Path loss data normalization for growth management of cellular system

Description: TECHNICAL FIELD [Detailed description of the invention]

【0001】

[Industrial application field]

The present invention relates to methods of managing channel interference in mobile phone systems, and more particularly to methods of analyzing RF propagation in mobile phone systems for better deployment, structure and operation of antennas.

【0002】

[Previous technology]

Today's commercial mobile communication systems typically have multiple fixed-location base stations arranged according to a pattern, within which each station is divided between multiple sectors and has multiple antennas. , Sends and receives on many frequencies. The distribution of antenna, sector and frequency allocation and antenna selection is designed to optimize coverage. The channels used by the base station are well separated from each other in frequency so that signals on different channels from that base station do not interfere.

【0003】

However, each cell is usually arranged so that its coverage area is adjacent and overlaps the coverage area of many other cells so that the mobile unit can send and receive telephone communications over a wide area.

【0004】

When a mobile unit moves from an area covered by one base station to an area covered by another base station, communication is transferred from the former station to the latter station within an area where different cells overlap.

【0005】

Because of this overlap coverage, the channels assigned to the cells are carefully selected to prevent adjacent cells from transmitting and receiving on the same frequency. The channels used by adjacent base stations are also well separated from each other in frequency so that signals from any base station do not interfere with signals from other base stations. It assigns channels to the central cell, which are widely separated in frequency as described above. Then, channels are assigned to the cells surrounding this central cell in a pattern that optimizes the cover and minimizes interference.

【0006】

The pattern of channel allocation is similarly continuous in other cells close to the center cell. The technique of separating each channel assigned to a cell from the next channel assigned to that cell is to place a large number of cells with completely different frequencies in the system before any frequency is repeated. It is possible to do. The pattern is often referred to as a frequency reuse pattern and varies widely.

【0007】

In some systems, especially those with cells in busy urban areas, each cell is further divided into three sectors, each sector having a channel allocation. Antennas in each sector are arranged to provide 120 degree coverage in the cell. Slightly over 400 channels are available, repeating patterns can be achieved with a hexagonal arrangement, using 7 cells, each with 3 sectors.

【0008】

Theoretically, according to this cell arrangement form and channel allocation proposal, frequency reuse is repeated at sufficiently separated distances to minimize interference between mobile units.

【0009】

Unfortunately, interference occurs for several reasons. Antenna patterns, power levels, dispersion and wave diffraction are different from cell to cell. The signal strength transmitted by physical objects such as buildings, hills, mountains, and trees varies over the area covered by the cells. Therefore, the signal strength boundaries of channels that fall below a level sufficient to support communication with mobile units vary widely from cell to cell.

【0010】

For this reason, cells that are close to each other do not actually form the exact geographical boundaries described above, and cover gaps can occur. Overlapping coverage often occurs between channels because cell boundaries overlap to provide complete coverage of the area and cell boundaries are inaccurately defined. As a result, channels with overlapping covers may interfere with each other.

【0011】

Due to the odd-shaped boundaries of cells and the need for cells to have overlapping coverage areas, multiple signals on the same channel are more likely to interfere with each other, even if they are generated by cells that are not close together or are widely separated. To do. This is especially true when sectorized cell patterns are used because the cells are closer together than in simple cell patterns.

【0012】

Signals originating from remote locations (also known as interferol I) are between the two signals and a second stronger signal carrying mobile communications (also known as carrier signal S) on the same channel. Interferes when the difference between is less than the threshold (measured as the S / I ratio in decibels). In addition, signals on adjacent channels are carried by the abutting cells according to the frequency reuse pattern described above. Frequency filters are not sufficient to completely eliminate the reception of adjacent frequencies. Therefore, there will be proximity channel interference.

【0013】

When the difference between the two signals is less than the second (usually low) threshold, the signals on adjacent frequencies interfere with the communication link. Channels that are usually in close proximity can have a signal strength close to that of the communication link without causing significant interference. This is because the frequency filter on the receiving side excludes all parts of the signal strength of the adjacent channel.

【0014】

When designing or reconstructing the cover of a mobile cell system, in order to optimize the cover and overcome interference, the mobile system operator uses predictive software to use predictive software from each of the specific cell populations across the mobile system. Determine if signal strength is expected. The software can utilize data showing the physical characteristics of the terrain surrounding each cell site and the physical characteristics of the cell station that plots the predicted signal strength around the cell site.

【0015】

The signal strength predictions are overlaid on the graph plot to determine where the antenna should be placed to provide optimal coverage with appropriate overlap areas of the handoff. Once the antenna site is determined, the operator assigns the channel group to the cell by the method described above.

