VIEWS: 3 PAGES: 19 POSTED ON: 8/16/2012
INTERNATIONAL TELECOMMUNICATION UNION RADIOCOMMUNICATION Document 8F/908-E STUDY GROUPS 27 March 2003 English only 10TH MEETING OF WORKING PARTY 8F PORTO SEGURO, 26 MARCH – 3 APRIL 2003 Source: Documents 6S/317, 6S/320, 6S/337 SPECTRUM Working Party 6S LIAISON STATEMENT TO WP 8F FROM WP 6S PDNR “PROBABILISTIC METHODOLOGY FOR THE ASSESSMENT OF INTERFERENCE FROM NON-GSO BSS (SOUND) INTO TERRESTRIAL IMT- 2000 IN THE BAND AROUND 2.6 GHz” The attached PDNR “Probabilistic methodology for the assessment of interference from non-GSO BSS (sound) into terrestrial IMT-2000 in the band around 2.6 GHz” has been reviewed by WP 6S. Within the discussions of WP 6S and also JRG 6S-8F (See Document 6S/337), several areas of concern were identified as well as areas where further work is necessary on this PDNR. These concerns and areas for further work are summarized in the attached PDNR. WP 8F is invited to review the contents of this PDNR and forward any comments to WP 6S in an effort to progress the work on this methodology. WP 6S notes that this PDNR should be developed jointly with WP 8F. Attachment 1: Document 6S/TEMP/176 (edited) C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 -2- 8F/908-E Attachment Source: Document 6S/TEMP/176 PRELIMINARY DRAFT NEW RECOMMENDATION ITU-R BO. Probabilistic methodology for the assessment of interference from non-GSO BSS (sound) systems into terrestrial IMT-2000 systems in the band around 2.6 GHz The PDNR in Attachment 2 contains a methodology for the assessment of interference from non- GSO BSS (sound) systems into terrestrial IMT-2000 systems in the frequency band around 2.6 GHz. It is noted that WP 6S and WP 8F have jointly developed a PDNR that contains another methodology that addresses this same topic. The methodology in the attached PDNR takes a more probabilistic approach to the assessment of interference than that methodology. During the WP 6S meeting, several concerns and areas for further work were identified related to the attached PDNR and these are included as Attachment 1. Attachment 1 Comments on PDNR “Probabilistic methodology for the assessment of interference from non-GSO BSS (sound) systems into terrestrial IMT-2000 systems in the band around 2.6 GHz” The statistical approaches proposed in the PDNR in Attachment 2 are basically averages of “interference levels” received at IMT-2000 stations located evenly on Earth, and suffering aggregate interference from an assumed constellation of non-GSO BSS (sound) satellites/systems. The “statistics of interference” calculated at an IMT-2000 station are proposed to be evaluated in terms of Isat/Nth statistics, or in terms of “availability loss” statistics using the methodology contained in 6S/238, attachment 11: “working document towards a PDNR on the methodology for determining the availability loss of terrestrial mobile base stations due to satellite interference” (see 6S-8F/22, sections 3.1.2 and 3.1.3). 1 Isat/Nth statistics The Isat/Nth statistics using “Approach A” could be labelled “risk assessment method”: it is a geographical averaging of Isat/Nth levels obtained with 6S-8F/22’ Approach A in section 3.1.1 on each IMT-2000 station, hence including averaging over the IMT-2000 station orientations, latitudes and longitudes. Hence the records obtained with this approach could be interpreted as the “risk” an IMT-2000 station may receive a given Isat/Nth level… Several input documents to WP 6S and to the JRG 6S-8F show that the BSS (sound) systems’ need for high power levels on their service area will necessitate the use high performance directive antennas on board BSS (sound) space-stations. These documents show then that the areas on Earth receiving the most important levels of interference and seen with high elevation angles by the satellites will be distributed on a “belt” of several tens or thousands of kilometres around the BSS (sound) service area (the belt’s wideness depending on the antenna roll-off performance). This means that the “risk” would certainly not be C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 -3- 8F/908-E equally distributed over the Earth, and the averaging technique as proposed is therefore likely to hidden evident very high risks of interference in potential victim countries whishing to implement IMT-2000 stations and situated in the vicinity of a potential BSS (sound) service area. Therefore it is anticipated that the results produced by this averaging technique would provide very poor information or even misleading information that could be used in the context of specific co- frequency interference situations assessment involving BSS (sound) and IMT-2000 systems. 2 Availability loss using Annex 2 of PDNR The “availability loss” methodology contained in Annex 2 of the PDNR aims to be applied in any case involving terrestrial cellular systems in the Mobile Service and satellite interferers, special considerations being made regarding terrestrial cellular CDMA systems. Working Party 6S has received several liaisons statements making observations on this document, at its September 02 meeting in Geneva (see 6S/253, 6S/256). The observations mainly address concerns about the limitation of this document’s considerations to the outage of one single link, and its subsequent lack of accuracy to meaningfully address the “availability loss” (or capacity loss) of terrestrial cellular mobile systems, and in particular in CDMA IMT-2000 systems. In summary, the outage of one single link is expressed as a signal availability, which is calculated in absence, and in