Technical Annex. This criterion corresponds to the aggregate interference from a co-primary allocation for month.

Transcription

1 RKF Engineering Solutions, LLC th St. NW, Washington, DC Phone Fax Protection of In-band FSS Earth Stations Technical Annex 1.1 In-band Interference Protection Criteria for FSS Earth Stations Both long-term and short-term interference criteria are considered when assessing the in-band interference into FSS receiver earth stations. In line with Recommendation ITU-R S.1432, the following criterion is identified for use for the longterm case when assessing in-band interference into FSS receive earth stations: This criterion corresponds to the aggregate interference from a co-primary allocation for month. of a given In line with Recommendation ITU-R SF.1006, the following criterion is identified for use for the short-term case when assessing single-entry in-band interference into FSS receive earth stations: which may be exceeded up to time. For cases in which the long-term interference criterion applied, of the allowable interference to the FSS earth station receiver was allocated to CBSD systems. 1 This results in a reduction of the protection criterion by db: For cases in which the short-term interference criterion was applicable, allowance was allocated to the CBSD system. of the interference The SAS will have to calculate the total interference power at each FSS earth station in order to determine if the protection criterion is exceeded. This determination can be performed using an aggregate Effective Power Flux Density (EPFD) calculation methodology. Use of EPFD rather than PFD is preferable because the FSS earth station pointing direction is taken into account in the EPFD interference calculation. In contrast, using PFD would require using the worst-case assumption regarding the FSS earth station pointing angle (5% elevation angle). 1 This apportionment reflects the fact that there are other co-primary users in the band (e.g., Federal allocations).

2 1.2 Challenges in Protecting FSS Earth Stations The total received interference power from CBSDs at an FSS earth station receiver will be a function of: (i) the EIRP density of each CBSD transmitter in the direction of the FSS receiver (which depends on the CBSDs maximum EIRP density and the antenna pattern and orientation); (ii) the FSS receiver gain in the direction of each CBSD transmitter (which depends on the FSS receivers antenna pattern and orientation); (iii) the distance between the FSS receiver and each CBSD transmitter; and (iv) the intervening terrain between each CBSD transmitter and the FSS receiver. In order to accurately determine whether the prescribed aggregate protection criterion is met, the following technical challenges and shortcomings have to be addressed: a. Many technical and operational characteristics of CBSDs that will significantly affect potential interference into the FSS earth stations have not been defined, including whether the CBSDs will be TDD or FDD, can use beamforming to increase terminal gain, use power control, dynamically assign traffic, etc. b. In theory, if an SAS database could calculate the aggregate interference from all CBSD transmitters in real time at a given FSS receiver location, it would be able to ensure that the aggregate CBSD transmissions will not exceed the prescribed aggregate criterion by instructing individual CBSDs either to not transmit on a particular frequency or to reduce power. This is at best a very complicated calculation that will have to be updated constantly in real time to account for a broad variety of changing circumstances. Some of the modeling challenges are discussed in the paragraphs that follow. c. The SAS will have to take into account constantly changing CBSD deployments. The number and location of active CBSDs will change over time, with more significant impacts from CBSDs dynamic traffic scheduling on a frame basis (e.g. msec or msec frames). On each frame, the CBSD Base Stations (BS) can be communicating with different user terminals, and the user terminals may also have different resource block assignments on each frame. Modulation will be changing, and user terminals may use power control. The practical feasibility of taking into account this variation in traffic over time has not been proven. d. Another significant technical obstacle is with regards to changes in CBSD deployment and traffic variations. Conservative propagation models that include significant fade margins to account for inaccuracies and variability would be needed to ensure protection of the FSS earth stations. Propagation modeling will be non-line-of-site (NLOS), which means a rapidly changing multipath environment. Without conservative propagation models, it is not clear how the SAS deployment modeling will handle this variability. Most propagation models are two dimensional, but it will be necessary to take into account the multipath in three dimensions to model NLOS environments. It will be important to investigate the accuracy of proposed modeling approaches, as the variation in propagation loss can be quite large even in small geographic areas. In addition, these models cannot be expected to include up-to- RKF Engineering Solutions, LLC th Street NW, Washington DC Page 2

