Orbital Angular Momentum for Wireless Communications

Orbital Angular Momentum for W ireless Communications W enchi Cheng, Member , IEEE , W ei Zhang, F ellow , IEEE , Haiyue Jing, Student Member , IEEE , Shanghua Gao, and Hailin Zhang, Member , IEEE Abstract As the traditional resources (frequency , time, space, etc.) are efficiently utilized, it becomes more and more challenging to satisfy the e ver -lasting capacity-growing and users-boosting demand in wireless networks. Recently , the electromagnetic (EM) wave was found to possess not only linear momentum, but also angular momentum. The orbital angular momentum (OAM) is a kind of wav efront with helical phase. The O AM-based vortex wav e has different topological charges, which are orthogonal to each other , bridging a ne w way for multiple access in wireless communications. In this article, we introduce the fundamental theory of O AM and the O AM based wireless communications. The research challenges regarding OAM signal generation, OAM beam con ver ging, and OAM signal reception are discussed. Further , we propose a new multiuser access with different O AM-modes in wireless networks, where multiple O AM-modes are used as a new orthogonal dimension for interference av oidance. Simulation results rev eal the inherent property of OAM wav es and show that OAM based radio transmission can significantly increase the spectrum ef ficiency in wireless networks. Index T erms Orbital angular momentum (O AM), radio vortex wireless communications, multiuser access, inter - ference avoidance. W enchi Cheng, Haiyue Jing, Shanghua Gao, and Hailin Zhang are with the State Ke y Laboratory of Integrated Services Networks, Xidian Univ ersity , Xi’an, 710071, China (e-mails: wccheng@xidian.edu.cn; hyjing@stu.xidian.edu.cn; shgao@stu.xidian.edu.cn; hlzhang@xidian.edu.cn). W ei Zhang is with the School of Electrical Engineering and T elecommunications, Uni versity of New South W ales Sydney , NSW , Australia (w .zhang@unsw .edu.au). 1 Orbital Angular Momentum for W ireless Communications I . I N T R O D U C T I O N As wireless communications migrate from the fourth-generation (4G) to the fifth-generation (5G) and beyond, it is highly demanded to meet the requirements of explosi ve data traffic. For example, the aggregate data rate is expected to be increased by roughly 1000 times for 5G as compared with 4G [1]. It is expected using 5G and 5G-beyond New Radio (NR) for spectrum ef ficiency enhancement with adv anced techniques, such as massiv e multiple-input- multiple-output (MIMO), co-frequency co-time full-duplex, and millimeter -wa ve (mmW av e) [1], [2]. F or 5G NR, the promise of significant spectrum ef ficiency enhancement, v ast spatial div ersity , and simple transmit/receiv e structure has ele v ated massiv e MIMO to a central position in 5G wireless communications netw orks, with a foreseen role of coexisting with mmW ave [3]. Co- frequency co-time full-duplex, which potentially double the spectrum efficienc y , is e xpected to be integrated into future 5G-beyond wireless communications networks [4]. Howe ver , during the past fe w decades, multiple orthogonal resources, such as frequency , time, and space, were extensi vely explored. Nowadays, it becomes more and more dif ficult to increase capacity or support more users with the traditional access techniques such as time-division-multiple-access and frequency-division-multiple-access. In fact, until now wireless communication is being built on the plane-electromagnetic (PE) wa ve. Howe ver , the electromagnetic (EM) wa ve possesses not only linear momentum, but also angular momentum, which contains the spin angular momentum (SAM) and orbital angular momentum (O AM). O AM, as a kind of wav efront with helical phase, has attracted much research attention [5]. O AM has a great number of topological char ges, i.e., O AM-modes. Beams with dif ferent OAM-modes are orthogonal to each other and they can be multiplexed/demultiple xed together , thus increasing the capacity without relying on the traditional resources such as time and frequency . OAM, which has multiple orthogonal topological charges, bridges a new way to significantly increase spectrum ef ficiency and is expected to be used in 5G-beyond or e ven more future wireless communications networks. Recently , some experiments hav e verified the feasibility of O AM based wireless communica- tions [5]–[8]. The authors of [6] studied tw o O AM-modes (O AM-modes 0 and 1 ), which share the same frequency band. The authors of [7] performed the O AM based high capacity transmission with 60 GHz and 17 GHz carrier frequencies, respectiv ely . The authors of [9] demonstrated April 23, 2018 DRAFT 2 that O AM multiplexing can achiev e high capacity in mmW ave communications. Moreover , the research on O AM based wireless communications extends to mode detection, mode separation, axis estimation and alignment, mode modulation, O AM-beams con verging, and mode hopping, etc. Generally , MIMO multiplexing can be jointly used with OAM, thus significantly increasing spectrum efficienc y . The authors of [8] and [10] experimentally and theoretically demonstrated that jointly using MIMO based spatial multiplexing and OAM multiplexing can increase the spectrum efficiency of wireless communications. Although the feasibility of O AM based wireless communications is v alidated, there are many research problems unsettled. For example, in order to support simultaneous transmission with multiple OAM-modes, transmit and receive antennas need to support the generation and recep- tion, respectiv ely , of multiple O AM-modes mixed signals. For another example, because the EM wa ve with O AM is v orticose hollow and di ver gent [5], the O AM beam needs to be con verged for relati vely long distance transmission. Moreover , the phase errors due to the non-alignment or fading are very hard to be estimated at the receiv er . Although O AM beam is v orticose hollow and div er gent, the di vergence of O AM beams greatly reduces as the frequency increases. W ith small div ergence, the received signal-to-noise ratio (SNR) is relativ ely large, which is beneficial for the reception of OAM signal. Thus, it is expected to use O AM in mmW a ve networks with high frequency such as 70GHz. The authors of [10] proposed the framew ork for O AM embedded massi ve MIMO communication to obtain the multiplicativ e spectrum ef ficiency gain for joint O AM and massiv e MIMO mmW av e wireless communications, which is larger than that for the traditional massive MIMO mmW av e communications. In this paper , we surve y the fundamental issues of using O AM in wireless communications. W e show the advantages and challenges of O AM based wireless communications. Furthermore, we gi ve a no vel O AM-modes based orthogonal multiuser access framework and e valuate the obtained spectrum efficienc y in the case study . I I . W H A T I S OA M ? O AM is one basic physical property of EM wa ve. It describes the orbital property for EM rotational degree of freedom and rotation characteristic for energy . OAM is interpreted as a beam with a number of O AM-modes which can theoretically take not only any integer value but also any non-integer v alue. Inherently , the EM wav e carried OAM can be generated by PE wav e with one phase rotation f actor exp( ilϕ ) , where i = √ − 1 , l is the order/inde x of O AM-mode, and DRAFT April 23, 2018 3 ϕ is the azimuthal angle (defined as the angular position on a plane perpendicular to the axis of propagation). A pure OAM-mode is characterized by integer and different O AM-modes are orthogonal with each other . When the OAM-mode is a non-integer , the phase term exp( il ϕ ) can be expressed by the sum of Fourier series of orthogonal OAM-modes. Af fected by the rotation phase factor , the wa vefront phase is spiral structure instead of planar structure. The w a vefront phase rotates around the beam propagation direction and the phase changes 2 π l after a full turn. Figure 1 shows the wa vefront and 3 dimensional (3D) profile for OAM wa ves with different modes, where the transmit antenna is uniform circular array (UCA) antenna with 16 array- elements. Figs. 1(a)-1(d) show the wav efront phase corresponding to O AM-modes 0, 1, 2, and 3, respectiv ely . In fact, O AM-mode 0 represents the PE wa ve as sho wn in Fig. 1(a). Based on Figs. 1(a)-1(d), we can observe that the spiral characteristic of OAM wa ve becomes complicated and the phase changes sharply as the the order/index of O AM-modes increases within the same distance. Figs. 1(e)-1(h) sho w the 3D profiles of O AM wav es for dif ferent O AM-modes 0, 1, 2, and 3, respectiv ely . There exist central hollo w for different OAM-modes except OAM-mode 0. This is because the O AM wa ve of mode 0 is in fact the PE wa ve. The central hollo w increases as the order of O