【0016】

(Gist of the Invention) Even if only the prediction software is used, the cover cannot be optimized in general, and the planned interference between the signals cannot be excluded. To find the difference between the predicted system characteristics and the actual system characteristics, collect data, measure the signal strength of each sector at many points within the range in question, and use those measurements as the predicted characteristics. It will be necessary to compare. Data can be collected by many methods of measuring the signal strength of a channel.

【0017】

During data collection to determine RF signal strength, indications of RF signal strength at multiple locations throughout the system are collected. In one form of collection (called a driving test), a mobile unit equipped with a scanning receiver travels on the road in the test area. As the mobile unit travels throughout the system, the scan receiver scans, receives, identifies and measures the strength of the signal transmitted by the cell and / or sector. For each measurement point, the intensity measurement of the signal transmitted by each cell is obtained.

【0018】

The signal strength is recorded for the transmitted cell and sector. The data collected shows the RF cover of the system. Also by comparing the specific strengths of signals originating from different sectors and measured at a common point, which sector potentially interferes with each other sector when using common or close channel frequencies. Can be determined.

【0019】

Thus, an effective cover pattern can be used to optimize the cover, and the actual points of interference can be used to determine if there is sufficient interference to change the channel group assigned to a particular region. can do. The collected data can also be used for many other analyzes and installation specifications. The present invention is not dependent on specific data collection and is not limited to specific analytical specifications. The present invention is a post-processing technique that can be widely applied to network growth management.

【0020】

If the geographic area of interference is large enough, the channel allocation pattern for the system will change. That is, when a frequency or frequency group is assigned to an antenna, the sector or cell is completely changed to another non-interfering frequency group. Occasionally, interference can be eliminated by changing the cell characteristics without changing the channel allocation.

【0021】

The present invention processes data collected after data collection and before additional analysis. By identifying interference between cell channels using both measurement and predictive data for the system, the present invention presents prior art problems in designing or reconstructing antenna gain patterns, orientation, placement and selection, etc. It is something to overcome. The present invention identifies planned interference based on dynamic information that corresponds to the behavior in use.

【0022】

The method of the present invention analyzes and normalizes the received data to remove the effect of antenna orientation from the propagating profile of the measured and modeled cell system. In addition, the present invention is used to reassign radio frequency (RF) channels and power settings based on measured system behavior and predicted parameter variations. This maximizes capacity and minimizes interference throughout the system. The analysis method here enables system expansion planning and optimum antenna gain pattern selection.

【0023】

When collecting data for analysis and optimization of wireless systems, the measured path loss data is used as the basis for cell system analysis. Path loss is defined as the attenuation between the source antenna (sector) and the terrain point of the radio system. To measure path loss, the Received Signal Strength Display (RSSI) is subtracted from the known transmitted signal level for the received signal at a terrain location. The path loss formula is expressed as follows. Here P_txThe pattern is the transmitted power.

【0024】

[Number 1]

<img file="JP2002204468A_D0001.tif" />【0025】

RSSI values can be obtained at various locations in the cell system by appropriate methods. Once the data is collected, it is post-processed. In this post-processing, each measured value is merged with the sector transmission so that the route loss is calculated. In this merger, the measurement RF channel is matched with the RF channel used for the sector. This path loss calculation requires knowledge of at least one reference RF channel used in the sector and its transmit power.

【0026】

A complete analysis of the cell system requires path loss information at each measurement location for each sector to be serviced and for sectors that interfere at a particular location. For example, if the system requires a signal-to-interference ratio (S / I) of 20 dB for non-interfering communication, the collection should include an RSSI power level within 20 dB of the signal level recognizable and serviced at the location. It is necessary to obtain the measured value of the sector signal held. This information can be used to automate optimization tasks such as determining frequencies and power allocations that provide the desired quality of service.

【0027】

In practice, it is not always possible to obtain a measurement path loss at all potential measurement points. Therefore, in the present invention, it is possible to increase the measurement path loss information that can be obtained together with the predicted path loss information obtained by using the propagation model. Form regression analysis can be applied to the preliminary results provided by the propagation model to better align the propagation model with the actual cell system.

【0028】

Other techniques such as geographic analysis can be used to allow weighted measurements and interactive interpolation of propagated data. Once the propagation model has been refined, the path loss information generated by the model is combined with the measurement information. The combined information is then used to improve or model the behavior of the cell system. As mentioned above, path loss is measured by subtracting the signal received signal strength indication (RSSI) of the signal at some point on the terrain from the known transmitted signal level for the received signal. This is shown by the following equation. Here P_txIs the transmitted power.

