Читать книгу Advanced Antenna Array Engineering for 6G and Beyond Wireless Communications - Richard W. Ziolkowski - Страница 16

Clearly, the number of antenna ports and radios for 5G systems will grow dramatically with the increasing numbers of massive antenna arrays and operating bands. This growth implies that the number of cables that connect the radios to the antennas would increase accordingly. This increase necessarily leads to increased fabrication complexities, losses in the cables and connectors, and difficulties in the control of passive intermodulation (PIM) and testing. To mitigate these problems, one needs to change antenna system design methodologies to introduce much higher levels of integration. To this end, there has been a high expectation that 5G antennas, the mm‐wave band antennas in particular, will become highly integrated systems.

Integrated antenna and radio systems eliminate the need for multiple cables between the radios and antennas, thus increasing their reliability by reducing part counts and handling, and simplifying their testing and installation. As a result, there has been an increasing need for effective antenna‐in‐package (AiP) solutions. In addition to managing the radiation performance of the antenna elements and arrays, one must consider several issues for AiP designs. These include, for instance, the materials; process selection and control; power and heat management; and new testing techniques. As an example, Figure 1.1 shows a 64‐element AiP system at 28 GHz. It has four flip‐chip‐mounted transceiver ICs that support its dual‐polarized operation [8]. For clarity, the heat sink below the ball‐grid‐array (BGA) interface is not shown.

Figure 1.1 An illustration of a 64‐element antenna‐in‐package (AiP) assembly breakout.

Source: From [8] / with permission of IEEE.

One particular new challenge associated with highly integrated 5G antenna arrays is obtaining accurate antenna beam patterns. Depending on their actual implementation, methods for testing active antennas vary. Current examples include the following [4]:

 a) Sample Testing

This approach involves the fabrication of a number of fixed analog beamforming circuits that provide the requisite amplitude and phase excitations to the antenna array to produce the desired beams including narrow beams for user traffic and broad beams for user management. Each circuit produces one specific beam. This allows one to sample each of the desired beam types and steering directions. For practical reasons, it is difficult to perform a comprehensive test of all of the possible beams generated by a large array. Therefore, only those beams of greatest interest are likely to be tested.

 b) Element‐by‐Element Testing

The far‐field vectorial pattern of each element, i.e., the amplitude and phase distribution in the far‐field of the array, can be measured with respect to a common reference. Any beamforming pattern can then be synthesized numerically by adding all the element patterns with the corresponding appropriate complex weights. This approach is the most flexible method since all possible patterns can be tested. Nevertheless, one can argue realistically that the synthesized beam patterns may differ from the real ones to a certain extent because all of the actual interactions are not explicitly included.

 c) Employ Beam Testers

Beam testers are effectively flexible beamforming networks. By connecting a beam tester to an antenna array, one can test a variety of the beams defined by the beam tester using a traditional method for antenna pattern testing. The 3rd Generation Partnership Project (3GPP), which unites seven telecommunications standard development organizations (ARIB, ATIS, CCSA, ETSI, TSDSI, TTA, and TTC), has defined three Over‐the‐Air (OTA) test methods for MIMO antennas: the direct far‐field (DFF) method using a far‐field chamber, the indirect far‐field (IFF) method using a compact range, and the near‐field to far‐field transform (NFTF) method using a near‐field chamber. All three OTA approaches are conventional methods familiar to antenna engineers.

It must be recognized that when active electronics are added to a radiating aperture to form a MIMO antenna, the antenna ports are now embedded in the system. As a result, it becomes much more difficult to measure the true gain and antenna efficiency. Because a massive MIMO antenna has a large number of antenna elements and its radiating aperture can be excited in many ways to create different beams, both narrow and broad, it is truly difficult to fully test and validate beam performance in terms of conventional figures of merit, e.g., pattern characteristics, beam shapes, beam steering, side lobe levels, and null locations. Testing is further complicated because measurements for both the transmit case and the receive case must be performed to understand the operating characteristics of both RF chains.

To facilitate the manufacturing and adoption of large antenna arrays in 5G and beyond systems, the wireless industry is pushing to increase the level of integration of the system frontend modules (FEM). Figure 1.2 shows the AiP roadmap of the TMY Technology (TMYTEK) company for their 5G mm‐wave products [9]. Each enclosure block represents one particular level of component integration. The industry trend is to integrate the antenna arrays with all of the radio frequency (RF) and intermediate frequency (IF) modules into one package. Characterization of all of the beams produced by such modules is undoubtedly a new challenge for antenna designers.

Figure 1.2 Three levels of AiP implementation by TMYTECH.

Source: From [9] / with permission of TMY Technology Inc.