AM-mode increases. Also, the po wer gain decreases as the order of OAM- mode increases. This indicates that it is impossible for long distance O AM wa ve transmission by directly using O AM-modes. For long distance transmission, we need to con ver ge the hollow O AM wav e. I I I . T H E OA M B A S E D W I R E L E S S C O M M U N I C AT I O N S Dif ferent from frequency/time/code-domain based orthogonal di vision, OAM offers a ne w mode domain to support the orthogonal access of multiple users. W ith OAM, we can re-design the wireless communications because many aspects in wireless communications can be improved with the new orthogonal dimension. Three basic advantages reg arding O AM based wireless communications are revie wed as follows: Advantage 1: High spectrum efficiency — Different O AM-modes are orthogonal with each other . Thus, in ideal case there is no interference among different OAM-modes. With the orthogonality , the parallel transmission can be performed among multiple OAM-modes. The orthogonality among different O AM-modes can be used to increase the spectrum ef ficiency in wireless communications without consuming more traditional frequency/time/code/po wer- April 23, 2018 DRAFT 4 (a) W avefront of mode 0. (b) W avefront of mode 1. (c) W avefront of mode 2. (d) W avefront of mode 3 (e) 3D profile of mode 0. (f) 3D profile of mode 1. (g) 3D profile of mode 2. (h) 3D profile of mode 3 Fig. 1. W av efront and 3D profile for OAM wav es with different modes. domain resources. Also, mode-domain resources can be jointly used with frequency/time/code- domain resources to significantly increase the spectrum efficienc y in wireless communications. Advantage 2: More users access — O AM pro vides a nov el multiple access method, i.e., mode di vision multiple access (MDMA), without consuming more frequency and time resources. W ith MDMA, dif ferent users can employ dif ferent O AM-modes to orthogonally access the wireless networks. Instead of non-orthogonal multiple access, which uses power domain to distinguish multiple users, it is expected to get back to orthogonal multiple access using mode-domain resource in future wireless communications. DRAFT April 23, 2018 5 (a) Spiral Phase Plate (SPP) antenna. (b) Uniform Circular Array (UCA) antenna. (c) Metasurface. Fig. 2. Three kinds of antenna structures for radio vorte x signal generation. Advantage 3: High reliability f or anti-jamming — Faced with the more and more crowded spectrum pressure, there exist limitations using the con ventional frequency hopping techniques for anti-jamming. Ho we ver , OAM-mode hopping technique has the potential for anti-jamming in future wireless communication. O AM can not only be used within the narrow band, but also jointly used with frequency-hopping in wide band to improve the ability of anti-jamming for wireless communications. Although OAM has the potential to increase the spectrum efficienc y , support more users, and impro ve the reliability of anti-jamming, there are still some important research challenges remaining unsettled. The issues can be classified into three categories: radio vortex signal generation, transmission, and reception. Radio V ortex Signal Generation. Different from traditional PE w a ve based signals, radio vorte x signals hav e the phase rotation factor exp( il ϕ ) . There are some popular facilities can be used to generate radio v ortex signal such as Spiral Phase Plate (SPP) antenna, Uniform Circular Array (UCA) antenna, and metasurfaces, as shown in Fig. 2. • SPP antenna [8]: An example of SPP antenna is giv en in Fig. 2(a). The SPP antenna generates the phase delay by increasing the antenna thickness in proportion to the azimuthal angle or by drilling inhomogeneous holes in dielectric plate to change the equi v alent permitti vity . The SPP antenna has the advantages of small diver gence and lo w attenuation as well as the disadvantages of not applicable for relativ ely low frequency transmission and cannot generate multiple OAM-modes simultaneously . • UCA antenna [11]: An example of UCA antenna is giv en in Fig. 2(b). The phase information of adjacent array-element of UCA antenna is linearly increased by 2 π l / N , where N is the April 23, 2018 DRAFT 6 number of array-elements. The UCA antennas are low profile, lo w weight, and easy to manufacture with rectangular patch arrays. Also, the UCA antennas can simultaneously generate multiple vorte