presence, of satellite external interference. The loss of availability of this link is then calculated from the difference between the two signal availability results: this approach is akin to an approach in terms of “C/N+I” degradation for the link into consideration. The methodology then proposes averaging methods to assess a “loss of availability” on a terrestrial cellular deployed network. Basically, the document considers that a cell in a cellular network is a set of several single links, each of which is living independently in a given environment (characterised by a level of noise rise, determined through parametrical equations taking into consideration a “cell load” factor and an other factor standing for “perfect, or imperfect, power control”). This assumption is the main reason why the proposed methodology does not reflect the specificities of a terrestrial CDMA cellular systems, and in particular the necessary power control associated with this technology. A cell and its load or number of users, is not to be considered as a set of independent links, but rather as a set of links which are dependant to each other and monitored through power control algorithms to optimise the overall CDMA system performance, which is compound of three interlinked parameters (cell capacity or cell load, quality of service offered to the users, and the coverage of the cells). Document 6S/238 attachment 11 concludes in particular that the “loss of availability” due to satellite interference over a cell is independent of its load, whereas it has been shown in other studies that the effect of satellite interference over heavily loaded cells (which are capacity limited) is lower in terms of capacity loss than over lightly loaded cells (which are coverage limited), all other parameters being fixed (cell range and quality of service). This shows that this “signal availability loss”-derived methodology is unproper to assess the global impact of satellite interference into cellular systems, that is : on an overall network of cells, each cell containing a number of users whose individual quality of service, distance from the base stations, and transmit power are monitored globally by the base stations, with dedicated power control algorithms. In addition to that, documents 6S/253 and 6S/256 (section 4) also underlines that the working document in 6S/238 Attachment 11 conveys some erroneous ideas about terrestrial CDMA cellular systems. Furthermore, it is unclear in the proposed methodology how the results in terms of “availability loss” are meant to be presented, but would the intent be to further average again the results over location and orientation of the IMT-2000 base stations, then the comments concerning the “risk availability approach” would apply in addition to the ones presented here. C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 -4- 8F/908-E Attachment 2 PRELIMINARY DRAFT NEW RECOMMENDATION ITU-R BO. Probabilistic methodology for the assessment of interference from non-GSO BSS (sound) systems into terrestrial IMT-2000 systems in the band around 2.6 GHz1 (QUESTIONS ITU-R 204-1/10 AND 229/8) The ITU Radiocommunication Assembly, considering a) that a methodology is required to assess the aggregate interference to IMT-2000 from non- GSO BSS (sound) for the technical study as invited in Resolution 539 (WRC-2000), recognizing a) that as far as space services are concerned, the band 2 535-2 655 MHz is allocated on a primary basis to BSS (sound) and complementary terrestrial broadcasting service under RR No. 5.418; b) that BSS (sound) systems under RR No. 5.418 are subject to RR Resolution 528 (WARC- 92) and that these systems are not subject to the pfd limits in RR Table 21-4; c) that WRC-2000 adopted Resolution 539 which, inter alia, contains provisional pfd threshold levels for non-GSO BSS (sound) systems under RR No. 5.418 and invited ITU-R to conduct the necessary technical and regulatory studies in time for WRC-03 relating to frequency sharing in the band 2535-2655 MHz with a view to avoiding placing undue constraints on either service (see Res. 539); d) that the band 2 500-2 690 MHz band is also identified in RR No. 5.384A for use by Administrations wishing to implement IMT-2000 in accordance to Resolution 223 (WRC-2000); e) that RR No. 5.384A states this identification does not preclude the use of these bands by any application of the service to which they are allocated and does not establish priority in the Radio Regulations; further considering that the preliminary draft new recommendation “A methodology to assess interference from BSS (sound) into terrestrial IMT-2000 systems intending to use the band 2 630-2 655 MHz” can also be used to assess interference from, and possible impact of, BSS (sound) on terrestrial IMT- 2000 systems intending to use the band 2 630 - 2 655 MHz in the context of co-frequency operation through the development of pfd masks, recommends 1 that the methodology described in Annex 1 of this Recommendation could be used for the assessment of interference from non-GSO BSS (sound) systems into terrestrial IMT-2000 systems. ____________________ 1 It should be noted that, in accordance with Resolution 528 (WARC-92), broadcasting-satellite service (sound) systems may only be introduced within the upper 25 MHz of the band 2 535-2 655 MHz. C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 -5- 8F/908-E Annex 1 Probabilistic methodology for assessing the potential interference from non- GSO BSS (sound) systems into terrestrial IMT-2000 1 Input data and scenarios 1.1 Characteristics of the systems A given scenario will consist of non-GSO BSS (sound) systems interfering into IMT-2000 systems (base and/or mobile stations). 