ï&XÚ�O¥�ü�`z¯KÚ�'`z�{ 57
Äuþã�dC/¿|^�þ t �AÏ(�, ©z [61] JÑXe�'u¯K (2.2) �
SDR∗:
mint,T
Q •T + 2cTt
s.t. At = en, t > 0,
T � ttT,
Ti,i = Diag(ti), i = 1, 2, · · · , n,
(2.16)
Ù¥Cþ t ∈ RMn, T ∈ RMn×Mn, Ti,i ∈ RM×M ´Ý T �1 i �é�¬. �����
å Ti,i = Diag(ti) �¦ Ti,i ´��é�Ý�Ùé���� ti. ÃØ´3/ª�´í�
�ªþ, SDR (2.10)Ú SDR (2.16)ÑØ��.k¿g�´,·�3©z [62]¥y²ùü
� SDR �m��d'X.
½n 2.3 SDR (2.10) Ú SDR (2.16) ��18�m÷vXe�'X:
STST = Y Ú St = y, (2.17)
Ù¥ S 3 (2.15) ¥�Ñ. ¤± SDR (2.10) Ú SDR (2.16) ´�d�.
þã^� (2.17)�y�A�ü� SDR�8I¼ê���.(ܽn 2.2Ú 2.3,·�
��: � (2.1) ¥�ëê H Ú v ÷v (2.14) �, SDR (2.16) 'u¯K (2.2) ´;�. ½
n 2.3 Ó��« SDR (2.16) ¥äk�½P{5. l (2.17) ¥�±��, p��m (T, t)
¥�k^&E3$��m (y,Y, t)¥Ñ���±.AO/, SDR (2.16)¥ÝCþ��
ê´ Mn×Mn SDR (2.10) ¥ÝCþ��ê´ 2n× 2n. ¤±¦) SDR (2.10) ò¬
'¦) SDR (2.16) 3ê�þk��õ (AO´� M ��u 2 �).
Äuþã\r� SDR,·��uÐk��©|½.�{¦)MIMOuÿ¯K (2.2).
'uù�¡�?Ð,�ë�©z [63].���J�´,3�^���¹e,~XÑ\ÚÑÑ
�ê8��½XÚ�&D'�$�,·�uÐ�Äu\r SDR�©|½.�{'¥)è
�{3¦)�Ýþkêþ?�Jp. ·�uÐ�©|½.�{�±í2^5¦)Xe�
����àE�g5y¯K:
minx∈Cn
1
2x†Qx + Re
(c†x)
s.t. `i 6 |xi| 6 ui, i = 1, 2, · · · , n,arg (xi) ∈ Ai, i = 1, 2, · · · , n,
Ù¥ |xi| L«Eê xi �ÌÝ, ui Ú `i (i = 1, 2, · · · , n) ´ 2n �¢ê÷v ui > `i > 0,
Ai (i = 1, 2, · · · , n) ´ n �lÑ/ëY�8Ü.
'u¦) MIMO uÿ¯K (2.2) �Ù¦k��{!�#?б9nã, �ë�©z
[47,48,50-53,63-68]; 'u��½5y/��½tµ9Ù3&Ò?nÚÃ�Ï&¥A^�n
ãÚ�#?Ð, �ë�©z [11,69,70],
∗SDR (2.16) Ú [61] ¥� SDR (Model III) ���«O3u SDR (Model III) ¥|^z��þ ti ¦Ú�u 1 �5�����Cþ.
58 4æ¹ 23ò
3 ( å �
Ã�Ï&XÚ�O¥�Nõ¯K8�(.Ñ´`z¯K. |^`z��{(ܯK
�AÏ(�¦)Ú?nÃ�Ï&XÚ�O¯K´Cc5Æâ.ïÄ�9:. �©X0
�üaÃ�Ï&XÚ�O¥�¯K, =õ^rZ6&�����OKe�éÜDÑ/�
ÂÅå¤/�O¯KÚ MIMO uÿ¯K, Ì�0�ùüa¯K�`z�.±9|^`
zEâ(ܯK�AÏ(�3¦)Ú?nþãüa¯K��#?Ð.
Cc5,15�(5G)£ÄÏ&XÚ®¤�.��SÆâ.Úó�.�ïÄ9:[71,72].
