Wi-Gig Communication System for NASA Mission Support

Transcription

1 Wi-Gig Communication System for NASA Mission Support ITP Capstone Research Paper 04/28/2017 Gaurav Diwate Kunal Aniruddha Parth Shah Sree Krishna Vinay Natarajan Authors Interdisciplinary Telecom Program University of Colorado Boulder Dr. Kevin Gifford Faculty Advisor Interdisciplinary Telecom Program University of Colorado Boulder Adam Schlesinger Industry Advisor XS Exploration Development Integration Office Abstract This research investigates the utilization of IEEE ad (Wi-Gig) wireless communications protocol to support high data-rate communications in the proximity of the International Space Station (ISS). Specific emphasis is placed upon supporting the transport of high-rate HD video external to the ISS via based systems (ac, n, ad). RF modeling of the ISS external environment and site survey planning is performed to determine optimal transceiver location for an external wireless communications system. Simulation results are verified with hardware testing of commercial-grade Wi-Gig access points. Results show significant Wi-Gig throughput gains as compared to an ac or n system, and surprisingly, greater than expected data throughput and range. Index Terms Millimeter wave propagation, path loss, multipath, fading, delay spread, throughput. I. INTRODUCTION The wireless standard that is currently in use on the International Space Station (ISS) is n. However, with the improvement in quality and size of the media captured in outer space, it is essential that a new technology be deployed on the ISS that can better the current max data rates of approximately 300 Mbps with n. This paper proposes using ad (Wi-Gig) technology to be deployed around the ISS that operates in 60 GHz frequency band. One of the main reasons for using ad is that it can provide multi-gigabit data rates per second. This would facilitate in faster data offloading of the High-Definition (HD) pictures and video. Another reason in using one of the higher frequency protocols is due to the ability of mm-wave to support Internet of things (IoT). With the giant leaps towards having millions of devices connected, mm-wave can aid fast data transfer and device-to-device connectivity in outer space. Adopting ad would make sure that decreased time is spent collecting data by the crew members in dangerous conditions in outer space. Real time biometrics monitoring of the crew members can also be implemented using ad. Any indication of irregularity in the heart rate, blood pressure or any other vital body function of the crew member, he/she would be called back inside the space station at that very instant. The main problem with using Wi-Gig technology is very high path loss associated with 60 GHz frequency. The communication will be limited to line of sight. Currently, the ad standard is used indoors for Audio-Visual applications such as transmission of a video from a digital camera to a projector, High-Definition TV or a monitor. This paper discusses the feasibility of deploying Wi- Gig in space with the help of RF propagation models associated with mm-wave and extending those models to using Wi-Gig. First, outdoor tests were performed using ad access points, to determine the behavior of this technology over variable distance and nodes. Building on this, ultimately the aim is to come up with an RF propagation model that is specific to Wi-Gig communication in deep space. For research purposes, the feasibility of deploying ad for the ISS is considered. However, this technology can be deployed on other spacecrafts as well. 1