【0029】

[Number 2]

<img file="JP2002204468A_D0002.tif" />【0030】

P_txIs the transmission power radiated by the transmitting antenna, so the above path loss equation does not give propagation path loss. Instead, the equation gives the coupled path loss of the RF link including the transmitting antenna and the propagation medium, and assumes that the propagation path loss of the receiving system has already been factored into the value of RSSI. Propagation path loss can also be obtained from the above equation by incorporating the gain given by the transmitting antenna as a factor. Once the antenna gain is removed from the path loss calculation for the RF link, the resulting path loss value is used to model the propagation characteristics of the communication signal between the transmit and receive points, regardless of the transmit antenna gain or beam pattern. be able to.

【0031】

An embodiment of the present invention relates to a process of optimizing the behavior of a wireless communication system and includes many steps such as:

【0032】

Identify the driving test area. Signals from each of multiple sectors of multiple cell sites for the driving test area are measured to determine the receiving power of the measured signals. Normalize each measurement signal to a conceptual antenna with an isotropic radiation pattern, except for the effective antenna gain associated with each measurement signal.

【0033】

Correlate each measurement signal with the corresponding modeled conceptual signal. The propagation model is modified based on the correlation of each measurement signal to the corresponding modeled conceptual signal to generate a modified propagation model that better predicts the expected path loss of the conceptual signal. A modified propagation model is applied to the conceptual signals from each sector to determine the predicted path loss for additional non-overlapping spatial bins. Plan antenna changes, changes and replacements based on conceptual models to improve coverage through more effective use of directional antennas, antenna gain patterns and antenna arrangements.

【0034】

A modified propagation model is applied to the planned signal from each planned sector to determine the planned path loss of the planned signal for each bin. And / or repeat these steps many times to improve the co-channel signal-to-noise ratio for the planned signal in each bin.

【0035】

In the present invention, a database including antenna specification data corresponding to the antennas deployed in the system is used to normalize the data excluding the influence of the antennas. The database may contain additional information about other antennas, which allows the system to equip multiple propagation analyzes at the cell site location through replacing the antenna with known performance to improve system coverage. Can be done.

【0036】

(Example) Here, terms used in the present specification are defined. The following performance of mobile systems refers to the physical characteristics of the system.

【0037】

The term "place" refers to a cell site site on the earth. The term "RF mount" refers to the position of the antenna above a certain point on the earth. The term "RF path" refers to the directional installation of the antenna on the RF mount, and more specifically to the physical information needed to determine how the antenna concentrates energy in the geographic area. .. The term "path loss point" refers to the path loss value and associated geographic information for a particular point in space. These are shown in Table 1.

【0038】

[table 1]

<img file="JP2002204468A_D0003.tif" />【0039】

The properties shown in Table 1 have the following meanings in the present specification. However, the physical properties of the cell system vary greatly from the systems exemplified here, and they are also within the scope of the present invention.

【0040】

Longitude: Each coordinate position in centimeter seconds, positive for the eastern hemisphere. Latitude: Each coordinate position in centimeter seconds, positive for the Northern Hemisphere.

【0041】

Elevation: The height from sea level to land level on the meter. Radius center: The height from land to the center of the antenna on the meter, often abbreviated as radctr. Array: The physical direction of the antenna in the horizontal plane of the earth, the unit is degrees with respect to true north, and the array increases clockwise from north.

【0042】

Mechanical Tilt: The physical orientation of the antenna in the vertical plane of the earth, in degrees with respect to the horizon, positive for angles below the horizon, negative for angles above the horizon. Is the value of. This parameter is often abbreviated as mtilt. Electrical tilt: The shift of the main lobe of the antenna pattern in the vertical plane of the earth, in degrees, following the same representation as mechanical tilt, this parameter is often abbreviated as etilt.

【0043】

Antenna: A group of characteristics that indicate the characteristics of an antenna. Path loss: The difference in signal level between transmit power and receive power, expressed in dB. Fluctuation: A range of recorded path loss values, expressed in ± dB. Actual measurements will fluctuate with respect to a given confidence level.

【0044】

In the present invention, the predicted path loss value is generated from the normalized path loss value, and a group of data for different antenna gain structure analysis is created. The formation of predicted path loss values and the processing of interference patterns for various antenna specifications improve the cell system cover.

【0045】

The entire process of the system of the present invention is shown in FIG. A cell system analysis program capable of carrying out the invention is executed by a computer to evaluate the qualitative and quantitative concepts of the cell system. If the analytical program requires data about the RF path, the measurements and predictions are combined to form the binned path loss value.

【0046】

This interpolation process uses a path loss value that does not include the influence of the antenna. The effect of the antenna is absent because it is removed from the measured path loss value or not initially added to the predicted path loss value. If the gain characteristics of the RF path antenna are required for system analysis, they are additionally incorporated into the binned path loss value.

【0047】

The invention introduces the concept of RF mounts and is used to characterize all RF paths at the same location and altitude. Moreover, the influence of each antenna is excluded. The measured RF path data is converted to RF mount data by removing the gain given by the antenna pattern. The gain characteristics of the isotropic source in the propagation model are used to generate the predicted data for the RF mount. After both types of RF mount information have been generated, the interpolation method combines them to generate bin RF mount data.