x beams with multiple O AM-modes ev en in the radio frequency band. Howe ver , the vortex beams generated by UCA is div ergent and centrally hollo w . Thus, the UCA antennas need to be jointly used with the con verging schemes to combat the signal attenuation during the propagation. • Metasurf aces [12]: An example of metasurfaces is gi ven in Fig. 2(c). In the metasurfaces based OAM signal generation schemes, the wa vefront of electromagnetic wa ves are con- trolled by regulating phase shift to the incoming wa ves. These schemes hav e the advantages of lo w profile, small mass, and lo w manufacturing cost. Ho we ver , it is hard to accurately control the phase for signal modulation and thus not applicable to multiple O AM-modes transmission in wireless communications. Radio V ortex Signal T ransmission. Three typical problems need to be considered for radio vorte x signal transmission. • T ransmitter -r eceiver alignment. It is required that the transmitter and receiv er are aligned with each other to decompose signals with different O AM-modes [13]. If the transmitter and the recei ver are not aligned, the phase of receiv ed signal contains not only the phase of O AM mode, but also the phase turb ulence due to unequal distance transmission at different places of the recei ver . Therefore, for non-aligned scenarios, it is demanded to add phase turbulence adapti ve estimation algorithm at the receiv er . • F ading . Current OAM related researches mainly focus on the line-of-sight scenario, where no fading has been taken into account. Howe ver , there exists fading in many practical scenarios, leading to the randomness of wa vefront phases at the recei ver . If the signals of dif ferent OAM-modes undergo fading, the phase change corresponding to each O AM-mode needs to be estimated. • Con ver gence. Because an O AM beam becomes more and more di ver gent as the order of O AM-mode increases, it se verely reduces the transmission distance and decreases the spectrum ef ficiency of O AM based wireless communications. It is required to make OAM beams con vergent so that all O AM-modes including both high and lo w order O AM-modes can be efficiently used. Ho wev er , it is a dif ficult task to design efficient antenna structure and algorithms to con ver ge the O AM beams without changing the original wa vefront phases of DRAFT April 23, 2018 7 O AM-modes. There are two typical methods for conv erging OAM beams: parabolic antenna and lens antenna. The parabolic antenna reforms the di ver ged OAM beam to an con ver ged beam while keeping remain the angular identification information of each O AM-mode. The parabolic antenna enjoys relati vely low attenuation for O AM beams. Howe ver , the size of the parabolic antenna is relativ ely large. Through refracting, the lens antenna can reform the di ver ged O AM beams into the con ver ged O AM beams. The lens antenna fits all frequency band. Howe ver , it is bulk y and needs to endure relativ ely large attenuation. Figure 3 sho ws the E field of uncon verged and con verged O AM beams observed from the horizontal direction and the phase profiles of the O AM-modes. N patch-elements of UCA are equally distributed on the circle, where the radius is set as 25 mm. The relativ e permitti vity of patch material is set as 2.2. The length and the width of patch element are set as 2.947 mm and 3.388 mm, respecti vely . Then, we place the plane at 100 mm to observe the results. Figs. 3(a) and 3(b) show the E field of uncon ver ged and con ver ged O AM beams, respectiv ely , observed from the horizontal direction. Fig. 3(a) confirms that the uncon ver ged O AM beams ha ve a large hollow in the central area. Moreov er , the hollo w increases as the order of OAM-mode ( l ) increases. Fig. 3(b) verifies that the O AM beams can be effecti vely con v erged with the lens antenna. The central hollo w is very small as compared with Fig. 3(a) and the intensity of OAM beams ef ficiently increases. Fig. 3(c) sho ws the phase profiles of the O AM-modes 0, 1, 2, and 3. W e set the frequency of OAM beams as 35 GHz. Radio V ortex Signal Reception. At the receiv er , the phase detection is a key to distinguish the order of different O AM-modes [14]. The radio vorte x signal reception schemes for SPP antenna, UCA antenna, as well as other schemes are discussed in the following. • Reception with SPP antenna. One SPP antenna can detect the signal with one O