引言
?数字化是进一步经济增长的关键。但是，如果有合适的电信基础架构，那么数字化只能提供所需的改进。5G将提供此基础架构。 公开讨论主要是由诸如传感器和大数据，工业4.0，自动驾驶，智能轨道等想法推动的。当然，工件与铣床之间的通信不需要卫星。同样，对于自动驾驶或对时延要求苛刻的其他应用，也不能使用对地静止卫星。但是，将来的大部分应用程序仅需要高数据吞吐量。 当前，约60％的移动数据量是由视频流引起的。 在接下来的5年中，预计将增加到74％[1]。 视频流是对地静止卫星的核心业务。
?其他重要方面是要克服数字鸿沟和现代工作模式的推动力。 地面基础设施的现代化将需要大量资源。因此，它将主要发生在城市和大都市地区。 存在已经存在的数字鸿沟将增大的风险。 迁入城市将增加生活成本，而农村人口外流将导致农村地区生活质量进一步下降。 例如在德国，许多高度创新的中小企业都位于农村地区。 他们的贡献对于经济增长至关重要。 对于农业数字化而言，这也是未来成功的关键动力。
?卫星可以补充地面电信基础设施。 在没有或没有现代地面基础设施的地区，卫星通信至少可以用于不需要低延迟的应用。 典型示例包括视频流，用于制造的数字模型交换，软件维护和更新，远程教学和社交网络。 视当地情况而定，在地面基础设施可用之前，卫星通信可以是一种桥接技术，也可以是地面解决方案过于昂贵的地区的永久解决方案。 该卫星将用于5G网络的回程。 不管回程是通过有线，地面无线还是通过卫星实现，用户侧都是相同的
II. requirements and concepts for future high throughput satellites
A. Antenna scenarios
High throughput satellites cover the service area with small high gain spot beams. For their antennas usually the single feed per beam (SFB) or multiple feeds per beam (MFB) principle are used [2] [3]. The feed systems are complex assemblies of waveguide components manufactured by milling, turning, spark erosion or other traditional technologies [4]
高通量卫星以小的高增益点波束覆盖了服务区域。 对于它们的天线，通常使用单波束每馈电（SFB）或多波束每波束（MFB）的原理[2] [3]。 馈送系统是通过铣削，车削，电火花腐蚀或其他传统技术制造的波导组件的复杂组件[4]
Key parameter for satellite providers are the costs per bit. Eutelsat-KaSat, one of the currently largest satellites serving Europe, has a capacity of about 90 Gbps, spread over 82 beams [2]. Such satellites inclusive launch and ground segment needn typical Capex of 300-500 Me. Target value for the cost of future high throughput satellites are 1 Me/Gbps and below [5]. This means a massive cost reduction is required or a significant increase of capacity without additional costs.
卫星提供商的关键参数是每比特成本。 Eutelsat-KaSat是目前服务于欧洲的最大卫星之一，容量约90 Gbps，分布在82束波束上[2]。 包括发射和地面部分在内的此类卫星通常需要300-500 Me的资本支出。 未来高吞吐量卫星成本的目标值为1 Me / Gbps或更低[5]。 这意味着需要大量降低成本，或显着增加容量而无需额外成本。
A promising way to increase capacity significantly, is a higher frequency re-use factor. This can be achieved, if the same coverage is served by more but smaller beams. If the beam diameter is reduced by a factor of 2, the area is reduced by a factor of 4 and so frequency re-use factor and capacity increase by a factor of 4. This increase is achieved without an increase of RF power. This means there are no additional costs for power amplifiers or thermal hardware.
显着提高容量的一种有前途的方式是更高的频率复用系数。如果相同的覆盖范围由更多但更小的波束提供服务，则可以实现此目的。如果将光束直径减小2倍，则将面积减小4倍，因此频率重用因子和容量增加4倍。在不增加RF功率的情况下实现了这种增加。这意味着功率放大器或热硬件没有额外的成本。
Larger reflectors enable smaller beams, but the number of beams and so the number of feeds has to be increased. As the available space for the accommodation of the feed clusters does not increase, the feeds have to become smaller. For the same reason also a lower mass is required. Our heritage polariser for single feed per beam antennas was designed for a beam spacing of 60 mm and has a mass of 300 g [6]. In order to increase the number of feed by a factor of 4 without an increase of required room or mass, a new small and lightweight polariser has been developed in the frame of the Artes C&G programme Ka-band User Feed (KaUF). The new polariser (Fig. 1) enables a horn spacing of 30 mm and has a mass of only 70 g. Compared to our heritage polariser four times more polarisers can be accommodated without an increase of total mass or space. Further savings are possible by the use of waveguide panels instead of single waveguide runs. Within a panel, waveguides can share side walls. Additionally to the distribution of the RF power, the waveguide panel can also be used as structural support for the feed systems.