【0048】

By adjusting the calculated path loss value by removing the antenna gain from the measurement data, the RF path path loss data is converted into the RF mount path loss data. Figure 2 shows the overall flow. This allows the data to be grouped by RF mount, which is more easily combined with the predicted data and more easily adjusted when small changes occur in the RF path structure. It is also possible to quantify the environmental propagation characteristics of the RF path and change the fluctuation characteristics of the path loss value according to the propagation characteristics by using the process of removing the influence of the antenna gain from the RF path path loss value.

【0049】

Predicted path loss data is generated for the RF mount using the propagation model. The resulting values are easily merged with the RF mount data by considering the relative variation of the source. Predictions are generated using factors derived from the relevant measurements of the RF path. Each measured or predicted value has an associated variation value, which reflects the reliability of the value. Measured values are generally assigned greater confidence than predicted values. The reliability of the measured value depends on the distance between the measurement location and the antenna, the arrangement of the antennas, and so on.

【0050】

It is difficult to determine the relative number of measured and predicted path loss values required to accurately characterize a particular region of the cell system. For this reason, each path loss value is given a variable characteristic. Figure 3 shows the interaction between the reliability of the forecast and measurement data. Relatively high variability is assigned to the predicted path loss value and relatively low variability is assigned to the measured path loss value. This is because the reliability of the measured data is greater than that of the predicted data.

【0051】

If sufficient measurement data is available in the region, the interpolation method can ignore the predicted high fluctuation values, so it is better to include the predicted values and let them decide where to use it for the interpolation process. Therefore, the path loss interpolation method can accept both measured and predicted data, and the variation can lead to an interpolation algorithm that selectively weights significant differences in the source data.

【0052】

Fluctuations assigned to the measurement data are affected by the reliability of the antenna manufacturing specifications. Antenna specifications are more accurate in the main anterior lobe than in the ancillary posterior lobe, which can be significantly affected by antenna installation. When the antenna gain pattern is removed in the post-processing of this invention of the collected data, inaccuracies are mixed by the antenna reliability factor, especially in the lateral or posterior lobes. Therefore, the fluctuation of the measured value may fluctuate due to the antenna fluctuation factor.

【0053】

In order to better understand the purpose and means of eliminating the influence of the antenna gain for each bin path loss value, the characteristics that define the antenna will be described. Of most interest are the properties that define electrical properties. The electrical properties are provided by the antenna manufacturer to identify the ability of the antenna to concentrate RF energy in three dimensions. This information is defined by the maximum gain and the gain offset value surrounding the two sets of antennas. The gain offset value identifies the difference in maximum gain at points along the horizontal and vertical planes to form horizontal and vertical gain patterns.

【0054】

The gain pattern is represented by a 360 gain offset value, the unit of which is decibel (dB). The gain offset value represents the difference in gain from the maximum gain point of the antenna to a certain gain point, and is displayed as an angle (0 to 359 degrees) from the bore site of the antenna. Therefore, the gain offset value is zero for the horizontal and vertical gain patterns at the maximum gain point of the antenna, and negative or zero for all other points in the gain pattern.

【0055】

The bore sight of an antenna refers to the physical direction (aiming direction) of the antenna. Boresites correspond to angles 0 and 0 degrees in horizontal and vertical planes. The bore site is not necessarily the point of maximum gain on the gain pattern. The gain pattern is a complex three-dimensional shape. To represent this shape, the manufacturer provides cross sections of the gain pattern in horizontal and vertical planes. The 3D gain pattern is constructed by combining their horizontal and vertical cross sections. Figure 4 shows the three-dimensional gain pattern for the simplified antenna pattern.

【0056】

The gain offset for the receiving point in space is determined by the angle between the antenna and the receiving point. When the receiver point is located on a horizontal or vertical plane, the gain offset can be obtained by calculating the angle on the gain pattern and collecting the gain offset value. In many cases, the receiver points are not located in their planes, so interpolation is required to determine the gain offset. Figure 5 shows the receiver points off the horizontal and vertical planes.

【0057】

A line is drawn from the antenna to the receiver point to interpolate the gain offset for the receiver point in space. The intersection of this line with the horizontal and vertical gain patterns is then determined. The gain offset value forms two points on the elliptic curve that meets the intersection. The surface of this elliptic curve is orthogonal to the bore site of the antenna as shown in FIG.