2 II. RESEARCH QUESTION The research question addressed in this paper is How to establish high rate (multi-gbps) data communication in proximity to International Space Station by using IEEE ad (Wi-Gig) technology to record ultra-high definition media and space crew biometrics for NASA mission support? High quality data and video is fundamental scientific information in search expeditions and successful analyses. NASA has mounted recording devices on the body of ISS, which are supposed to stream data collected externally to the controlling unit inside. Only a very high throughput wireless communication standard is capable of supporting such ultrahigh definition (4K) data streaming, which makes it a necessity to employ a suitable communication standard. IEEE has developed such a standard called ad (Wi-Gig) which works in unlicensed 60GHz frequency spectrum and is capable of providing a multi-gbps throughput. Therefore, this standard is a candidate option to be further investigated for NASA mission support. Following the scientific approach, the main research problem has been divided into sub-problems, which help to analyze each aspect of the project thoroughly: 1) Design Testing: The first sub-problem of Design Testing was developed to address feasibility of a highly efficient Wi-Gig network for ISS. This includes baseline link tests, network design/implementation and network testing. It has been aimed to design a complete Wi-Gig network and carry out extensive tests based upon a specific test plan. Baseline link tests ensure initial configurations and proper orientation to achieve desired throughput. Network testing will evaluate the applicability of Wi-Gig for the ISS. 2) Wi-Gig Channel and Propagation Modeling: The second sub-problem is developing the RF channel and propagation model specific to Wi-Gig communication external to ISS. An industry standard RF modeling and simulation software (Wireless InSite) is utilized for this purpose. 3) Market Research: A market assessment is performed to compare ad access points from different commercial vendors. This analysis is required for selecting a device that is expected to perform optimally in an ISS-emulated test bed. From this, the current ad market scenario was understood in depth, which is essential in providing an optimum research solution. III. LITERATURE REVIEW High frequency communication comes with its share of problems, and 60 GHz is no different. One of the characteristics of wireless channel at 60 GHz is limited range [1]. Friss formula for free space path loss when calculated for 60 GHz shows a power penalty of about db for the same distance at operating frequencies below 6 GHz. The lack of atmosphere in outer space reduces transmit power attenuation making it easier to extend the 60 GHz ad model for communication external to ISS ad signal propagation inside the International Space Station is different as multipath comes into consideration. Catherine Sham, while working for Johnson Space Center in Houston performed some practical experiments implementing a wireless network inside the ISS in her publication Propagation Characteristics of International Space Station Wireless LAN [2]. The electromagnetic field (E) for direct, reflected and diffracted rays is perfectly computed with the below equation: E = Ei(r) * D * A(s) * e-jks.. (1) where, Ei(r) is the field incident on the reflection or diffraction point r, D is a dyadic reflection or diffraction coefficient, A(s) is a spreading factor and s is the distance from the reflection or diffraction point r to the field point r. D and A(s) are found from the geometry of the structure at reflection or diffraction point r and the properties of the incident wave there. When indoor communication is concerned, material attenuation is also responsible for deteriorated received signals. Moreover, different materials will have different multipath characteristics. To understand the wireless channel at 60 GHz, it is required to determine appropriate system design parameters such as path loss, material attenuation, multi-path effects, antennas and spatial and temporal changes [3]. 60 GHz 2.4 GHz Material 2.4 db/cm 2.1 db/cm Drywall 5.0 db/cm 0.3 db/cm Whiteboard 11.3 db/cm 20.0 db/cm Glass 31.9 db/cm 24.1 db/cm Mesh Glass Table 1 Material attenuation at 60 GHz and 2.5 GHz [4] Significant research has been conducted or proposed regarding the protocol analysis of ad. Since this is a recently deployed standard, there is still a potential for modifications in the PHY/MAC layers and many ideas are being researched to streamline the working of this technology. 2

3 In the initial development stages of Wi-Gig, it was found that traditional omni-directional communication was not possible like the popular 2.4/5 GHz wireless LAN systems. This was due to the fact that the power received in 60 GHz band was dramatically less due to the smaller receive antenna aperture compared to 2.4/5 GHz bands, for identical transmit power and antenna gains [5]. Furthermore, due to other factors like high penetration losses and formation of sharp shadow zones for 60 GHz frequency, the requirement of highly directional antennas or beamforming was a necessity for a reliable communication link. CSMA/CA will be effective for bursty-type applications like web-browsing since it provides lower average latency. TDMA will be useful for applications like video and voice due to better QoS support and efficiency, a feature that is extremely useful in this project since the crew members of ISS performing EVAs will be in constant contact with the ISS preferably through video calls. Polling will be beneficial in dealing with dynamic re-allocation of channel time that is scheduled but unused, which might happen in applications like compressed wireless display [5]. This attribute also will be very convenient for this research as astronauts floating around with wireless display devices will need uninterrupted video playing wirelessly on the devices. The major use of this project will be the transfer of large quantities of high quality video data between crew members and ISS. Although it is possible to stream or transfer uncompressed video over 60 GHz, it is not practical to do so. Uncompressed video consumes huge amount of bandwidth and does not leave enough for other applications. In this case, other critical data like the astronaut s biometrics will need to have sufficient bandwidth at all times, which makes video compression mandatory. As a result, it is necessary to develop 60 GHz communication which supports compressed or variable-bit-rate (VBR) video. Figure 1: Compressed video bit rate [5] This project makes use of Modulation Coding Scheme (MCS) index and the corresponding data rates and compares them with the throughput requirements of VBR video as shown in Figure 1. It is possible to calculate the PHY rates corresponding to the MCS index using the following equation (2): TPphy=Rsym * NCBPS * Nrepetitions -1 * Rcode * Oblocking (2) [6] where: TPphy= PHY data-rate (bits per second) Rsym = Symbol rate (1760 sym/s) NCBPS= Number of bits per symbol Nrepetitions=Number of redundant symbol repetitions Rcode = Coding rate Oblocking= Blocking overhead PHY rates (Mbps) Blocking Overhead Coding rate Redundant symbol repetitions Bit/Sym MCS Table 2 Calculated PHY rates for corresponding MCS for ad [6] For PHY rate calculations, typical coding rate and blocking overhead values were chosen [6]. Looking at the PHY rates in Table 2 and compressed video bit-rates in Figure 1, it can be concluded that MCS 4-12 support variable bit-rate transmission video over a 60 GHz link. A comprehensive set of ad performance simulations have been presented by Carlos Cordiero et.al [7]. These simulations form the basis of our design test setup. The range tests validate the link budget calculations and indoor tests with multiple clients simulate an environment inside the ISS. For calculations inside of ISS, a human blockage model for ad was extended to ISS after referring to calculations performed by Martin Jacob et.al [8]. This prompted us to perform NLOS throughput tests along with human obstruction tests. This research project uses Talon AD7200 Multi-Band Wi-Fi router to perform the design testing. According to the specification sheet for the above access point, the 60 GHz antenna was made up of 32 elements with a gain of 16.2 dbi [9]. This value was taken under consideration for the RF propagation modeling using Wireless InSite. This project involves theoretical RF propagation analysis, software based RSSI predictions and hardware based throughput testing for International Space Station communication. Since ad deployments are still in 3