1.1.1 IMT-2000 stations2 • Receiver characteristics: – thermal noise level (dBW/MHz); – noise factor (dB); • Antenna characteristics: – maximum gain (dBi); – polarization; – feed loss; – 3 dB beamwidth3; – vertical and azimuthal antenna radiation patterns over a range of elevation angles2; – downtilt of the antenna2; – site sectorization2; • Location of the receivers (for example, an area bounded by latitude(s) and longitude(s) data). 1.1.2 Non-GSO BSS (sound) systems Although the orbital simulation approach (including time variation of the satellite) requires more complex simulation tools, this approach will produce more accurate results and is recommended for the accurate pfd assessment. The following orbital elements and characteristics are required. • Orbital elements (Inclination angle, altitude of apogee, altitude of perigee, argument of perigee, longitude of the ascending node) • Orbital characteristics (the number of active satellites in the active arc, start of active arc, and end of active arc etc.) • The respective positions of non-GSO satellites are defined by longitude, latitude and height which can be determined from orbital elements and characteristics. Information on the polarization used by the satellite transmitters would be required to assess polarization discrimination if needed (see factor Pi in equations 1 in section 3.1.1). The following example models could be used in the analysis: ____________________ 2 Additional system specific input parameters for IMT-2000 stations would be required for usage of methods in section 3.1.3. 3 Only applies to base station receivers. C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 -6- 8F/908-E – Model 1 : A number of equally spaced non-GSO systems with synchronization: All have their “active arc” located in the northern hemisphere. The orbital parameters are identical to those given in the above table with the constellations separated synchronously. – Model 2 : A number of equally spaced non-GSO systems without synchronization: All have their “active arc” located in the northern hemisphere. The orbital parameters are identical to those given in the above table with the constellations separated non-synchronously. Model 1 is easier one to perform calculations and not far from the realistic situation because the size of the “active arcs” of the each non-GSO BSS (sound) systems are small to have the minimum elevation angles of not less than 40 deg over their service areas and the difference between the results of Model 1 and those of Model 2 is expected to be small. Model 2 is expected to bring the most realistic results. 2 Results Presentation of the results The results are expressed as statistics of Isat/Nth received at IMT-2000 stations. The results are also expressed as statistics of the reduction in the signal availability (see Annex 2). 3 Method of calculation and production of the results 3.1 Assessment of the interference into an IMT-2000 base station In the analysis for the pfd assessment, IMT-2000 base stations at latitudes varying in 10 increments from XoN to YoN are evaluated (X and Y can be determined from the visible surface area of the Earth from the non-GSO BSS (sound) satellite). At each latitude, it is assumed that there is an IMT-2000 base station located at every 1 of longitude. Each base station antenna azimuth is also varied from 0 to 360 in 10 steps. 3.1.1 Aggregation of the interference from multiple satellites into a given IMT-2000 base station receiver The calculation steps for the aggregation of the satellites interference are summarized below: – considering a set of non-GSO satellites distributed on the 360° longitude around Earth; – considering given pfd threshold values applicable to non-GSO satellite; – considering an IMT-2000 base station with sectoral antenna, characterized by its latitude, longitude and orientation; – the azimuth and elevation, hence the gain of the antenna depends on the location of the base station (latitude, longitude) with regard to the satellite under consideration; – calculation of the aggregate interference at the receiver entrance from all co-frequency visible satellites (whose elevation is positive) and the subsequent Isat/Nth IMT-2000 base station receiver (sector) at a given latitude and longitude (lat, long), and pointing in a given direction (orientation, tilt angle) from all co-frequency visible satellites and the subsequent Isat/Nth is given by the following formula: pfd ( elevation _ ti ( t )) G ( azimuth _ oi ( t ),elevation _ ti ( t ))10 log 2 FL Pi 1 n _ sat / 10 Isat lat, long, orientation, t 10 log 10 4 i (1) Nth Nth i 1 C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 -7- 8F/908-E where: Isat Nth lat, long, orientation, t (dB) is the resulting aggregate Isat Nth from all co-frequency satellites at the IMT-2000 receiver; Isat The appropriate approach to calculate Isat/Nth should relate to the appropriate Nth criterion of Isat/Nth for the co-primary sharing. Isat/Nth is calculated at a given base station location as a result of the transmissions of multiple non-GSO BSS (sound) systems ; pfdi(elevation) (dBW/m2/MHz) is the pfd at the IMT-2000 station from visible BSS (sound) satellite i; G(azimuth_oi ,elevation_ti) (dBi)is the gain of the IMT-2000 base station antenna for the off-axis angle towards the satellite i; Elevation_ti (°) is the absolute elevation together with the tilt angle from the base station to the satellite i; it is the elevation angle (the angle of arrival of the satellite i incident wave to the IMT-2000 station, above the horizontal plane), plus the tilt angle (a downtilt angle is a negative value); Azimuth_oi (°) is determined by the longitude and latitude of the base station together with its orientation relative to the satellite i; it is the azimuth angle (North pole, IMT-2000 station, satellite i), plus the orientation of the IMT-2000 station with regard to the North direction on Earth; (m) is the wavelength; FL (dB) is the IMT-2000 receiver feeder loss; Pi (dB) is the expected averaged polarization discrimination between transmitting antenna of satellite i and the IMT-2000 base station receiving antenna; n_sat is the number of satellites ; Nth (W/MHz) is the IMT-2000 station receiver thermal noise. 