【0058】

The antenna gain associated with the RF path path loss at the binned receiver point is determined by the following steps. (1) Calculate the horizontal angle between the binned receiver point and the antenna to determine the horizontal gain offset. (2) Calculate the vertical angle between the binned receiver point and the antenna to determine the vertical gain offset. (3) Calculate the angle between the binned receiver point and the antenna on the plane orthogonal to the bore site and interpolate the spatial gain offset. (4) Calculate the antenna gain by subtracting the spatial gain offset from the maximum antenna gain.

【0059】

In FIG. 7, in order to determine the horizontal gain offset from the horizontal antenna pattern, the angle θ between the antenna bore site and the binned receiver point must first be determined. In fact, this involves calculating the angle from true north to the binned receiver point and subtracting the angle between true north and the boresight arrangement from this angle. These calculations are performed as follows.

【0060】

The angle between the binned receiver point and true north is given by atan (dX / dY). Where dX = X_P-X_S, DY = Y_P-Y_SAnd X_PAnd Y_SIs the plane coordinates of the binned receiver point, X_SAnd Y_SIs the plane coordinates of the source antenna. The angle θ is determined by the following mathematical formula.

【0061】

[Number 3]

<img file="JP2002204468A_D0004.tif" />【0062】

Here, the RF path sequence is the angle between the antenna bore site and true north. If θ is known, then the antenna's horizontal gain offset G_{ant, hor}[θ] is obtained by examining the horizontal gain pattern of the antenna.

【0063】

In FIGS. 8A and 8B, in order to determine the vertical gain offset from the vertical antenna pattern, the angle σ between the effective bore site of the antenna and the binned receiver point must first be determined. In fact, this involves calculating the angle between the RF path horizon and the path loss data point and subtracting the RF path mechanical and electrical tilt angles (ie, the effective boresight angle) from this value. This calculation is performed by the following formula.

【0064】

[Number 4]

<img file="JP2002204468A_D0005.tif" />【0065】

Here X_P, Y_P, Z_PAre the coordinates of each binned receiver point, X_S, Y_S, Z_SIs the coordinates of the source antenna, the RF path etilt is the electrical tilt angle between the effective boresight and the actual boresight, and the RF path mtilt is the angle between the horizon and the actual boresight.

【0066】

Antenna vertical gain offset G by examining the antenna vertical gain pattern using known σ_{ant, ver}[σ] can be obtained. Horizontal and vertical gain patterns are stored in a database with information about each antenna deployment in the system. This database may also have information about a large number of replacement antennas for use in processing the proposed antenna replacements.

【0067】

FIG. 10 shows an example of typical antenna pattern data. G_{ant, ver}In determining [σ], it is necessary to consider whether the binned receiver point is before or after the antenna, but this consideration is omitted. Those skilled in the art will understand how to incorporate the consideration of whether the binned receiver point is before or after the antenna.

【0068】

After checking the horizontal and vertical gain offset values, elliptical approximation is used to interpolate the gain offset. In order to perform this interpolation, it is necessary to calculate the angle φ from the antenna bore site to the path loss data point. Figure 9 shows this angle in the plane orthogonal to the boresight and containing the ellipse. To calculate the angle φ, use the following formula.

【0069】

[Number 5]

<img file="JP2002204468A_D0006.tif" />【0070】

To calculate the gain offset, it is necessary to solve the radius of the ellipse at the angle φ. This is done by the following formula.

【0071】

[Number 6]

<img file="JP2002204468A_D0007.tif" />【0072】

If the path loss value includes the effect of the antenna, the antenna gain is subtracted from the path loss data point.

【0073】

[Number 7]

<img file="JP2002204468A_D0008.tif" />【0074】

To remove the influence of the antenna in the path loss value, add the antenna gain to the path loss data point as shown in the following formula. In either case, the path loss is a positive value.

【0075】

[Number 8]

<img file="JP2002204468A_D0009.tif" />【0076】

The variation is additive because the antenna pattern variation is independent of the path loss data point variation. Therefore, the fluctuation of the antenna pattern is added to the fluctuation of the path loss data point. The process of calculating the antenna fluctuation uses the same procedure as in the case of the gain offset, and has the following steps in particular. (1) VAR_an_{t, hor}Calculation of [θ], (2) VAR_{ant, ver}Calculation of [σ], (3) VAR_antCalculation of [φ] and (4) Calculation by the following formula.

【0077】

[Number 9]

<img file="JP2002204468A_D0010.tif" />【0078】

Path loss data normalization (PLDN) allows you to add or remove antenna effects from a set of path loss data. In this process, for each path loss data point, the path loss value is offset from the antenna gain in relation to each path loss data point location. In addition, the variation associated with each path loss data point can be adjusted by the variation of the corresponding point on the antenna pattern.

【0079】

A typical cell tower has three sectors, each served by a different transmitting antenna. Each antenna forms a unique RF path. The measurement path loss data is stored separately for each RF path. To combine the data from all three RF paths into a single dataset, the data must be normalized by removing the antenna effect from each path loss point measurement.