AM-mode. The receiv er generates the conjugate phase related to the corresponding OAM-mode carried by the signal. Then, multiplied with the recei ved one OAM-mode signal, the OAM wav e can be transformed into plane wav e. An SPP antenna can only decompose one O AM-mode among different O AM-modes due to the characteristic of SPP antenna [8]. • Reception with UCA antenna. The spatial fast F ourier transform (FFT) can be used to obtain the signal carried by the corresponding O AM-mode [10]. This is because it has a property that after spatially sampling the sum is zero within the interv al length except the designed April 23, 2018 DRAFT 8 l = 0 (PE beam) Large hollow l = 1 Large hollow l = 2 Large hollow l = 3 (a) The E field of uncon ver ged O AM beams. Small hollow l = 0 (PE beam) Small hollow l = 1 Small hollow Small hollow l = 2 Small hollow Small hollow l = 3 (b) The E field of lens conv erged O AM beams. l = 0 (PE beam) l = 1 l = 2 l = 3 (c) The phase profiles of O AM-mode 0, 1, 2, and 3. Fig. 3. The E field of uncon verged and conv erged OAM beams observed from the horizontal direction and the phase profiles of the O AM-modes. DRAFT April 23, 2018 9 O AM-mode. The UCA antenna based reception scheme supports the detection for multiple O AM-modes. Howe ver , it is highly required that the transmit UCA and the recei ve UCA are aligned with each other . • Other schemes. There exist other schemes to obtain the signals carried by O AM-modes. The phase gradient method (PGM) is dev eloped to identify the O AM-modes. The PGM is dependent on the circular separation among recei ver antennas. Because the helical phase fronts of different O AM beams are different, the OAM-mode can be detected using a two- point phase measurement. Other reception schemes include direct torque measurement and triangulation. I V . O A M - M O D E S B A S E D M U LT I P L E A C C E S S : A C A S E S T U DY As a case study , we in vestig ate the O AM-modes based multiple access in wireless networks, as shown in Fig. 4, consisting one macrocell and se veral small cells. L 1 - L 2 denote the PE beam used in macrocell and L 3 - L 6 represent the OAM beams corresponding to a group of O AM-modes used in small cells. Theoretically , the cross-layer and co-layer interference can be solved with sufficiently large number of OAM-modes in wireless networks. The macrocells and small cells can utilize the PE wa ve and O AM wav es, respectively , without causing cross-layer interference between macrocells and small cells. Also, different small cells can use different O AM-modes so that in principle there is no interference among small cells if the number of av ailable O AM-modes is larger than the number of small cells. If not, we can dev elop the O AM-modes allocation/scheduling schemes for small cells to minimize the interference. For instance, the small cells far away use the same O AM-modes while the neighbouring small cells utilize different O AM-modes. As a result, the spectrum ef ficiency of wireless networks can be significantly increased to meet the demand of tremendous data traffic. W e consider the 3GPP Rel-12 small cells scenarios for performance ev aluation. Figure 5 sho ws the comparison in spectrum efficiencies of traditional frequency-di vision-multiple-access (FDMA) and the newly proposed MDMA with different number of O AM-modes, where we set the density of users as 0.2 users/m 2 and the number of frequency-orthogonal channels is two. From Fig. 5, we can observe that the spectrum efficienc y of more than one O AM-modes is larger than that of traditional FDMA and the spectrum efficienc y significantly increases as the number of O AM-modes increases. This is because the orthogonality of O AM-modes makes the interference partly av oided. When the SNR is relativ ely large, there exists the ceiling effect for April 23, 2018 DRAFT 10 1 2 3 4 5 6 L 1 L 2 L 3 L 4 L 5 L 6 L 1 L 2 L 3 L 4 L 5 L 6 O F D M A s u b c h a n n e l s 1 2 3 4 5 6 O F D M A s u b c h a n n e l s P l a ne w a ve i n t he upl i nk V or t e x w a ve i n t he upl i nk P l a ne w a ve i n t he dow nl i nk V or t e x w a ve i n t he dow nl i nk P l a ne w a ve i n t he u pl i nk P l a ne w a ve i n t he dow nl i nk V or t e x w a ve i n t he upl i nk V or t e x w a ve i n t he dow nl i nk s ubc ha nne l s unus e d s ubc ha nne l s us e d ... Fig. 4. The O AM beams based multiple users access in wireless networks. 