4 infancy, existing research may not be applicable to most network topologies. Existing research cannot be extended for 60 GHz communication for mission critical applications in space. IV. RESEARCH METHODOLOGY The research planning is divided into two main stages per below: 1) Design Testing: Validating performance of a Wi-Gig (802.11ad) protocol based communication system was an important part of the project. It was needed to validate a complete design and a precise methodology to firmly establish final results so that this state of the art technology could be deployed upon the International Space Station (ISS) successfully. Design Testing is subdivided into three main sections and developed methodologies accordingly. The below subsections explain testing requirements and methods in detail to validate the complete design. Hardware and Software Requirements: Software and hardware equipment to test the network design: IX Chariot Traffic Generator Wi-Gig transceiver sets with MIMO configuration Iperf3 Ethernet cables Trolley loaded with Batteries A) Baseline Link Tests: Baseline link tests are necessary to ensure proper orientation of Wi-Gig transceiver sets, link setup and throughput. This does not need any experimentation in the field, but could be done in a lab environment itself. i. Power-up all Wi-Gig transceiver sets and configure them to set carrier frequency within Wi-Gig band (57GHz - 64GHz, USA standard), set propagation channel 2 (total 3 are available) with bandwidth of 2.1 GHz and to set transmitting power at maximum level (Up to 25dBm). ii. Place Wi-Gig transceivers at a distance of 5 meters from each other and plug in IX Chariot traffic generator with a baseline throughput configuration script. iii. Initiate a traffic stream through Wi-Gig transceiver and observe throughput values. iv. Orient transceiver sets till we get a steady throughput as defined in IX Chariot scripts under run configuration. B) Network Implementation: To establish ultra high-speed video streaming and continuously recording space crew biometrics, we needed a fully fledged network over ISS. The below procedure was developed to design the network and to implement it. i. Communicate with the Faculty adviser and Industry adviser to completely understand ISS design with respect to total size and different modules with their functionalities. ii. Study NASA documents to understand currently utilized video streaming technology, communication system and existing network architecture. iii. Based upon the primary study results and inputs from advisors, design location points for W-Gig transceiver sets to cover entire ISS area. This would satisfy communication with space crew and biometrics recording. iv. After careful study of ISS modules, select a suitable Experimental module to place ultra high definition cameras along with Wi-Gig transmitters. C) Network Testing: This is an important validation process to test the designed network for optimum throughput and to ensure overall performance. The below methodology has been developed to test Network sections. a) Distance vs. Throughput: Main purpose of distance vs. throughput test is to determine the maximum possible distance of propagation with Wi-Gig communication system at high throughput levels. i. Initially place Wi-Gig transceiver sets at a distance of 5 meters from each other and configure IX Chariot traffic generators script for a ultra high throughput. ii. Measure maximum throughput at a short distance. iii. Keep on increasing the distance in steps of 10 meters and note down throughput values. iv. Continue the process till the throughput value drops significantly below a set minimum threshold. b) Network Design Test: Once the maximum possible distance at optimum 4