�'u�c�£ÄÏ&XÚ, 5G ØI�?�ÚJp^rN��Ç, �I�|±pë
��ݱ¢y£Äêâ�ÖþZ�O�!DÑà��Âà�Ï&�òü$ 90%!�¢y
ªÌ�ÇÚõÇ�ÇJ,���8I. �¢yd8I, 5G £ÄÏ&3Ã�DÑEâÚÃ
��äe��¡?1�·5�EâM#.ùC���¡U¢yþã8I,,��¡
��ïÄö�JÑ#��ä]Ô5�Ä:nØÚ'��ƯK. ù¯K¥éõÑ´
`z¯K, ~X�Ã��\� (Cloud Radio Access Network) ¥���`z[73,74]!�5
�U�� (Massive MIMO) ¥� 1 'Aý?è (One-Bit Precoding) �O[75,76]!ÑÖ
���ä��ä�¡ (Network Slicing)[77,78]!Ôé�¥��Å�\uÿ (Random Access
Detection)[79,80] �, )ûù¯K�Ì�(JÚ]Ô3u�Op�Ø%ûü`z�{é
Ï&] ?1Ä�©�±·A¯�Cz�Ã��¸! r±96þG�. o ó�, `z
�{òUY3 5G £ÄÏ&XÚ�O¥�üØ�O����Ú.
�u�öY²k�, �©J�k¢¦Ú ¹�?, �Öö1µ��.
��
�©¤�9�ö�ïÄó�Ñ´�Ü�ö�å�¤�. AO/, �©'u MIMO Z
6&�����OKe�éÜDÑ/�ÂÅå¤/�O�ó�Ì�´Ú�ö����Í
ôïÄÚ�l¥©�Æ (�e) Û���ÇÜ��¤�; �©'u MIMO uÿ¯K�
�½tµ�ó�Ì�´Úu�>å�Æ´§Æ¬Ü��¤�. 3d�¤kÜ�ö�¿L
«a�. �ö�/dŬAOa�¥I�Æ�êÆ�XÚ�ÆïÄ��æ��¬Ú`z
�K|�P�ÓÆ�.
ë � © z
[1] �æ�, �©Õ. �`znØ��{ [M]. �®µ�ÆÑ��, 1997.
[2] Nocedal J, Wright S J. Numerical Optimization [M]. New York: Springer Science+Business
Media, 2006.
[3] Bertsekas D. Nonlinear Programming [M]. Belmont: Athena Scientific, 1999.
[4] Tse D, Viswanath P. Fundamental of Wireless Communications [M]. New York: Cambridge
University Press, 2005.
[5] Cover T M, Thomas J A. Elements of Information Theory [M]. New York: John Wiley & Sons,
Inc., 2006.
[6] Goldsmith A. Wireless Communications [M]. New York: Cambridge University Press, 2006.
[7] Shannon C E. A mathematical theory of communication [J]. The Bell System Technical Journal,
1948, 5: 379–423, 623-656.
3Ï Ã�Ï&XÚ�O¥�ü�`z¯KÚ�'`z�{ 59
[8] Garey M R, Johnson D S. Computers and Intractability: A Guide to the Theory of NP-
Completeness [M]. San Francisco: W. H. Freeman and Company, 1979.
[9] Luo Z Q, Yu W. An introduction to convex optimization for communications and signal pro-
cessing [J]. IEEE Journal on Selected Areas in Communications, 2006, 24(8): 1426-1438.
[10] Luo Z Q. Applications of convex optimization in signal processing and digital communication
[J]. Mathematical Programming, 2003, 97(1-2): 177-207.
[11] Luo Z Q, Ma W K, So A M C, et al. Semidefinite relaxation of quadratic optimization problems
[J]. IEEE Signal Processing Magazine, 2010, 27(3): 20-34.
[12] Gershman A B, Sidiropoulos N D, Shahbazpanahi S, et al. Convex optimization-based beam-
forming [J]. IEEE Signal Processing Magazine, 2010, 27(3): 62-75.
[13] Bjornson E, Jorswieck E. Optimal resource allocation in coordinated multi-cell systems [J].
Foundations and Trends R© in Communications and Information Theory, 2013, 9(2–3): 113-
381.
[14] Hong M, Luo Z Q. Signal processing and optimal resource allocation for the interference channel
[M]//Academic Press Library in Signal Processing. Elsevier, 2014, 409-469.
[15] Gesbert D, Hanly S, Huang H, et al. Multi-cell MIMO cooperative networks: A new look at
interference [J]. IEEE Journal on Selected Areas in Communications, 2010, 28(9): 1380-1408.
[16] 4æ¹. Ã�Ï&¥��`] ©�)E,5©Û��{�O [J]. ¥I�Æ: êÆ, 2013, 43(10):
953-964.