3.1.2 Evaluation using Isat/Nth Statistics This approach determines the probability that an IMT-2000 base station will experience different Isat/Nth levels (see Section 3.1.1). For the analysis, simulations are performed to calculate the interference into an individual base station. The simulation takes into account each satellite that is visible to the base station and is located within its “active arc”. For a given latitude, longitude and azimuth combination for the IMT-2000 base station, the orbital positions of each satellite (in all of the constellations) is updated for each time increment. At each time increment, the point angles (azimuth and elevation) from the base station location to the active satellite in each constellation are calculated. For each satellite that is visible to the base station location, the azimuth and elevation off-pointing angles (i.e., the angles between the direction the base station is pointing and the direction to the satellite) are calculated. The gain of the base station antenna in the direction of each visible active satellite is then calculated. The satellite power-flux density is determined from the power-flux density mask that is C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 -8- 8F/908-E assumed in the analysis and the elevation angle from the base station location to the satellite. The interfering signal power received by the base station from each visible satellite is calculated. The aggregate interfering signal power received by the base station from all visible active satellites is then calculated, as is the Isat/Nth due to the interference from all visible active satellites. This process is repeated for each azimuth at a given latitude and longitude location. This process is repeated for each longitude at a given latitude in accordance with the approach described in Section 3.1.1. The maximum and average Isat/Nth levels received for all base stations at a given latitude are also calculated. The percentage of base stations where Isat/Nth exceeds at the interest level should be compared to the appropriate criterion for the co-primary sharing. Further study will be required to establish the appropriate criterion for the co-primary sharing. 3.1.3 Evaluation using signal availability After Isat/Nth level is obtained at the point of interest, the methodology in the Annex 2 should be taken for the signal availability evaluation. The reduction of signal availability with respect to the design value should be compared to the appropriate criterion for the co-primary sharing. Further study will be required to establish the appropriate criterion for the co-primary sharing. 3.2 Assessment of the interference into an IMT-2000 mobile station In the analysis for the pfd assessment, IMT-2000 mobile stations at latitudes varying in 10 increments from XoN to YoN are evaluated (X and Y can be determined from the visible surface area of the Earth from the non-GSO BSS (sound) satellite). At each latitude, it is assumed that there is an IMT-2000 mobile station located at every 1 of longitude. For the I/N evaluation, the calculation methodology in 3.1.1 and 3.1.2 could be used. The evaluation methodology for the signal availability of mobile stations should be further studied. Annex 2 Methodology for determining the availability loss of terrestrial mobile base stations due to satellite interference 1 Introduction This contribution presents a methodology for estimating the effect of potential satellite interference, Isat, from either GSO or non-GSO satellites on the performance of cellular systems. The methodology is demonstrated for the case of cellular uplinks (mobile to base station) using CDMA signalling, but has application to the downlink (base to mobile) as well as to other signalling techniques. In the analysis presented, the effect of Isat (interference power from satellites) is measured in terms of reduced cellular system signal availability as a function of the ratio of satellite interference to system thermal noise Isat/Nth. It is demonstrated that the reduction in signal availability due to Isat is the most appropriate measure of loss in cellular system. The analysis presented shows that the effect on signal availability is small for Isat/Nth < 0 and that for future cellular systems economic measures are available to offset the potential effect of Isat. The low impact of Isat is due, in large part, to the margin required in the design of the cellular system to address the statistical nature of the propagation medium. The typical design goal of 95% C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 -9- 8F/908-E availability at the edge of coverage requires that the cellular system be designed to accommodate statistical variations of greater than 30 dB. This is very large relative to a