5 throughput has been successfully established, different sections of the network with respect to throughput at varying elevations and varying distances are tested. i. Set elevations and distances as per design. ii. Set IX Chariot for ultra high throughput and inject traffic streams through Wi-Gig transceiver sets. iii. Measure throughput values for all sections and configurations. iv. Analyze all results. Repeat above steps for maximum possible throughput values for each section and make necessary changes in the network design. 2) Wi-Gig Channel and Propagation Modeling: Channel modeling outside of ISS is simple as compared to that of inside ISS. The channel will behave similar to how it would behave in free space conditions and follow the Friss transmission equations for free space. As a precaution to reduce channel interference, it is necessary that the communications internal to ISS operate at a frequency located further away in space from the frequency where communications outside ISS take place. In order to understand the path loss effects with deployment of ad, Table 3 compares free space path losses at different distances at 2.4 GHz, 5 GHz, and 60 GHz. 60 GHz 5 GHz 2.4 GHz Distance (in meters) db db db db db db db db db db db db db db db db db db db db db 1000 Table 3 Free Space Path loss in db at different distances As per Table 3, there is about 22 db of path loss more at 60 GHz than at 5 GHz and about 28 db path loss more than at 2.4 GHz. To tackle the increased path loss, ad radio will have to transmit signal at an increased power as opposed to 2.4 GHz and 5 GHz radios. However, with highly improved throughput, power is efficiently used as observed by Swetank Kumar Saha et.al [10]. Simulation was initially performed using a single transmitter and receiver. This served as a baseline to provide familiarity with the software and verify the basic distance vs. throughput tests conducted with access points. This simulation was designed with two different antennas, first with omnidirectional antenna and later with MIMO array. This would verify the improvement in the behavior of ad in presence of MIMO antenna array. Following the baseline simulation, the currently deployed network design and technology would be implemented in the software. This test is required to get an idea about the present situation at the ISS. Results obtained after running this simulation can be used for comparison with the final ad simulation results, thus establishing the superiority of ad over n. The last simulation is based on the proposed network design. Placing 5 access points according to the network topology, their behavior will be inspected with a receiver sphere covering major sections of ISS. This will demonstrate ad behavior with multiple nodes. It will also provide all the coverage and capacity parameters. 1) Design Testing V. RESULTS Following the method and procedures designed for the ISS and as described in previous section, extensive tests were performed in both indoor and outdoor environments. A) Baseline Link Tests: Baseline link tests were performed to ensure proper orientations and throughput in a lab environment. Figure 2, below, shows the testing results and a resulting average throughput of 882 Mbps. Figure 2: Baseline link test throughput The above result indicates a proper orientation of Wi-Gig transceiver sets. B) Network Implementation: Specific NASA documents were utilized to determine current (802.11n) access point locations and functionality of ISS modules. Wi-Gig transceiver sets are depicted as CP in the below Figure 3 along with the Kibo back porch which depicts 5

6 the location of a Wi-Gig transceiver set. A series of tests was performed upon the network design for the ISS. Following the procedure as described in the previous section, results were achieved that satisfied the testing requirements for an external wireless communication system. 1) Link Test between CP3 and CP13: Total Distance: 20 meters C) Network Design Test: Figure 3: Initial Network Design for ISS a) Distance vs. Throughput: Through experimentations, we found that throughput drops significantly after 60 meters of a distance between Wi- Gig transceivers. Below figures explains test setup and results. Figure 5: Throughput results for a 20 meters link An average throughput of 882 Mbps was achieved at a range of 20 meters. 2) Link Test between EWC Antenna and Kibo Back Porch: EWC indicates External Wireless Communication Total Distance: 30 meters Figure 4a: Wi-Gig ad test setup Figure 4b: Wi-Gig throughput rate at 60 meters An average throughput of 701 Mbps was achieved at a range of 60 meters. b) Network Design Tests: Figure 7. Throughput results for a 30 meters link 6