较大的反射镜可实现较小的光束，但是光束的数量以及进给的数量必须增加。由于容纳馈进团的可用空间没有增加，因此馈给必须变小。出于相同的原因，还需要较小的质量。我们用于每波束单馈天线的传统偏振器的波束间距为60 mm，质量为300 g [6]。为了在不增加所需空间或质量的情况下将馈送数量增加4倍，在Artes C＆G计划Ka波段用户馈送（KaUF）框架中开发了一种新型的小型轻量型偏振器。新的偏振器（图1）使喇叭间距为30 mm，质量仅为70 g。与我们的传统偏光片相比，偏光片可以多四倍容纳而没有增加总质量或空间。通过使用波导面板而不是单个波导线路，可以进一步节省成本。在面板内，波导可以共享侧壁。除了分配射频功率外，波导面板还可用作馈进系统的结构支撑。
The new polariser can be used for SFB scenarios as well as MFB scenarios. As a test case we chose a four colour MFB application in Ka-band, where each beam is fed by four feeds. Adjacent beams share feeds with opposite polarisation. In this case we can limit the number of reflectors to two, whereas SFB applications usually need three or four reflectors [3]. The reflector has a diameter of 3:5 m and a focal length of 6:7 m. Such an antenna geometry can be accommodated on an Airbus Eurostar Neo bus using a solid shell reflector. For larger diameters unfurlable reflectors are required. The coverage is shown in Fig. 2. Service area are Europe and North Africa. In total 403 spots with a diameter of 0:3° were realised.
新的偏振器可用于SFB场景以及MFB场景。 作为测试用例，我们选择了Ka波段的四色MFB应用，其中每个光束由四个馈源馈送。 相邻光束共享具有相反极化的馈源。 在这种情况下，我们可以将反射器的数量限制为两个，而SFB应用程序通常需要三个或四个反射器[3]。 反射器的直径为3：5 m，焦距为6：7 m。
这样的天线几何形状可以使用固体反射器安装在 Airbus Eurostar Neo客车上。 对于更大的直径，需要不可卷曲的反射器。 覆盖范围如图2所示。服务区域为欧洲和北非。 总共实现了403个直径为0：3°的地点。
Fig. 3 and Fig. 4 show the predicted directivity for transmit and receive, respectively. Thanks to the high transmit directivity of more than 51 dBi for 95 % of the coverage, less the 10 W RF power per beam are needed. This enables the replacement of travelling wave tube amplifiers (TWTA) by solid state power amplifiers (SSPA). Gallium Nitrate SSPA can provide this power in the 20 GHz band [7].
图3和图4分别显示了发射和接收的预测方向性。 由于95％的覆盖范围具有超过51 dBi的高发射方向性，因此每束光所需的10 W RF功率更少。 这样就可以用固态功率放大器（SSPA）代替行波管放大器（TWTA）。 硝酸镓SSPA可以在20 GHz频段提供这种功率[7]。
Additionally to the directivity, the carrier to interferer ratio (C/I) is a key parameter for the available capacity. Because of the highly asymmetric data use for typical broadband application, the C/I for the user downlink is of special importance.
The predicted values for the transmit C/I is shown in Fig 5 and the receive C/I in Fig. 6. The achieved values of 13 dB for transmit and 11 dB for receive for 95 % of the coverage enable the use of efficient modulation schemes and therefore a high data capacity。
除方向性外，载波与干扰源之比（C / I）是可用容量的关键参数。 由于典型的宽带应用使用高度不对称的数据，因此用户下行链路的C / I尤为重要。 发射C / I的预测值如图5所示，接收C / I的预测值如图6所示。对于95％的覆盖范围，发射的13 dB的接收值和接收的11 dB的实现值使得能够使用有效的调制 方案，因此具有很高的数据容量。
B. Manufacturing methods
With the increasing number of feeds, their manufacturing and assembly cost become a driver for the overall cost. Expensive and time consuming manual work must be removed where possible.
For an efficient use of robots the number of items is not high enough. A promising approach is the use of additive manufacturing methods. Many additive manufacturing methods provide a higher level of complexity for the manufactured products without extra costs. This enables the manufacturing of entire feed system as one piece. There is no need for assembly and integration work. Also cost for drawings, documents and procurement reduce significantly. Design engineers can send their digital models directly to the manufacturer and receive an entire sub-assembly. Classical feed systems for high throughput satellites would consist of several hundred of RF components and tens of thousands screws, nuts, Steel and Invar washers. All assembly steps including half and full torque or head locking would have to bedocumented. Additive manufacturing provides here a massive potential for cost reduction.