3 date clutter. Construction and other changes to the landscape would have to be kept up-todate. If aggregate measurements into the FSS receiver are not accurate or if worst case modeling assumptions are not used, FSS earth stations may experience interference despite SAS analysis showing the applicable interference criteria are met. It is not clear that the SAS can then measure and identify the terminals causing the highest levels of interference in a deployment of thousands of terminals. The ability of an SAS to police compliance with the aggregate criterion would, of course, have to be validated (with strong security measures to prevent bypass), as discussed in Section 3 below. 1.3 Sample Receive Power Limit Calculation For any given CBSD deployment, the aggregate interference from all of the CBSDs will have to be evaluated. Given an I/N threshold and depending on both FSS earth station and CBSD system characteristics and deployment scenarios, a received power limit for the aggregate interference power from all CBSDs managed by the SAS can be calculated. For example, for an I/N = -13 db, assuming an FSS earth station receiver system noise temperature of and a CBSD bandwidth of, we can evaluate the FSS earth station noise power. Therefore in this case, the aggregate interference power from all CBSDs managed by the SAS would have to be less than at the FSS receiver in order not to exceed the long-term threshold of. This aggregate power may change depending on the FSS earth station and CBSD system characteristics and deployment scenarios. As described in Section 1.2, the SAS will need to manage and calculate all of these changing characteristics in real time to ensure protection of FSS receivers based on the appropriate receive power characteristics for that particular FSS earth station and CBSD deployment characteristics Single Entry CBSD Interference Simulations It is important to note that a single user could exceed the interference criteria if deployed within a certain proximity to an FSS earth station. The following analysis calculates a protection area (or contour) around each of the 37 in-band FSS receive earth stations 2 in which a single CSBD could exceed the interference criteria. If deployment of CBSDs is permitted within this protection area, the SAS will need to be capable of implementing the necessary measures (as described in discussion points a-d in Section 1.2) in order to ensure that the aggregate interference criteria is not exceeded. The protection areas were determined as described below, using the modeling assumptions of the FSS earth stations, CBSD transmitters and propagation path described in sections 1 through 3 of the Appendix. For each FSS earth station (ES) pointing direction to the GEO arc (with ES elevation angle varying from 5 East to 5 West), a contour was generated via the following methodology: 2 Per FNPRM Appendix 2, table of FSS earth stations in MHz. RKF Engineering Solutions, LLC th Street NW, Washington DC Page 3

4 o Two CBSD terminal types (a non-rural base station (BS) and a fixed point-topoint station) were modeled with EIRP densities of and, respectively. o A single CBSD interferer was oriented at a given azimuth angle relative to the ES location. The CBSD station distance was varied relative to the ES location to find the maximum distance where the CBSD interference exceeded the interference criteria. o This calculation was repeated for every azimuth direction from the ES to produce a maximum interference contour. The protection area contour is the maximum envelope of all the above contours. The resulting contours around each FSS earth station, for single-entry interference simulation described above, are illustrated in Figures 1 and 2 for the long-term interference case ( db not exceeded for more than of the time) and Figures 3 and 4 for the short-term interference case ( db not exceeded for more than of the time), where the FSS earth station is at in all figures. Figure 1: Single-Entry Protection Area Contour in km (East(right), West(left), North(top), South(bottom)) - Non-rural CBSD (BS) EIRP density = 30 dbm/10 MHz for the long-term interference case. RKF Engineering Solutions, LLC th Street NW, Washington DC Page 4

5 Figure 2: Single-Entry Protection Area Contour in km (East(right), West(left), North(top), South(bottom)) Fixed Point-to- Point CBSD EIRP density = 53 dbm/10 MHz for the long-term interference case. Figure 3: Single-Entry Protection Area Contour in km (East(right), West(left), North(top), South(bottom)) - Non-rural CBSD (BS) EIRP density = 30 dbm/10 MHz for the short-term interference case. RKF Engineering Solutions, LLC th Street NW, Washington DC Page 5

6 Figure 4: Single-Entry Protection Area Contour in km (East(right), West(left), North(top), South(bottom)) Fixed Point-to- Point CBSD EIRP density = 53 dbm/10 MHz for the short-term interference case. Figure 5 and Figure 7 show the four protection area contours around two representative FSS earth stations: #12 in Medley, FL, and # 29 in Alexandria, VA, in Google Earth. Figure 6 and Figure 8 show the size of the corresponding contours. Figure 5: Single-Entry Protection Area Contours around FSS earth station in Medley, FL (ES #12) at N, W. From innermost contour, the contours correspond to Long-term interference with CBSD EIRP density=30dbm/10 MHz (Green), Long-term interference with CBSD EIRP density= 53dBm/10 MHz (Blue), Short-term interference with CBSD EIRP density=30 dbm/10 MHz (Red filled), and Short-term interference with CBSD EIRP density=53 dbm/10 MHz (Red unfilled). RKF Engineering Solutions, LLC th Street NW, Washington DC Page 6