possible increase of 1 dB in overall system noise plus interference (N+I). Precedence exists for using an acceptable reduction in system performance as the basis for determining acceptable levels of Isat with specific examples being: 1) In establishing pfd limits in the band 3 700-4 200 MHz for satellite systems, the effect of potential interference into the terrestrial fixed point to point microwave stations was evaluated on the basis of allowing a certain level of interference for a given percentage of time. 2) In considering the introduction of the Skybridge system in sharing with BSS systems in the 12.2 to 12.7 GHz band, the allowable levels of interference from Skybridge into existing BSS systems was based on an increase in BSS outage time as contained in Recommendation ITU-R BO.1444. These two cases dealt with the accommodation of new technologies within the ITU frequency plan. Basing allowable levels of interference on a performance basis is appropriate for the analysis of sharing between co-primary services. 2 Basic approach First, the performance of a cell at edge of coverage is considered. A basic design parameter for the cell is the signal availability at the edge of coverage, Ase, which is the probability that a typical user can achieve service with the minimum required Eb/No. A typical design value for Ase is 95%. In order to analyse cellular system performance it is helpful determine the required mean signal at the base station receiver for a user transmitting at full power from the edge of coverage. This can be determined from Ase, the required Eb/No, the noise and interference levels at the base station receiver and the statistical variation of the desired signal. The availability at edge of coverage when Isat is present and not present can then be determined. Figure 1 illustrates the process for determining the impact of Isat at the edge of coverage as the difference in system availability with and without Isat. The reduction in signal availability is illustrated as the darkened portion of the signal statistical distribution at the edge of coverage. The propagation model then provides the means for extending this basic approach to the analysis of performance at any distance from the base station. The cellular system performance measures used in evaluating the impact of Isat are: Ase = Signal availability at the edge of the cell in %. Asa = Signal availability averaged over the entire cell in %. 2.1 Signal, noise and interference at edge of coverage Table 1 presents computations of the basic signal, noise and interference levels for a typical cellular system. The computation of N+I is shown on lines 1-24 for the cases of "without Isat" and "with Isat". The impact of Isat on overall N+I is given on line 25. The determination of mean signal at edge of coverage, based on a specified Ase, is given in lines 26-31. In a cellular, spread spectrum system, interference is contributed from other users. The internally generated interference or interference resulting from other users, relative to the sum of thermal noise plus Isat at the base station receiver as given on line 20 is given by: Ii = 10 log (L/(1L)) (1) C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 10 - 8F/908-E Where L is the normalized loading relative to the theoretical maximum, or pole capacity. In selecting the appropriate value for L, allowance is made for interference from users in adjacent cells. The cellular system will be operating with power control so that all user signals arrive at the base station with the same signal level. The user transmitter levels are set so that the received signals are the minimum level to maintain the required Eb/No for all users. However, to represent realistic conditions in which the other user signals can exceed the desired value since power control is not perfect, an allowance for power control error of 1 dB is made. The noise from other users is modelled as being 1 dB higher than that for the case of all user signal arriving at the base station receiver with equal level. The signal variation, in location, from any distance to the base station is modelled as a lognormal distribution with 8-10 dB standard deviation [See References 1 and 2]. The mean required signal level to provide the desired Ase, can then be determined as given in line 31. The first column of Table 1 presents the basic signal, noise and interference parameters for the case where Isat is not present with the second column presenting the values for the parameters, which change due to the presence of Isat. Isat is specified in terms of the system thermal noise level as shown in row 13. The net effect of the presence of Isat relative to the overall N+I, is shown on line 25. 2.2 Signal, noise and interference at any point in the cell The noise and interference levels at the base station receiver are independent of user location within the cell. The internal interference, Ii (Equation 1) is only dependent upon the number of users within the cell as specified by the loading ratio, L. The propagation model is illustrated in Figure 1 with the mean signal received at the base station receiver, for points not close to the base station, Pr, given by: Pr k / D b (2) Where: D= distance of the user from the base station, b= power law coefficient. k= a constant including mobile power and antenna heights and gain Pr= mean signal level at the base station receiver For the Hata and CCIR propagation models, 3.52 <b<3.84. The signal variation in location, for any distance to the base station is modelled as a lognormal distribution with an 8 to 10 