7 An average throughput of 774 Mbps was achieved at a range of 30 meters. 3) Link Test between EWC Antenna and CP9: Total Distance: 40 meters 2) Wi-Gig Propagation Modeling: Along with the hardware testing in normal conditions on Earth, an RF propagation model was created for the ISS to simulate Wi-Gig behavior in space. Initially, a 3D ISS model was developed, around which the propagation modeling will be performed. An STL file, defining all materials of the ISS, was developed. Using this 3D model file, a series of simulations were conducted. The transmitters and receivers used for these simulations were setup as per ad specifications. Channel 2 of the ad frequency band is globally used, so that channel was chosen with a center frequency of GHz and bandwidth 2.16 GHz. The transmit power was set to 30 dbm. As specified in the methodology previously, both omnidirectional and MIMO antennas were used. The first run consisted of a simple point-to-point simulation with one transmitter and receiver. This was similar to the baseline test in the design-testing sub-problem, just to verify software functionality as well as behavior of Wi-Gig with a single node. Figure 7: Throughput results for a 40 meters link An average throughput of 754 Mbps was achieved at a range of 40 meters. 4) Link Test between EWC Antenna and CP8: Total Distance: 50 meters Figure 8: Throughput results for a 50 meters link An average throughput of 729 Mbps was achieved at a range of 50 meters. Fig. 9. Point-to-point simulation The transmitter was placed on JEM back porch and propagation was measured in a sphere of radius 60m around that transmitter. As seen in figure 9, the received power can be seen in the sphere. The scale ranges from color red for -73 7

8 dbm and blue for -124 dbm. As seen here, the received power at a distance of 60m is very good and sufficient for any industry-grade receiver to detect and process without any corruption or errors. Fig. 10. Path Loss Figure 10 corresponds to the path loss. The path loss remains in the range db, which is expected at such high Wi-Gig frequencies. There are some high spikes, which are minor simulation abnormalities and it remains constant over the measured distance. The delay spread, shown in the Figure 11, goes up to a maximum of 50 microseconds, which is manageable for processing a multipath output. All these parameter calculation outputs demonstrate that point-to-point communication using Wi-Gig in space are reliable at a distance of 60 meters. Fig. 12. Simulation with Omni directional Antenna First, simulation was done using omnidirectional antennas for transmitters and receivers. The above figure shows received power, with the scale ranging from -101 dbm (dark blue/purple) to -68 dbm (red). Significant portion of the receiver sphere is purple in color. Received power lower than - 85 dbm is not sufficient to be processed without any corruption. As a result, ad with omnidirectional antennas is not feasible. Fig. 11. Delay Spread The second run consisted of a full network implementation, with 5 transmitters installed on the ISS, according to the topology shown in Figure 3. A receiver sphere of 180 meters in radius from the center of ISS was used, which covered almost the entire ISS. Fig. 13. Simulation with MIMO Antenna Configurations The next step was implementing the same network with MIMO-capable transmitters and receiver. A MIMO array 8

9 of 2X2 was used. Figure 13 displays the received power for this simulation. The scale for this simulation is from -153 dbm (purple) to -71 dbm (red). As seen in the figure, almost the entire sphere is reddish orange, which signifies that the received power is in the range -71 to -85 dbm, which is adequate power for any receiver to process the signal. 3) Market Research: Through a quantitative approach, the best transceiver for our requirements was found. The characteristics of the transceivers were evaluated along with their specifications and therefore decide the applications that can utilize these transceivers. Currently, Qualcomm produces ad chipsets on a commercial scale. The first company to use it on their devices is Acer, a laptop manufacturer. Currently, there are multiple vendors in the Wi-Fi market ad Access Points from multiple vendors were assessed for their performance characteristics. The devices, which could perform optimally in the unpredictable conditions of outer space, were chosen. All the test results were obtained using TP-Link Talon AD7200 routers. Currently, there are 2 major vendors who sell ad routers in the market. They are NetGear and TP-Link. Fig. 14. Simulation for Path Loss Figure 14 shows path loss for the full network implementation. The scale for this simulation is dark blue (99 db) to red (193 db). Almost the complete sphere is a dark blue shade. Therefore, it puts the path loss in the range db, similar to the point-to-point implementation. Fig. 15. Path Loss Analysis Figure 15 shows path loss vs. distance. As observed, only 2 abnormal spikes are present over the entire simulation. Overall, it remains within the manageable limit of db. These are favorable results for the simulation and provides evidence that ad is sustainable at much farther distances than theoretically computed. The TP-Link AD7200 router was the first ad router to be released in the market. This router supports 4X4 MU-MIMO and can achieve data rates upto 4.6 Gbps in the 60 GHz domain and achieve a cumulative data rate of 7.2 Gbps in the tri-band domain. The NetGear Nighthawk X10 is an alternative to TP-LinkAD7200. This is the first router from Netgear to feature ad. This router also provides a cumulative data rate of 7.2 Gbps and supports quad stream MU-MIMO. An interesting feature of this router is that the amplifiers are present in the antennas of the router thereby reducing internal noise and interference and leads to better signal quality. Both the routers use beam-steering technology, which is crucial to the functioning of ad. Some basic tests on the Access Points were carried out to verify various parameters such as throughput vs. distance. The chosen Access Point was decided using a comparative analysis of the data gathered from these PHY layer tests. A precise network design, extensive network tests and substantial market research are the output of state-of-theart interdisciplinary coursework. The project has made significant contribution to the current knowledge of Wi-Gig communication system. Currently, the market research and vendors project Wi-Gig (802.11ad) as a short distance protocol (10 meters). But, test results of this project clearly show that Wi-Gig performs well even at the range of 60 meters. VI. FUTURE SCOPE This project provides a decent scope to address time sensitive applications in a space environment. Current network design can be further improved to communicate and coordinate with space shuttles approaching ISS such as Orion capsule. Such a network would require advanced transmission 9