【0080】

RF path data is used to (1) combine measurement data from many RF paths into a single RF path, (2) combine measurement data with modeled data, and (3) existing RF. The mount is rewritten to RF mount data for various reasons, such as to model a new RF path.

【0081】

The PLDN is initialized by providing the desired RF route. PLDN works with a single path loss point at a time. Each path loss point is provided individually for computer processing. Using the above procedure, the horizontal and vertical offset angles {θ, σ} with respect to the path loss point are calculated. The horizontal and vertical offset gain values are then retrieved from the antenna database.

【0082】

The antenna database contains the same offset gain information as the horizontal and vertical gain pattern graphs. But in the database, the antenna value is loaded in one increment. Gain G of an antenna at an angle φ in three-dimensional space using horizontal and vertical gain offsets_antInterpolate [φ]. Here, φ = atan (sin (σ) / sin (θ)). Once the offset gain is calculated, the RF mount path loss value is calculated.

【0083】

For example, suppose the RF path loss for a path loss point is derived from a measurement at the point and is known to have a value of 102 dB. Furthermore, the RF path antenna gain G for the angle φ_antIt is assumed that [φ] has a value of 5.2 dB. The RF mount path loss is calculated as the sum of the RF path path loss and the RF path antenna gain and has a value of 107.2 dB. Although it does not actually affect the path loss of the antenna gain propagation link, summing the RF path path loss with the antenna gain is useful for characterizing the RF link propagation loss.

【0084】

This propagation loss is characterized by RF transmission originating from an isotropic radiator and is commonly referred to as free space loss. RF path path loss includes propagation effects of both free space loss and antenna gain. Rewriting RF mount data with RF route data is the reversal of the above process.

【0085】

The gain pattern provided by the manufacturer represents the theoretical gain behavior of the antenna model. These gain patterns are usually measured in a laboratory environment under tuned conditions.

【0086】

When the antenna is deployed at the cell site, the actual gain pattern is influenced by the erection state and the geographic state and frequency of the neighborhood. Due to these effects, the actual gain pattern changes from the theoretical gain pattern reported by the manufacturer. In particular, the gain pattern on the outside (side lobe) of the main beam of the antenna is liable to change in an actual environment. Within the main beam of the antenna, the gain pattern matches the theoretical gain pattern.

【0087】

The variation pattern identifies the data associated with the gain pattern. This variation is included when the gain pattern data is incorporated into the data by adding or removing gain patterns for a series of path loss data points. Fluctuation patterns are represented by a sequence of 360 decimal values (dB) depending on the horizontal and vertical gain patterns.

【0088】

Since the antenna pattern fluctuation is independent of the path loss point fluctuation, the fluctuation is additive and is expressed by the following formula. Here VAR_ant[θ, σ] is the fluctuation value in the direction of the path loss point, and the elliptical approximation method that does not separate the fluctuation for the horizontal and vertical antenna patterns is used.

【0089】

[Number 10]

<img file="JP2002204468A_D0011.tif" />【0090】

Since the variation increases with each PLDN rewrite, it is important to use the minimum number of rewrites to model a particular RF path or RF mount. For example, when modeling 10 different RF paths for the same RF mount, the first measurement data should be used as an input for all 10 rather than the output from each model as an input to the next model. Is.

【0091】

The propagation model calculates the path loss between the RF mount and the path loss point. Many different propagation models have been developed to capture a variety of different propagation effects. Since it is not possible to model all the factors that affect path loss, the calculated values only estimate the path loss within the statistical variation. An example of the propagation model is as follows. Where h is the effective height (meters) of the RF mount, r is the distance (meters) between the RF mount and the path loss point, and c₀, C₁, C₂, C₃Is a constant, diffraction is a diffraction component, local effect is a correction factor for adjusting local terrain, w₀, W₁, W₂, W₃Is a weighted factor with a default value of 1.

【0092】

[Number 11]

<img file="JP2002204468A_D0012.tif" />【0093】

The general formula is extended as follows and the following formula is obtained by combining common terms. In the final form of this general expression b₀~ b₄Is the path loss coefficient.

【0094】

[Number 12]

<img file="JP2002204468A_D0013.tif" />【0095】

If the location of the path loss point is known, the path loss data can be generated for a single path loss point or a series of path loss points. In one application of the generation process, when the modeled path loss value must be calculated for each measurement path loss point used in the coefficient regression, it is used to generate the path loss coefficient. To generate path loss data For each path loss point, RF mount, receiver height (meters), signal frequency (MHZ), propagation coefficient (b)₀-b₄) And path loss fluctuation (dB), etc. are needed.