dB standard deviation [Reference 1 page 247, Section 2.2.3]. In the sample computations presented below, the propagation model is based on a power law coefficient of 3.84 and a standard deviation of 8 dB. From the basic propagation model given in Equation 2, the mean signal at any distance from the base station can be determined from: Pr Pe( Rd )b (3) Where: Pe = Mean signal level at edge of cell coverage Rd = (De/Ds) is the distance ratio De = distance to the edge of coverage C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 11 - 8F/908-E Ds = distance to the point of interest 3 Signal availability with uniform traffic density For convenience in computing the impact of Isat, the cell is divided into rings with the impact of Isat computed for each ring. Figure 2 presents the procedure for dividing the cell area into a number of rings. The width of each ring is determined such that little variation in system performance occurs throughout the ring. Equation 3 provides the basic method for performing the analysis on a normalized basis independently of the actual cell size. The radii of each ring need only be specified relative to De. The performance within each ring is represented as the performance at the average radius of the ring. The mean signal levels at the center of each ring can be determined from Equation 3 since the signal level at edge of coverage has been determined. The statistical distribution of the received signal at a point removed from the edge of coverage is illustrated in Figure 1. The widths of the outer rings are smaller than the width of the inner rings since the impact of Isat on the performance in the outer rings is suggesting a higher degree of resolution. The performances within each ring are weighted by the area of the ring relative to the total cell area and summed to determine the impact on signal availability due to the presence of Isat. Weighting signal availability by the area of the ring is consistent with an assumption of uniform traffic, or user, density. 3.1 CDMA system operation and user available power margin In a CDMA spread spectrum cellular system, power control of the mobile transmitters is used to equalize the signal received at the base station receiver for all users, with the mobile or user transmitter power being set to the minimum required to achieve the desired Eb/No at the base station receiver. The user available margin, shown on Figure 1, is the difference between the signal at the base station for maximum user transmitter power available to a mobile and the signal at the base station necessary to achieve the required Eb/No at the base station receiver. If the user has available power margin greater than the noise increase due to Isat, the transmitted power is increased to overcome the effect of Isat. In the analysis to follow the impact of Isat is measured in terms of the reduction in signal availability at the edge of the cell, Ase, and the reduction in signal availability averaged over the entire cell, Asa. Both parameters are determined based on the assumption that the mobile transmitter is free to transmit at full power if required. 3.2 Computation of Signal Availability Table 2 presents sample computations for the % of users not served within each ring with each column corresponding to a ring. Lines 34-42 and 43-51 show computations of availability loss with and without Isat, respectively. The difference is the availability loss due to Isat as given on line 52. The maximum radius of each ring is presented in both lines 34 and 43 on a normalized basis relative to the cell radius. The percent of the total cell area within each ring is given on lines 36 and 45. The mean signal level at edge of coverage is given in line 31 of Table 1 and Equation 2 provide the basis for computing the mean signal at the average radius of each ring as given in rows 38 and 47 of Table 2. Basically, the user not served does not have sufficient power to overcome N+I and an unfavorable propagation condition. As the radius of the rings decrease the users have additional power margin and the probability of not being served is reduced. The probability (in %) within the ring of not being served is given in lines 40 and 49. The probability of no service within the ring is weighted by the normalized area to give the percentage of the total cell subscribers not being served as shown in lines 41 and 50. The percentage of unserved users within each ring can then be summed to obtain the average C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 12 - 8F/908-E unavailability within the cell as given in lines 42 and 51. The difference is the overall reduction in signal availability due to the presence of Isat, as shown on line 52. Figure 3 presents the reduction in Asa due to Isat as a function of Isat/Nth and Ase. Figure 4 presents results for different propagation model parameters illustrating that availability loss due to Isat is not sensitive to power law coefficient and signal standard deviation over the range of accepted value. The reduction in signal availability is independent of the design value for cell loading when the cell is operating at design cell loading. When the cell is operating with less loading than the design value, the reduction in availability due to Isat is less than that when operating at design loading. Accordingly, the reduction in availability presented in Figures 3 and 4 is the upper bound on availability loss for a given Ase and Isat/Nth. For a given Ase, the loss in availability is also found to be independent of the system thermal noise level and the