10 time synchronization ability to maintain a high throughput all the time. Further research in this specific direction with more test cases would surely help to address these applications. VII. CONCLUSION This research project was successful by performing a strong background research, implementing detailed Wi-Gig network for ISS and extensively testing all network sections with Wi-Gig transceiver sets. This project also validates test results with simulations using Wireless InSite software. Based upon all results, this project confirms the feasibility of deploying Wi-Gig communication system in the space for ISS providing a very high throughput efficiently around the entire ISS for time sensitive applications such as ultra-high definition video recording and space crew biometric monitoring. VIII. ACKNOWLEDGEMENT IEEE ad: Introduction and Performance Evaluation of the First Multi-Gbps Wi-Fi Technology. In Proceedings of the 2010 ACM International Workshop on mmwave Communications: From Circuits to Networks, mmcom 10, pages 3 8, New York, NY, USA, ACM [8] M. Jacob et al., "Extension and validation of the IEEE ad 60 GHz human blockage model," th European Conference on Antennas and Propagation (EuCAP), Gothenburg, 2013, pp [9] "AD7200 Multi-Band Wi-Fi Router", [Online]. Available: [Accessed: 25- Apr- 2017] [10] Swetank Kumar Saha, Viral Vijay Vira, Anuj Garg, and Dimitrios Koutsonikolas, A Feasibility Study of 60 GHz Indoor WLANs, University at Buffalo, The State University of New York. We would like to thank Dr. Kevin Gifford, Adam Schlesinger and Dr. David Reed for providing valuable insight regarding all the technical aspects of our project and guiding us throughout the research. Also, a special thank you to Dr. Thomas Schwengler for providing us with the hardware required to conduct our design testing and providing us access to wireless testing chamber. IX. REFERENCES [1] R. C. Daniels and R. W. Heath Jr., "60 GHz wireless communications: emerging requirements and design recommendations," in IEEE Vehicular Technology Magazine, vol. 2, no. 3, pp , Sept doi: /MVT [2] C. Sham, S. Hwu and Y. Loh, "Propagation Characteristics of International Space Station Wireless Local Area Network,"Johnson Space Center Houston, Houston, [3] R. C. Daniels and R. W. Heath Jr., "60 GHz wireless communications: emerging requirements and design recommendations," in IEEE Vehicular Technology Magazine, vol.2,no.3,pp ,Sept doi: /mvt [4] Alexander Maltsev et al., Channel Models for 60 GHz WLAN Systems, IEEE , May [5] M.Park, E.Perahia, C.Cordeiro and L.L.Yang, QoS Considerations for 60 GHz Wireless Networks, in IEEE GLOBECOM Workshops, 2009 IEEE. doi: /GLOCOMW [6] D. Holmes-Mitra, "A Detailed Characterization of 60 GHz Wi-Fi (IEEE ad)", Hdl.handle.net, [Online]. Available: [Accessed: 25- Apr- 2017]. [7] Carlos Cordeiro, Dmitry Akhmetov, and Minyoung Park. 10