【0096】

The propagation coefficient and the path loss variation are calculated during the occurrence of the path loss coefficient. During this initial modeling, all coefficient values are set to 1. Diffraction values are calculated for the line segment area as described below.

【0097】

Path loss data generation can be used to calculate path loss for new path loss points that are uniformly distributed along a single ray, or for multiple rays that are evenly distributed within an azimuth. To generate path loss data using the specified radius method, RF mount, receiver height (float meters), signal frequency (float MHZ), propagation factor (float b)₀-b₄), Path loss variation (float dB), number of radii (int), radius length (int meters), step size along radius (int meters), starting azimuth angle (int degrees) and inscribed arc angle (int degree) Information about int degree) etc. is needed.

【0098】

Figure 11 shows the input parameters as a graph. The initial radius is formed using the radius length and the azimuth angle. The step size value is used to determine the location of the path loss point along the radius. The location of each path loss point along the radius is rewritten as longitude and latitude. The height is derived from the terrain database. Propagation and diffraction losses are calculated one after another for each path loss point. Divide the inscribed figure by the number of radii and add this value to the array of previous radii to calculate the array of radii after the second.

【0099】

Figure 12 shows the resulting output. The path loss points are evenly distributed along each radius, and the radii are evenly distributed within the inscribed circle. There are three variables to express the radius: total length, step size and number of steps. There are three variables that represent inscribed loneliness: total lone span, angular spacing, and number of half-width characters. As shown in the following table, these values are represented in nine different ways, showing the same distribution of path loss points.

【0100】

[Table 2]

<img file="JP2002204468A_D0014.tif" />【0101】

In the preferred embodiment, the first set of inputs is used to represent the distribution of path loss points. Similarly, other input sets can be supplemented.

【0102】

In propagation modeling, the height of the transmit and receive antennas has a direct effect on the total path loss for a given link. Path loss generally decreases as the height of the antenna increases. In a flat land model, antenna height is a characteristic that can be easily determined. In an actual environment with various terrains, there are methods for handling various antenna heights as shown in FIG. The movable antenna height is ignored in the processing of all antenna heights. The height of the movable antenna is usually in the range of 1.5 to 2.0 meters. The base station antenna height has significantly larger fluctuations.

【0103】

In the generation of path loss data, the effective antenna height is used for the overall propagation model. In the diffraction algorithm, the actual antenna height is used rather than the effective antenna height. However, the diffraction algorithm directly calculates the effect of the terrain on propagation, resulting in a more accurate measurement.

【0104】

For the global propagation model, the effective antenna height is calculated as if it were above the average terrain (HAAT). The average terrain altitude is derived from the terrain database. Figure 13 shows an example of calculating the effective antenna height using HAAT.

【0105】

The FCC defines standard methods for calculating average terrain used by public mobile services in all application areas. Using this method, eight evenly distributed radii are drawn from the RF mount starting from true north (0 degrees). A minimum of 50 altitude points are collected along each radius, evenly spaced between 3 and 16 km (2 to 10 miles). The average altitude is calculated for each radius. The radial altitudes are then averaged to give the final average terrain altitude. Figure 14 shows a statistical method for calculating average terrain altitude.

【0106】

As shown in FIG. 15, the propagation model diffraction algorithm calculates how obstacles interfere with the signal path between the RF mount and the path loss point. The diffraction algorithm uses the Picquenard model. In this model, the effective height of the first obstacle is calculated from the baseline formed by the line segments connecting the transmitter and receiver. The effective height of each subsequent obstacle is measured from the baseline line drawn from the top of the previous obstacle to the receiver.

【0107】

Diffraction values are calculated for the measurement path loss points during the coefficient generation. Certain RF mounts may have more than 40,000 path loss points. It is inefficient to calculate the altitude separately for each path loss point.

【0108】

As shown in FIG. 16, another method divides the area around the RF mount into segmented areas. A single diffraction value is calculated for each area and assigned to path loss points within that area. To implement this segmented area method, the diffraction algorithm uses RF mount, receiver height (float meters), signal frequency (float MHZ), number of radii (int) and step size along the radius (int metric). ) Is required.

【0109】

The number of radii is used to divide the area around the RF mount into radial regions with uniform angular spacing. Each path loss point is assigned to a specific radius region according to its location. The length of each radius region is set equal to the distance to the farthest path loss point within that region (each radius region may have different lengths). Figure 16 shows the measurement path loss points assigned to the radial region.

【0110】

Each radius region is then divided into concentric sections using the same uniform step size as shown in FIG. Finally, calculated diffraction values are assigned to all path loss points in the center of each section of each region. Figure 18 shows the final form.