required Eb/No. Both, however, are factors in determining Ase. As can be seen from Figures 3 and 4, the reductions in overall signal availability is small, especially for systems designed with high Ase. Typically, Ase 95% in cellular system design. Table 3 presents a summary of the reduction in both Asa and Ase for Ase = 95% and Isat/Nth = 6 and 10 dB. Figure 5 shows the variation of availability with loading expressed relative to the design loading. The reduction in loading leads to an increase in both Ase and Asa. 3.3 Considerations in cellular system design Although the impact of Isat on cellular system performance is small, it can be taken into account in the design of new systems with a corresponding increase in the design value of Ase. Methods for achieving this include reduction in base station receiver noise, increase in mobile transmitter power and with slight increases in base station antenna gain and height. The overall effect on receiver N + I for Isat/Nth = 6 dB is 0.97 dB. From the standard propagation models summarized in Reference 1, signal level increases approximately in proportion to the square of the receiving antenna height so that n increase of 5% would overcome a 1 dB increase in N+I. Other means include reduction in cell size and in the loading for each cell in accordance with Figure 5 which presented the variation of availability with system loading. It would be necessary to reduce the loading by approximately 25% in order to maintain Ase at 95%. Similarly a reduction in cell area of approximately 8% would be required to offset a 1 dB increase in N+I. It is concluded that compensating for Isat with a reduction in either cell loading or cell size should not be considered in estimating the economic impact of Isat as far more economic measures are available. 4 Effect of non-uniform or random traffic density on availability loss The results presented in Figures 3 and 4 are based on the assumption of geographically uniform user distribution over the cell coverage area. Additional analysis has been performed to investigate the effect that random user distributions throughout the cell would have on loss of signal availability in order to determine the sufficiency of the uniform user distribution assumption. Availability loss computations were performed for 50 and 20 users respectively with 50 trials being performed for each case. For each trial, random user locations were generated for each user. The ratio of the number of users in each ring to the total number of users was used in place of the normalized area on line numbers 36 and 45 in Table 2, and used to weigh the "Unavailability" in lines 40 and 49 to determine the percent of the total cell subscribers not being served within the ring. Table 4 presents the variation in availability loss results for the random user distribution trials. As seen in the results, the standard deviation and statistical range of signal availability loss for different percentiles is quite small. The variance of the results is slightly greater for the case of 20 randomly C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 13 - 8F/908-E positioned users. It is concluded that additional investigation of availability loss can be made with the "uniform user distribution model". 5 Summary and conclusions A method has been presented for determining the impact of satellite interference into cellular systems. The method is based on the use of edge of coverage signal availability, which allows for analysis to be performed on a normalized cell with results independent of actual cell dimensions. This method can be used in sharing studies since it presents a direct measure of the reduction in signal availability within a cell as a result of potential satellite interference. The results presented are based on uniform geographic user, or load, distribution over the cell service area under consideration. It has been demonstrated through a combination of statistical analysis and simulation techniques that random load distribution does not alter the results to any significant degree. Further the results are not sensitive to the power law coefficient and standard deviation of the lognormal distribution used to characterize the propagation model. For a given signal availability at edge of coverage, the loss in availability due to satellite interference is found to be independent of cellular loading, required Eb/No and system thermal noise level when the system is operating at design loading. For system loading less than the design value, the Availability loss due to Isat is always less than for the case when operating at design loading. The overall availability loss in the cell for Isat/Nth = 6 dB is less than 0.5 %, with the overall increase in N+I being 0.97 dB. The impact on cellular system design is small with several economic design measures available to compensate for Isat/Nth = 6 dB. It is concluded that compensating for Isat with a reduction in either cell loading or cell size should not be considered in estimating the economic impact of Isat on cellular operation, as far more economic measures for compensating for Isat are available. Methods for achieving this include reduction in base station receiver noise, increase in mobile transmitter power and with slight increases in base station antenna gain and height. The overall effect on receiver N + I for Isat/Nth = 6 dB is 0.97 dB. REFERENCES [1] CDMA Systems Engineering Handbook, Lee and Miller, Artech House, 1998. [2] Recommendation ITU-R M.1225, Guidelines for Evaluation of Radio Transmission Technologies for IMT-2000, (Question