7 Figure 6: Protection Area Contours around FSS earth station in Medley, FL (ES # 12). Figure 7: Single-Entry Protection Area Contours around FSS earth station in Alexandria, VA (ES #29) at N, W. From innermost contour, the contours correspond to Long-term interference with CBSD EIRP density=30dbm/10 MHz (Green), Long-term interference with CBSD EIRP density=53dbm/10 MHz (Blue), Short-term interference with CBSD EIRP density=30 dbm/10 MHz (Red filled), and Short-term interference with CBSD EIRP density=53 dbm/10 MHz (Red unfilled). RKF Engineering Solutions, LLC th Street NW, Washington DC Page 7

8 Figure 8: Protection Area Contours around FSS earth station in Alexandria, VA (ES # 29). As indicated by the simulation results, protection areas are large, and there is variation among the 37 FSS earth stations. Using the long-term interference criterion, the protection areas have radii tens to hundreds of kilometers (for the higher CBSD EIRP), and using the short-term interference criterion, the protection areas have radii hundreds of kilometers long. The protection areas vary with the ES azimuth direction toward the interferer and the ES pointing direction to the GEO arc. These results also imply that CBSDs deployed in the United States could cause interference into FSS earth stations deployed in neighboring countries. 1.5 Aggregate CBSD Interference Simulations In order to understand the magnitude of the interference problem that will have to be managed, an aggregate small-cell CBSD deployment was simulated around several FSS earth stations. The CBSDs deployment parameters were chosen according to the assumptions in Section 2 of the Appendix. As indicated in Section 2 of the Appendix, the CBSDs were deployed assuming they were supporting a macro cell operating in another frequency band. Macro cells were determined to be urban or suburban according to the population density in each area as described in the Appendix. For each Suburban macro cell, one CBRS small cell was deployed. For each urban macro cell, three CBRS small cells were deployed. In each small cell, a CBSD base station with an EIRP density of was assumed. The parameters used in the simulations were not based on a worst-case scenario. For instance, even though the FNPRM does not preclude macro cells in the 3.5 GHz band, macro cells were not included in the simulation. Additionally, even though the FNPRM is proposing to allow high-power fixed point-topoint links and rural terminals (with EIRP densities of and respectively), no such deployments were assumed. Finally, no user terminals were assumed in this simulation. If user terminals and higher power CBSDs were included in the simulations, this would lead to more stringent margins for protection for the FSS. RKF Engineering Solutions, LLC th Street NW, Washington DC Page 8

9 Table 1 lists the three FSS earth stations that were evaluated with characteristics per Section 1 of the Appendix. Table 1: List of FSS earth stations used in aggregate interference analysis FSS Earth Station City, State Call Sign Coordinates No. 3 2 Malibu, CA E N, W 13 Miami, FL KA N, W 20 Hagerstown, MD KA N, W The aggregate interference was calculated with CBSD deployment as described above and with no minimum distance between the CBSD and the FSS earth station. A second simulation was performed excluding CBSDs in the protection area, as determined in the single-entry interference analysis (Section 1.4). Assuming again an FSS earth station noise temperature of 100K and a CBSD bandwidth of 10MHz, resulting in an aggregate interference power limit from all CBSDs managed by the SAS to be less than Table 2 provides the calculated margins relative to the aggregate interference power limit. As can be seen from the negative margins, without additional protective measures, it is possible that a deployment will cause interference levels well above protection threshold limits. With CBSD deployment included within the protection area, the aggregate interference limit is exceeded by about db, and with the CBSD deployment excluded from the protection area, the limit is still exceeded by about db. In other words, even with CBSDs excluded from the protection area, significant constraints on CBSD operations will be needed to meet the aggregate interference criterion to the FSS earth stations. Table 2: Example calculated Margin relative to aggregate interference power limit (-121.6dBm/10 MHz) at FSS earth stations from a deployment of CBSDs (db) FSS Earth Station Location CBSD Deployment included within the protection area CBSD Deployment excluded within the protection area Malibu, CA Miami, FL Hagerstown, MD Based on the deployment scenarios described above, Table 3 and Figures 9 through 11 show the minimum number of CBSDs that would have to shut down to meet the example calculated aggregate interference power of dbm/ MHz at each FSS earth station (assuming the same FSS earth station and CBSD characteristics as above). These plots were provided to allow investigation into how 3 The Earth Station No. corresponds to the list in the FNRPM, Appendix 2, table of FSS earth stations in MHz. RKF Engineering Solutions, LLC th Street NW, Washington DC Page 9