【0111】

A typical propagation model is based on optimal calculation of measured field data. When applying a general model to different locations, the outcome depends on how the characteristics of the new location match the characteristics of the location from which the model was drawn. In the present invention, the measured data for each location is compared to the path loss value calculated by the path loss data algorithm. Based on this comparison, new coefficients are calculated to minimize the variation between the measured and model values. Various techniques such as simple substitution, iterative convergence and regression analysis are used for this comparison. Once the propagation model is optimized, new path loss points are modeled.

【0112】

Using the coefficient generation method, the following mathematical formula can be used to determine the weighting factor that minimizes the variation between the propagation model and the measured value.

【0113】

[Number 13]

<img file="JP2002204468A_D0015.tif" />【0114】

General formula b₀From b₄The final form to is the path loss coefficient, and the path loss equation fits the regression model. Here x₁Is log₁₀(h) and x₂Is log₁₀(h) log₁₀(r) and x₃Is diffraction and x₄Is a local effect.

【0115】

Analysis using the measured data should confirm that the path loss fluctuates linearly with diffraction and local effect values. To generate the path loss factor for an existing RF mount, the RF mount, receiver height (meters) and signal frequency (MHZ) are required as inputs.

【0116】

For each point in the path loss point list, the path loss coefficient algorithm is x₁~ x₄Calculate the value of. x and measurement path loss (y_iEach set of) forms a row in the observation matrix. Once all path loss points have been processed, a least squares matrix is formed. The path loss coefficient is calculated by solving this least squares matrix.

【0117】

To be usable, the coefficient generator must display how the modeled propagation value approximates the measured value. Also, the resulting path loss value must include a variable value that indicates a planned coupling for the path loss at a given path loss point. There are many ways to rewrite the results calculated according to the prediction factors by regression analysis. For the analysis of variability, the coefficient generator provides the variation of the modeled path loss value and the variation and standard deviation of the resulting path loss value.

【0118】

The same variation calculated in the coefficient generator algorithm applies to all path loss points modeled with the same coefficient. In point-to-point analysis, the modeled value will be smaller than the measured variation, as the measured value variation varies as a variable in the measurement environment. To be valid, variability must be analyzed based on the complete set of data.

【0119】

The coefficient generator algorithm is also used to calculate the propagation component for new (or simulated) RF mounts that do not have the measurement data as shown in Figure 19. In this case, a new regression model is needed that determines the effects of antenna height and altitude, effective antenna height, and the distance between the existing RF mount and the modeled RF mount.

【0120】

There are various methods for predicting each value, and each has an advantage in different altitude states. Crising is also a predictive method with spatial data. The data has various advantages and is used to calculate the coefficient values of the modeled RF mount.

【0121】

Although the present invention has been described above with reference to various examples, various modifications can be made within the range that can be inferred by those skilled in the art.

[Simple explanation of drawings]

[Figure 1]

It is a diagram which shows the flow of the whole process of this invention.

[Figure 2]

It is a diagram which shows the flow of the whole process of this invention. ..

[Fig. 3]

It is a diagram which shows the flow of the whole process of this invention.

[Fig. 4]

It is a figure which shows a typical three-dimensional gain pattern.

[Fig. 5]

It is a figure which shows the receiver point located in the horizontal and vertical planes of the gain pattern of FIG.

[Fig. 6]

It is a figure which shows the plane orthogonal to the bore site including the ellipse composed of the horizontal and vertical plane intersections about a receiver.

[Fig. 7]

It is a graph for determining an angle in a horizontal plane between an antenna bore sight and a receiver point.

[Fig. 8]

(A) It is a graph for determining the angle in a vertical plane between the effective bore sight of an antenna and a receiver point. (B) It is a graph for determining the angle in a vertical plane between the effective bore sight of an antenna and a receiver point.

[Fig. 9]

It is a graph for determining the intersection angle between the ellipse of FIG. 5 and the radius line drawn from the bore site to the receiver point.

[Fig. 10]

It is a figure which shows the typical antenna specification data.

[Fig. 11]

It is a figure which shows the parameter used to calculate the path loss value for a new path loss point along a ray using a specified radius method.

[Fig. 12]

It is a figure which shows the generation path loss point uniformly distributed along various radii of inscribed circles using the specified radius method.

[Fig. 13]

It is a figure which shows the example which calculates the effective antenna height using the height above the average terrain (HAAT) method.

[Fig. 14]

It is a figure which shows the static method about the calculation of the average orographic lift.

[Fig. 15]

It is a figure which shows a typical Picenard model.

[Fig. 16]

It is a figure which shows the measurement path loss point assigned to various radial regions.

[Fig. 17]

It is a figure which shows the radial area of FIG. 16 divided into the sector which had a uniform step size.

[Fig. 18]

It is a figure which shows the diffraction calculation method in each sector of FIG.

[Fig. 19]

There is a graph showing the measurement of the center of radius overlapping the modeled antennas in the system.