ITU-R 39/8) and 2)). C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 14 - 8F/908-E TABLE 1 System Signal, Thermal Noise and Interference System Thermal Noise at Receiver Input No Isat w/Isat 1 Rec Noise Figure dB 4 2 Feeder Loss dB 2 3 Ant Temp deg 290 K 4 Portion of Energy not absorbed by 0.63 feeder 5 Portion of Energy Absorbed by Feeder 0.37 6 Room Temp K 290.00 7 Rec temp K 438.45 8 Feeder Loss Temp K 107.02 9 Ant Temp @ Rec K 182.98 10 System Temp, Nth K 728.45 11 System Noise Figure dB 5.46 Noise + Interference 12 Sys Thermal Noise /MHz dBW 139.98 139.98 13 Isat/Nth dB 6.00 14 Isat dBW 145.98 15 Isat+Nth dBW 139.98 139.00 16 Cell Loading in % % 50.00 50.00 17 Voice/Data Average Activity % 75.00 75.00 18 Increase in Load from other Cells % 33.00 33.00 19 Overall Loading % 49.88 49.88 20 Noise due to cell loading re:(Isat+Nth) dB 0.02 0.02 21 Power Control Error dB 1.00 1.00 22 Noise due to cell loading w/o PCE dBW 140.00 139.02 23 Noise due to cell loading with PCE dBW 139.00 138.02 24 (N+I) total dB 136.45 135.48 25 Impact of Isat to tot Noise dB 0.97 Required Signal 26 Eb/No Required dB 5.00 5.00 27 Required Signal dBW 131.45 130.48 28 Signal Availability at Edge of Coverage % 95 95 29 Signal Standard Deviation dB 8 30 Standard Deviations Required 1.64 31 Mean Signal at edge of coverage dBW 118.29 C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 15 - 8F/908-E TABLE 2 Reduction in Signal Availability % Not Served With Isat 34 Max Ring Range/Max Cell Range 1.00 0.95 0.90 0.85 0.80 0.70 0.50 0.30 35 Center of Ring/Max Cell Range 0.98 0.93 0.88 0.83 0.75 0.60 0.40 0.15 36 % Total Area in Ring % 9.75 9.25 8.75 8.25 15.00 24.00 16.00 9.00 37 Acum Normalized Area(Users) 9.75 19.00 27.75 36.00 51.00 75.00 91.00 100.00 38 Mean signal in range dBW 117.87 116.99 116.06 115.08 113.49 109.77 103.01 86.65 39 Std Dev above req Signal 1.58 2.31 2.43 2.55 2.75 3.21 4.06 6.10 40 Unavailability % 5.752 4.592 3.581 2.716 1.688 0.483 0.030 0.000 41 Percent of Tot Users not served % 0.561 0.425 0.313 0.224 0.253 0.116 0.005 0.000 42 Percent Reduction in Availability % 1.897 % Not Served Without Isat 43 Max Ring Range/Max Cell Range 1.00 0.95 0.90 0.85 0.80 0.70 0.50 0.30 44 Center of Ring/Max Cell Range 0.98 0.93 0.88 0.83 0.75 0.60 0.40 0.15 45 % Total Area in Ring % 9.75 9.25 8.75 8.25 15.00 24.00 16.00 9.00 46 Acum Normalized Area(Users) 9.75 19.00 27.75 36.00 51.00 75.00 91.00 100.00 47 Mean signal in range dBW 117.87 116.99 116.06 115.08 113.49 109.77 103.01 86.65 48 Std Dev above req Signal 1.698 1.807 1.923 2.046 2.245 2.710 3.555 5.600 49 Unavailability % 4.479 3.535 2.723 2.038 1.240 0.337 0.019 0.000 50 Percent of Tot Users not served % 0.437 0.327 0.238 0.168 0.186 0.081 0.003 0.000 51 Total Users in Cell not served % 1.440 52 Reduction in Capacity % 0.457 TABLE 3 Change in Signal Availability due to Isat, Ase = 95%, Operating at Design Load Signal Availability Edge Average Signal Availability of Coverage (%) over Cell (%) Isat/N (dB) w/o Isat Reduction with Isat w/o Isat Reduction with Isat Due to Isat Due to Isat 6 95.00 1.38 93.62 98.56 0.46 98.10 10 95.00 0.56 94.44 98.56 0.18 98.38 C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 16 - 8F/908-E TABLE 4 Distribution of Isat Availability Loss with Random User Location, I/N =6 dB, Ase = 95% 50 Users, Operating at design Loading Percentile 0.05 0.1 0.3 0.5 0.7 0.9 0.95 0.99 Availability Loss(%) 0.35 0.37 0.41 0.44 0.50 0.53 0.54 0.60 Std Mean Dev 0.68 0.446 20 Users, Operating at design Loading Percentile 0.05 0.1 0.3 0.5 0.7 0.9 0.95 0.99 Availability Loss 0.31 0.33 0.40 0.42 0.49 0.54 0.56 0.62 (%) Std Mean Dev 0.88 0.443 C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 17 - 8F/908-E Mean Signal Distribution of Received Signal with Max Mobile Transmitter Power. Power law coefficient =3.8 Lognormal Distribution with 8 dB Std. Dev . Desired Signal System Noise Edge of Coverage Availability Typical Mobile and Interference Situation At the Base Station Receiver (dB) Typically 95% Available Power Required Signal Margin With Isat Required Signal Without Isat Due to Eb /No Loss of Signal Availability due to Isat Isat Interference due to other Users System Thermal Noise Distance from Base Station Edge of Coverage FIGURE 1 Effect of Isat on CDMA Performance Edge of Coverage Max Radius Ring i Min Radius Ring i Av Radius Ring i Cell Radius FIGURE 2 Dividing the Cell Into Rings of Uniform Performance C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 18 - 8F/908-E 3.6 3.2 Reduction in Signal r Availability, Asa (%) 2.8 2.4 Parameter: Signal Availability at Edge of CoverageWithout Isat 2.0 85 % 90% 95% 1.6 98% 1.2 0.8 0.4 0 - 20 - 15 - 10 -5 0 Isat /N (dB) FIGURE 3 Reduction in UpLink Availability Due to Isat, Cellular system Operating at Design Load 3.6 3.2 Reduction in Signal Availability, Asa (%) Parameter: Power Law Coefficient and Standard Deviation 2.8 b Std Dev 3.52 8 2.4 3.52 10 3.84 8 3.84 10 2.0 1.6 1.2 0.8 0.4 0 -20 -15 -10 -5 0 Isat /N (dB) FIGURE 4 Reduction in UpLink Availability Due to Isat, Cellular system Operating at Design Load, Ase = 95% C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12 - 19 - 8F/908-E 100 Signal Availability at Edge of Coverage (%) 99 98 Average Availability throughout cell 97 Availability at Edge of Coverage 96 Key: Ase = 95% 95 Isat /N= -6.0 dB Without Isat With Isat 94 93 0 0.25 0.50 0.75 1.00 Load Re: Maximum D esign Load FIGURE 5 Signal Availability at Edge of Coverage Vs. Loading Re: Design Load ______________ C:\DOCSTOC\WORKING\PDF\B2C5A9B4-C26A-4416-877C-E59BB6C5C62F.DOC 16.08.12 16.08.12