[17] Schubert M, Boche H. Solution of the multiuser downlink beamforming problem with individual
SINR constraints [J]. IEEE Transactions on Vehicular Technology, 2004, 53(1): 18-28.
[18] Wiesel A, Eldar Y C, Shamai S. Linear precoding via conic optimization for fixed MIMO
receivers [J]. IEEE Transactions on Signal Processing, 2006, 54(1): 161-176.
[19] Liu Y F, Dai Y H, Luo Z Q. Coordinated beamforming for MISO interference channel: Com-
plexity analysis and efficient algorithms [J]. IEEE Transactions on Signal Processing, 2011,
59(3): 1142-1157.
[20] Xia Y. A survey of hidden convex optimization [J]. 2019. Available: https://arxiv.org/abs/
1902.10921
[21] Liu Y F, Hong M, Dai Y H. Max-min fairness linear transceiver design problem for a multi-
user SIMO interference channel is polynomial time solvable [J]. IEEE Signal Processing Letters,
2013, 20(1): 27-30.
[22] Liu Y F. Dynamic spectrum management: A complete complexity characterization [J]. IEEE
Transactions on Information Theory, 2017, 63(1): 392-403.
[23] Powell M J. On search directions for minimization algorithms [J]. Mathematical Programming,
1973, 4(1): 193-201.
[24] Liu Y F, Dai Y H, Luo Z Q. Max-min fairness linear transceiver design for a multi-user MIMO
interference channel [C]//Proceedings of IEEE International Conference on Communications
(ICC), 2011, 1-5.
[25] Liu Y F, Dai Y H, Luo Z Q. Max-min fairness linear transceiver design for a multi-user MIMO
interference channel [J]. IEEE Transactions on Signal Processing, 2013, 61(9): 2413-2423.
[26] Yu W, Kwon T, Shin C. Adaptive resource allocation in cooperative cellular networks [M]//
Cooperative Cellular Wireless Networks. New York: Cambridge University Press, 2010, 233-
256.
[27] Zhang J, Chen R, Andrews J G, et al. Networked MIMO with clustered linear precoding [J].
IEEE Transactions on Wireless Communications, 2009, 8(4): 1910-1921.
[28] Chen E, Tao M, Liu Y F. Joint base station clustering and beamforming for non-orthogonal
multicast and unicast transmission with backhaul constraints [J]. IEEE Transactions on Wire-
less Communications, 2018, 17(9): 6265-6279.
60 4æ¹ 23ò
[29] Ng C T, Huang H. Linear precoding in cooperative MIMO cellular networks with limited
coordination clusters [J]. IEEE Journal on Selected Areas in Communications, 2010, 28(9):
1446-1454.
[30] Hong M, Sun R, Baligh H, et al. Joint base station clustering and beamformer design for partial
coordinated transmission in heterogeneous networks [J]. IEEE Journal on Selected Areas in
Communications, 2013, 31(2): 226-240.
[31] Dai B, Yu W. Sparse beamforming and user-centric clustering for downlink cloud radio access
network [J]. IEEE Access, 2014, 2: 1326-1339.
[32] Shi Y, Zhang J, Letaief K. Group sparse beamforming for green cloud-RAN [J]. IEEE Trans-
actions on Wireless Communications, 2014, 13(5): 2809-2823.
[33] Tao M, Chen E, Zhou H, et al. Content-centric sparse multicast beamforming for cache-enabled
cloud RAN [J]. IEEE Transactions on Wireless Communications, 2016, 15(9): 6118-6131.
[34] Liao W C, Hong M, Liu Y F, et al. Base station activation and linear transceiver design
for optimal resource management in heterogeneous networks [J]. IEEE Transactions on Signal
Processing, 2014, 62(15): 3939-3952.
[35] Shen K, Yu W. Distributed pricing-based user association for downlink heterogeneous cellular
networks [J]. IEEE Journal on Selected Areas in Communications, 2014, 32(6): 1100-1113.
[36] Shen K, Liu Y F, Ding D Y, et al. Flexible multiple base station association and activation for
downlink heterogeneous networks [J]. IEEE Signal Processing Letters, 2017, 24(10): 1498-1502.
[37] Sanjabi M, Razaviyayn M, Luo Z Q. Optimal joint base station assignment and beamforming for
heterogeneous networks [J]. IEEE Transactions on Signal Processing, 2014, 62(8): 1950-1961.