B.制造方法
随着饲料数量的增加，它们的制造和组装成本成为总成本的驱动力。在可能的情况下，必须取消昂贵且费时的手动工作。 为了有效利用机器人，物品数量不够高。一种有前途的方法是使用增材制造方法。许多增材制造方法为制成品提供了更高级别的复杂性，而没有额外的成本。这使得整个进料系统可以整体制造。无需进行组装和集成工作。图纸，文件和采购成本也大大降低。设计工程师可以将其数字模型直接发送给制造商，并接收整个子装配。用于高通量卫星的经典馈电系统将由数百个RF组件和数万个螺钉，螺母，钢制和因瓦垫圈组成。所有组装步骤（包括半扭矩和全扭矩或头部锁定）必须记录下来。增材制造为降低成本提供了巨大的潜力。
To the most developed additive manufacturing methods belongs direct metal laser sintering (DMLS) or direct metal laser melting (DMLM). The products are printed by a laser in a powder bed. We use the Aluminium powder AlSi10Mg. This material has electrical and mechanical properties very similar to the Aluminium alloy 6061, which we usually use for conventional waveguide components. Fig. 7 shows a Kuband cluster printed as one single piece on an EOS M400. An EOS M400 has a 1 kW Yb-laser and a building volume of 400 mm x 400 mm x 400 mm and is therefore suitable for printing of large metallic parts. The measured S-parameter of selected beams are shown in Fig. 8. Further details can be found in [8].
最发达的增材制造方法属于直接金属激光烧结（DMLS）或直接金属激光熔化（DMLM）。 产品通过激光在粉末床中印刷。 我们使用铝粉AlSi10Mg。 这种材料的电气和机械性能非常类似于我们通常用于常规波导组件的铝合金6061。 图7显示了在EOS M400上整体打印的Kuband集群。 EOS M400具有1 kW的Yb激光器，建筑体积为400 mm x 400 mm x 400 mm，因此适用于大型金属零件的印刷。 所选光束的实测S参数如图8所示。更多详细信息，请参见[8]。
lthough DMLM is meanwhile an established manufacturing method for complex feed systems, the surface roughness of the components limits its use for higher frequencies. A promising alternative for high frequencies is metal powder application (MPA). MPA technology stands for a thermal spray process where metal powder particles are compacted layer by layer to massive solids. To do so, the powder particles are accelerated to very high speed by means of a carrier gas and then deposited on the substrate via a laval nozzle [9].
尽管DMLM同时是复杂进给系统的既定制造方法，但组件的表面粗糙度限制了其在更高频率下的使用。 高频的一种有希望的替代方法是金属粉末应用（MPA）。 MPA技术代表热喷涂工艺，其中金属粉末颗粒逐层压实成块状固体。 为此，粉末颗粒借助载气加速至非常高的速度，然后通过拉瓦尔喷嘴[9]沉积在基材上。
The Ka-band feed cluster in Fig. 9 was developed in the frame of the DLR-sponsored programme “Akradia” and manufactured by a hybrid of milling and MPA. Thanks to this hybrid approach, accuracy and surface roughness are as good as for milled parts, but no screws are needed for assembly. Fig. 10 shows a comparison between measured and predicted antenna pattern at 18.75GHz. Co- and cross-polarisation agreed very well.
图9中的Ka波段进料集群是在DLR赞助的计划“ Akradia”的框架中开发的，并通过铣削和MPA的混合制造而成。 由于采用了这种混合方法，因此精度和表面粗糙度与铣削零件一样好，但是装配不需要螺钉。 图10显示了在18.75GHz处测得的天线图案和预测的天线方向图之间的比较。 共极化和交叉极化非常一致。
III. CONCLUSION
A significant part of the data volume of future 5Gapplications will be caused by video streaming or other applications without a need for low latency. Geostationary satellites are suitable to provide such services for areas where a terrestrial infrastructure is not or not yet available. To fulfil cost targets, the capacity of future satellites has to be increase by at least a factor of four, compared to the current satellites in service. This must be achieved without an increase of cost or mass. Digitalisation and innovative manufacturing technologies like additive manufacturing will make it possible to achieve these targets. Future geostationary satellites can deliver a significant contribution to the provision of 5G services especially in rural regions.
未来5G应用的数据量的很大一部分将由视频流或其他应用引起，而无需低延迟。 对地静止卫星适合为地面基础设施不可用或尚不可用的地区提供此类服务。 为了实现成本目标，与目前使用的卫星相比，未来卫星的容量必须至少增加四倍。 这必须在不增加成本或质量的情况下实现。 数字化和增材制造等创新制造技术将有可能实现这些目标。 未来的地球同步卫星将为5G服务的提供做出重大贡献，尤其是在农村地区。
ACKNOWLEDGMENT
This work has been supported by the DLR programme “Akradia” and the ESA Artes programme “Ka-band User Feed”. The authors are grateful for this valuable support.
DLR程序“ Akradia”和ESA Artes程序“ Ka-band User Feed”为这项工作提供了支持。 作者非常感谢您的宝贵支持.