10 difficult it will be for the SAS to identify and mitigate the interference. The figures show that for the selected FSS earth stations, a large population of CBSD terminals (on the order of 1000s or 10,000s) would have to be shut off or otherwise have their powers significantly reduced in order to meet the aggregate limit. Furthermore, as indicated, this is true even if there is no CBSD deployment in the protection area. For reference, the CBSDs were deployed up to from each FSS earth station as their contribution was assumed to be negligible beyond that. This resulted in the total number of CBSDs indicated in parentheses in Table 3. Table 3: The minimum number of CBSDs that have to be shut-down (out of the total population of CBSDs) to meet the longterm I/N criterion at each FSS earth station. FSS Earth Station CBSDs included within protection area CBSDs excluded within protection area Malibu, CA 24,906 (97,123) 20,776 (92,993) Miami, FL 11,229 (69,000) 6,180 (63,951) Hagerstown, MD 9,637 (229,835) 6,751 (226,949) Figure 9: Distribution of received power from deployed CBSDs at Malibu, CA FSS earth station showing the minimum number of CBSDs that would have to be shut-down to meet the long-term I/N criterion. RKF Engineering Solutions, LLC th Street NW, Washington DC Page 10

11 Figure 10: Distribution of received power from deployed CBSDs at Miami, FL FSS earth station showing the minimum number of CBSDs that would have to be shut-down to meet the long-term I/N criterion. Figure 11: Distribution of received power from deployed CBSDs at Hagerstown, MD FSS earth station showing the minimum number of CBSDs that would have to be shut-down to meet the long-term I/N criterion. 2. Protection of adjacent band FSS Earth Stations 2.1. Interference Protection Criterion In line with Recommendation ITU-R S.1432, the following criterion is identified for use when assessing the out-of-band interference into FSS receive earth stations: RKF Engineering Solutions, LLC th Street NW, Washington DC Page 11

12 The aggregate interference from all other sources of interference is considered for of the time where is the clear-sky satellite system noise as described in Recommendation ITU-R S To account for other interference sources, of the permissible interference to the FSS earth station receiver was allocated to CBSD systems. This results in a reduction of the protection criterion by 3 db: Therefore, for an, assuming an FSS earth station noise temperature of 100K and a CBSD bandwidth of, the calculated aggregate interference power limit from all CBSDs managed by the SAS would have to be less than at the FSS earth station Out-of-Band Emission Limit The FNPRM mentions three OOBE limits, and ) 4 and asks for comment on which is appropriate. Using the FSS earth station characteristics described in Section 1 of the Appendix, Figure 12 shows the required line-of-sight (LOS) separation distance between a CBSD and an FSS earth station as a function of OOBE limit and the FSS earth station off-axis angle, such that the out-of-band interference criterion ( ) is not exceeded. It is clear from the figure that significant separation distances will be needed to control aggregate interference with an OOBE limit of, while the required separation distances with a tighter OOBE limit of are between and depending on the FSS earth station off-axis angle. 10 Line-of-Sight Separation Distance (Km) 1-13 dbm/mhz -40 dbm/mhz -50 dbm/mhz Earth Station Off-axis Angle Figure 12: Required LOS separation distance between a CBSD and an FSS earth station as a function of OOBE limit and the FSS earth station off-axis angle (degrees), such that the out-of-band interference criterion is not exceeded. 4 See FNPRM at 81 & 83. RKF Engineering Solutions, LLC th Street NW, Washington DC Page 12

13 3. Validation of the SAS A multi-step process will be needed to confirm whether an SAS is capable of ensuring that interference protection levels are not exceeded. a. Initially, it is anticipated that an SAS will conduct simulations of the proposed CBSD deployments, model the associated multipath transmission loss of each of the CBSD users in the cells, and compute the aggregate interference at each of the FSS earth stations. It will be necessary for an SAS to have data regarding the CBSD deployment characteristics at each of the FSS earth stations, as well as the FSS earth station receivers antenna patterns over the complete hemisphere and their effective noise temperature in order to perform this modeling. Additionally, it will be necessary for an SAS to model the multipath transmission loss over the terrain for each of the CBSD terminals (both base stations and user terminals, which may be mobile). If the model shows that the shared performance is acceptable, the SAS would then authorize the CBSD deployment. The feasibility of performing all of this modeling of a highly uncertain reality must be proven prior to moving forward with CBSD deployment. b. If deployment of the CBSD system is ultimately approved, additional challenges emerge in the deployment phase. CBSD users must first be validated and regulated in a way that prevents abuse of protection levels and is capable of being modeled by an SAS. A deployed SAS will need to test actual interference at one or more reference FSS earth stations in order to validate the predicted levels of interference. This validation is necessary because of the limited accuracy of the multipath transmission loss models. c. It will be very difficult for an SAS to measure aggregate interference levels. A key concern regarding SAS monitoring is the extreme difficulty in measuring aggregate interference levels, as described in points a-d of section 1.2. It will be especially difficult to identify the terminals causing the protection level to be exceeded or significantly contributing to the interference exceedance, unless the interference is substantially above allowed limits d. For long-term interference, a possible reference measurement approach would require integration of a pseudo random noise signal, transmitted from each CBSD. This would be required to pull the interference above the FSS receiver noise. Before a CBSD is allowed to transmit traffic, it should first transmit a pseudo random code that could be received at the FSS earth station. The signal would probably have to be received for at least several days to characterize the long-term propagation statistics. This would help to calibrate the SAS propagation modeling for long-term statistics. Note that short-term statistics would still have to be extrapolated from the limited number of measurements taken. Furthermore, for this to work, someone would have to develop a box that could receive all the CBSD pseudo random codes at the reference FSS earth station(s). RKF Engineering Solutions, LLC th Street NW, Washington DC Page 13