0 2 4 6 8 10 12 14 16 18 20 Rec eiv ed SNR (dB ) 4 8 12 16 20 24 28 32 36 40 Spectru m e / c ienc y (bps/ Hz /m 2 ) Traditional FDMA Two OAM-modes Three OAM-modes Four OAM-modes Fig. 5. The spectrum efficiencies versus received SNR with different number of OAM-modes for multiple users access. DRAFT April 23, 2018 11 0.2 0.4 0.6 0.8 1.0 1.2 The de nsit y of users (No ./m 2 ) 0 4 8 12 16 20 24 28 Spectru m e / c ienc y (bps/ Hz /m 2 ) Traditional FDMA Two OAM-modes Three OAM-modes Four OAM-modes Fig. 6. The spectrum efficiencies versus the density of users with different number of OAM-modes for multiple users access. the spectrum efficienc y . This is because the interference caused by other users increases as the SNR increases. Figure 6 depicts the spectrum efficiencies of traditional FDMA and dif ferent number of OAM- modes, respecti vely , versus the density of users in wireless networks. The spectrum efficienc y increases as the density of users increases in wireless networks. W e can also observe that the spectrum ef ficiency of traditional FDMA is smaller than that of more than one OAM- modes. Ho we ver , for one O AM-mode, the spectrum efficienc y with 1.0 users/m 2 is similar to the spectrum efficienc y with 1.2 users/m 2 . When the density of users is relativ ely large, the spectrum ef ficiency of four OAM-modes is much higher than that of one O AM-mode. This is because as the number of O AM-modes increases, more users can be interference-free. Therefore, it is efficient for interference av oidance when applying wireless communications into wireless networks, thus greatly increasing the spectrum efficienc y of wireless networks. April 23, 2018 DRAFT 12 Ho we ver , there are some challenges regarding applying OAM beams into wireless networks. Challenge 1: Limited Number of A vailable O AM-Modes. — The OAM beams of high order O AM-modes experience sev ere attenuations in propagation. Thus, the number of av ailable O AM-modes is relativ ely small. Considering the high density of small cells, it is very likely the number of a v ailable O AM-modes is not enough. T o significantly increase the number of av ailable O AM-modes, the OAM beams corresponding to high order OAM-modes need to be con ver ged, which remains as a challenging problem. Now adays, some schemes have been dev eloped to con ver ge O AM beams. For example, the parabolic antenna can be used to reduce the div ergence of O AM beams without impacting the orthogonality of OAM-modes [10]. In addition, bifocal lens antenna also has good performance for con verging O AM beams. Both of these two methods need an antenna to generate O AM beams and another antenna for con verging. Alternativ ely , by transforming the singular UCA into the concentric UCAs, more low-order O AM-modes can be used as well as taking into account the inter-mode interference [13]. Challenge 2: J oint O AM-Mode and Fr equency/T ime Partition. — When the number of O AM-modes is not enough for wireless networks ev en after O AM-beams con ver ging, jointly using O AM-modes and frequency/time partition are highly needed to balance the traf fic over mode, frequency , and time domains to increase the spectrum efficienc y [15]. Moreo ver , when O AM-modes and other resources (frequency or time) are combined to maximize the spec- trum efficienc y , the priority of choosing O AM-modes or other resources should be considered. Mode-di vision-multiple-access based multi-channel medium access control (MA C) protocols can be jointly integrated with frequenc y-di vision-multiple-access/time-di vision-multiple-access/code- di vision-multiple-access for high throughput and flexible multiple access in O AM wireless communications networks. Challenge 3: Channel Estimation for Different O AM-Modes. — The number of channels corresponding to all OAM-modes is very lar ge, thus causing heav e ov erhead for channel es- timation. Moreover , it is possible that the phase front of O AM-modes will change during the propagation of OAM beams, thus causing the loss of phase identification feature at the receiv er . As a result, it is dif ficult to recover the transmit signals corresponding to different O AM-modes when the number of channels is relati vely large. Until no w , this area lacks v ery intriguing or solid results. For UCA antennas, the channel matrix for array-elements between the transmitter and the receiv er is the circular matrix, which makes the channel estimation easy to