[38] Shi Q, Razaviyayn M, Luo Z Q, et al. An iteratively weighted MMSE approach to distributed
sum-utility maximization for a MIMO interfering broadcast channel [J]. IEEE Transactions on
Signal Processing, 2011, 59(9): 4331-4340.
[39] Zhang X, Chang T H, Liu Y F, et al. Max-min fairness user scheduling and power allocation
in full-duplex OFDMA systems [J]. IEEE Transactions on Wireless Communications, 2019,
18(6): 3078õ3092.
[40] Codreanu M, Tolli A, Juntti M, et al. Joint design of Tx-Rx beamformers in MIMO downlink
channel [J]. IEEE Transactions on Signal Processing, 2007, 55(9): 4639-4655.
[41] Chang T H, Liu Y F, Lin S C. QoS-based linear transceiver optimization for full-duplex multi-
user communications [J]. IEEE Transactions on Signal Processing, 2018, 66(9): 2300-2313.
[42] Li Q, Hong M, Wai H T, et al. Transmit solutions for MIMO wiretap channels using alternating
optimization [J]. IEEE Journal on Selected Areas in Communications, 2013, 31(9): 1714-1727.
[43] Hong M, Li Q, Liu Y F. Decomposition by successive convex approximation: A unifying ap-
proach for linear transceiver design in heterogeneous networks [J]. IEEE Transactions on Wire-
less Communications, 2016, 15(2): 1377-1392.
[44] Wright S J. Coordinate descent algorithms [J]. Mathematical Programming, 2015, 151(1): 3-34.
[45] Hong M, Razaviyayn M, Luo Z Q, et al. A unified algorithmic framework for block-structured
optimization involving big data: With applications in machine learning and signal processing
[J]. IEEE Signal Processing Magazine, 2016, 33(1): 57-77.
[46] Verdu S. Multiuser Detection [M]. New York: Cambridge University Press, 1998.
[47] Yang S, Hanzo L. Fifty years of MIMO detection: The road to large-scale MIMOs [J]. IEEE
Communication Surveys & Tutorials, 2015, 17(4): 1941-1988.
[48] Verdu S. Computational complexity of optimum multiuser detection [J]. Algorithmica, 1989,
4(1-4): 303-312.
[49] Kisialiou M, Luo X, Luo Z Q. Efficient implementation of quasi-maximum-likelihood detection
based on semidefinite relaxation [J]. IEEE Transactions on Signal Processing, 2009, 57(12):
4811-4822.
3Ï Ã�Ï&XÚ�O¥�ü�`z¯KÚ�'`z�{ 61
[50] Fincke U, Pohst M. Improved methods for calculating vectors of short length in a lattice,
including a complexity analysis [J]. Mathematics of Computation, 1985, 44(170): 463-471.
[51] Viterbo E, Boutros J. A universal lattice code decoder for fading channels [J]. IEEE Transac-
tions on Information Theory, 1999, 45(5): 1639-1642.
[52] Damen M O, Gamal H E, Caire G. On maximum-likelihood detection and the search for the
closest lattice point [J]. IEEE Transactions on Information Theory, 2003, 49(10): 2389-2402.
[53] Jalden J, Ottersten B. On the complexity of sphere decoding in digital communications [J].
IEEE Transactions on Signal Processing, 2005, 53(4): 1474-1484.
[54] Lu C, Liu Y F, Zhang W Q, et al. Tightness of a new and enhanced semidefinite relaxation
for MIMO detection [J]. SIAM Journal on Optimization, 2019, 29(1): 719-742.
[55] Lu C, Liu Y F. An efficient global algorithm for single-group multicast beamforming [J]. IEEE
Transactions on Signal Processing, 2017, 65(14): 3761-3774.
[56] Lu C, Deng Z, Zhang W Q, et al. Argument division based branch-and-bound algorithm for
unit-modulus constrained complex quadratic programming [J]. Journal of Global Optimization,
2018, 70(1): 171-187.
[57] Jalden J. Detection for multiple input multiple output channels: Analysis of sphere decoding
and semidefinite relaxation. Ph.D. thesis. Stockholm: KTH, 2006.
[58] Jalden J, Martin C, Ottersten B. Semidefinite programming for detection in linear systems–
Optimality conditions and space-time decoding [C]//Proceedings of IEEE International Con-
ference on Acoustics, Speech, and Signal Processing (ICASSP), 2011, 9-12.
[59] Kisialiou M, Luo Z Q. Probabilistic analysis of semidefinite relaxation for binary quadratic
minimization [J]. SIAM Journal on Optimization, 2010, 20(4): 1906-1922.