14 e. The measurements described above will not verify compliance with the aggregate interference criterion or help identify the CBSDs causing the most significant interference over time. For this, there needs to be regular measurements. Since an FSS earth stations traffic and multipath propagation conditions are constantly changing, measurements of the CBSD interference would need to be taken at regular intervals to verify the continued compliance with the aggregate interference criterion. This could mean turning off CBSD traffic at regular intervals and transmitting pseudo codes that can be received at the FSS earth station. f. An additional validation method might be performed while the CBSD carried traffic. Each CBSD user would dedicate one or more resource blocks (180 KHz per resource block) to continuously transmit a pseudo-random code identifying the user during frame transmission. The number of resource blocks needed would have to be analyzed and have to be sufficient to pull the interference above the satellite receive signal and noise (at the FSS earth station) during a single CBSD frame. Results could then be averaged continuously to determine the variation of the multipath propagation. The validation receiver at the FSS earth station would have to be able to pick out the individual codes from all the CBSDs so that aggregate interference could be calculated and the CBSDs causing the most interference identified. RKF Engineering Solutions, LLC th Street NW, Washington DC Page 14

15 Assumptions Used in the Technical Annex: Appendix 1- FSS System Parameters and Criteria Table 1 contains typical downlink FSS parameters for the 3.5 GHz band that were used in the study. Table 4 Downlink FSS parameters Parameter Value Range of operating frequencies Elevation angle Antenna off-axis reference pattern Antenna diameter Antenna height above ground Receiving system noise temperature MHz (in-band) (out-of-band) 5 (aggregate interference simulation) See single-entry section for simulation assumption Recommendation ITU-R S m 3.0 m 100 K 5 is considered as the minimum operational elevation angle. 2- Deployment-related Parameters for CBSD Table 2 contains CBSD parameters that were used in the study. Table 5 CBSD Parameters Deployment density (small cell) Cell radius (macro-cell) Outdoor small cell suburban 1 per suburban macro cell 0.6 km per suburban macro cell Outdoor small cell urban 3 per urban macro cell 0.3 km per urban macro cell Antenna height (m) 6 Antenna pattern Base station transmit EIRP (dbm) Omni 30 (aggregate and single-entry simulations) 53 (single-entry simulation) Base station bandwidth (MHz) 10 RKF Engineering Solutions, LLC th Street NW, Washington DC Page 15

16 To make the CBSD deployment as realistic as possible, cell deployment related parameters (for bands between 3 and 6 GHz) from Table 4 of Report ITU-R M.2292 Characteristics of terrestrial IMT-Advanced systems for frequency sharing/interference analyses were used, as indicated in the first row (Cell radius/deployment density) above. To determine if a cell is Urban or Suburban, the population density database from the Census Bureau from 2000 was used. Urban was estimated to have 250 or more people per km 2. Suburban was estimated to have 50 to 249 people per km 2. The CBSD base station EIRP levels were based on the FNPRM proposal of for non-rural CBSDs ( 74) and for Fixed point-to-point stations ( 75), per bandwidth. 3- Propagation Model and Terrain The propagation model defined in Recommendation ITU-R P.452 was used. Terrain information was taken into account by using the Shuttle Radar Topography Mission (SRTM) database. The SRTM includes in addition to terrain information, building or vegetation heights. The SRTM is a surface database taken by radar measurements from a Space Shuttle mission and contains measurements of where the radar waves are reflected off the surface of the earth. RKF Engineering Solutions, LLC th Street NW, Washington DC Page 16