handle and is likely to be the breakthrough for channel estimation in O AM wireless communications. DRAFT April 23, 2018 13 V . C O N C L U S I O N S In this article, we ha ve introduced the fundamental theory of O AM and the O AM based wireless communications. The benefits and challenges in O AM based wireless communications are discussed. Further , we proposed a new multiuser access based on OAM beams and then we studied the case in two-tier wireless networks. The performance results verified the con vergence for O AM beams and showed the spectrum efficienc y enhancement for wireless networks. In conclusion, O AM pro vides the ne w mode domain, which gi ves many opportunities for future wireless communications research. R E F E R E N C E S [1] J. G. Andrews, S. Buzzi, W . Choi, S. V . Hanly , A. Lozano, A. C. K. Soong, and J. C. Zhang, “What will 5G be? ” IEEE Journal on Selected Ar eas in Communications , vol. 32, no. 6, pp. 1065–1082, June 2014. [2] ITU Document 5D/TEMP/625-E. “IMT vision-framework and ov erall objectives of the future de velopment of IMT for 2020 and beyond, ” 2015. [3] F . Boccardi, R. W . Heath, A. Lozano, T . L. Marzetta, and P . Popovski, “Five disruptiv e technology directions for 5G, ” IEEE Communications Magazine , vol. 52, no. 2, pp. 74–80, Feb . 2014. [4] S. Hong, J. Brand, J. I. Choi, M. Jain, J. Mehlman, S. Katti, and P . Le vis, “ Applications of self-interference cancellation in 5G and beyond, ” IEEE Communications Magazine , vol. 52, no. 2, pp. 114–121, Feb. 2014. [5] B. Thid ´ e, H. Then, J. Sj ¨ oholm, K. Palmer , J. Bergman, T . D. Carozzi, N. Istomin, N. H. Ibragimov , and R. Khamitov a, “Utilization of photon orbital angular momentum in the low-frequency radio domain, ” Phys. Rev . Lett. , vol. 99, no. 8, pp. 087701–1–087701–4, Aug. 2007. [6] M. T amagnone, C. Craeye, and J. Perruisseau-Carrier , “Comment on ‘Encoding many channels on the same frequency through radio vorticity: first experimental test’, ” New Journal of Physics , vol. 14, no. 11, 2012. [7] F . E. Mahmouli and S. D. W alker , “4-Gbps uncompressed video transmission over a 60-GHz orbital angular momentum wireless channel, ” IEEE W ir eless Communications Letters , vol. 2, no. 2, pp. 223–226, Apr. 2013. [8] Y . Ren, L. Li, G. Xie, Y . Y an, Y . Cao, H. Huang, N. Ahmed, Z. Zhao, P . Liao, C. Zhang, G. Caire, A. F . Molisch, M. T ur, and A. E. W illner, “Line-of-sight millimeter-wav e communications using orbital angular momentum multiplexing combined with con ventional spatial multiplexing, ” IEEE T ransactions on W ireless Communications , vol. 16, no. 5, pp. 3151–3161, May 2017. [9] Y . Y an, L. Li, Z. Zhao, G. Xie, Z. W ang, Y . Ren, N. Ahmed, S. Sajuyigbe, S. T alwar , M. T ur, N. Ashrafi, S. Ashrafi, A. F . Molisch, and A. E. W illner, “32-Gbit/s 60-Ghz millimeter-wa ve wireless communication using orbital angular momentum and polarization multiplexing, ” in 2016 IEEE International Conference on Communications (ICC) , May 2016, pp. 1–6. [10] W . Cheng, H. Zhang, L. Liang, H. Jing, and Z. Li, “Orbital-angular-momentum embedded massive MIMO: Achieving multiplicativ e spectrum-efficienc y for mmWav e communications, ” IEEE Access , vol. PP , no. 99, pp. 1–1, 2017. [11] M. Lin, Y . Gao, P . Liu, and J. Liu, “Theoretical analyses and design of circular array to generate orbital angular momentum, ” IEEE T ransactions on Antennas and Propagation , vol. 65, no. 7, pp. 3510–3519, July 2017. [12] S. Y u, L. Li, G. Shi, C. Zhu, X. Zhou, and Y . Shi, “Design, fabrication, and measurement of reflective metasurface for orbital angular momentum vortex wav e in radio frequency domain, ” Applied Physics Letters , vol. 108, no. 12, pp. 5448, 2016. April 23, 2018 DRAFT 14 [13] H. Jing, W . Cheng, Z. Li, and H. Zhang, “Concentric UCAs based lo w-order radio vortex wireless communications with co-mode interference, ” in IEEE/CIC ICCC , Qingdao, Oct. 2017 (Invited Paper). [14] M. Uchida and A. T onomura, “Generation of electron beams carrying orbital angular momentum, ” Nature , vol. 464, no. 7289, pp. 737–9, 2010. [15] Y . Y an, L. Li, G. Xie, M. Ziyadi, A. M. Ariaei, Y . Ren, O. Renaudin, Z. Zhao, Z. W ang, C. Liu, S. Sajuyigbe, S. T alwar , S. Ashrafi, A. F . Molisch, and A. E. Willner , “OFDM over mm-Wave O AM channels in a multipath en vironment with intersymbol interference, ” in 2016 IEEE Global Communications Confer ence (GLOBECOM) , Dec. 2016, pp. 1–6. DRAFT April 23, 2018