[60] So A M C. Probabilistic analysis of the semidefinite relaxation detector in digital communica-
tions [C]//Proceedings of twenty-first annual ACM-SIAM symposium on Discrete Algorithms
(SODA), 2010, 698-711.
[61] Mobasher A, Taherzadeh M, Sotirov R, et al. A near-maximum-likelihood decoding algorithm
for MIMO systems based on semi-definite programming [J]. IEEE Transactions on Information
Theory, 2007, 53(11): 3869-3886.
[62] Liu Y F, Xu Z, Lu C. On the equivalence of semidifinite relaxations for MIMO detection with
general constellations [C]//Proceedings of ICASSP, 2019, 4549-4553.
[63] Lu C, Liu Y F, Zhou J. An enhanced SDR based global algorithm for nonconvex complex
quadratic programs with signal processing applications [J]. arXiv:1902.04287.
[64] Liu H, Yue M C, So A M C, et al. A discrete first-order method for large-scale MIMO de-
tection with provable guarantees [C]//Proceedings of IEEE International Workshop on Signal
Processing Advances in Wireless Communications (SPAWC), 2017, 669-673.
[65] Burer S, Monteiro R D, Zhang Y. Rank-two relaxation heuristics for max-cut and other binary
quadratic programs [J]. SIAM Journal on Optimization, 2002, 12(2): 503-521.
[66] Luo Z Q, Luo X, Kisialiou M. An efficient quasi-maximum likelihood decoder for PSK signals
[C]//Proceedings of ICASSP, 2003, 561-564.
[67] Pan J, Ma W K, Jalden J. MIMO detection by Lagrangian dual maximum-likelihood relaxation:
Reinterpreting regularized lattice decoding [J]. IEEE Transactions on Signal Processing, 2014,
62(2): 511-524.
[68] Lu C, Liu Y F, Zhou J. An efficient global algorithm for nonconvex complex quadratic prob-
lems with applications in wireless communications [C]//Proceedings of IEEE/CIC International
Conference on Communications in China (ICCC), 2017, 1-5.
[69] Vandenberghe L, Boyd S. Semidefinite programming [J]. SIAM Review, 1996, 38(1): 49-95.
[70] Pu W, Liu Y F, Yan J, et al. Optimal estimation of sensor biases for asynchronous multi-sensor
data fusion [J]. Mathematical Programming, 2018, 170(1): 357-386.
62 4æ¹ 23ò
[71] Andrews J G, Buzzi S, Choi W, et al. What will 5G be? [J]. IEEE Journal on Selected Areas
in Communications, 2014, 32(6): 1065-1082.
[72] c�m,��©,pÜÛ,�. 5G£ÄÏ&uЪ³�eZ'�Eâ [J].¥I�Æ: &E�Æ, 2014,
44: 551–563.
[73] Tao M, Chen E, Yu W, et al. Cache-enabled cloud radio access networks [M]// Wireless Edge
Caching: Modeling, Analysis and Optimization, Cambridge University Press, 2019.
[74] Dai B, Liu Y F, Yu W. Optimized base-station cache allocation for cloud radio access network
with multicast backhaul [J]. IEEE Journal on Selected Areas in Communications, 2018, 36(8):
1737-1750.
[75] Jacobsson S, Durisi G, Coldrey M, et al. Quantized precoding for massive MU-MIMO [J].
IEEE Transactions on Communications, 2017, 65(11): 4670-4684.
[76] Sohrabi F, Liu Y F, Yu W. One-bit precoding and constellation range design for massive MIMO
with QAM signaling [J]. IEEE Journal of Selected Topics in Signal Processing, 2018, 12(3):
557-570.
[77] Zhang N, Liu Y F, Farmanbar H, et al. Network slicing for service-oriented networks under
resource constraints [J]. IEEE Journal on Selected Areas in Communications, 2017, 35(11):
2512-2521.
[78] Domenico A D, Liu Y F, Yu W. Optimal computational resource allocation and network slic-
ing deployment in 5G hybrid C-RAN [C]// Proceedings of IEEE International Conference on
Communications (ICC), 2019.
[79] Liu L, Larsson E G, Yu W, et al. Sparse signal processing for grant-free massive connectivity:
A future paradigm for random access protocols in the Internet of Things [J]. IEEE Signal
Processing Magazine, 2018, 35(5): 88-99.
[80] Chen Z, Sohrabi F, Liu Y F, et al. Covariance based joint activity and data detection for massive
random access with massive MIMO [C]// Proceedings of IEEE International Conference on
Communications (ICC), 2019.