Path loss determination using Hata model
Path loss is the degradation in received power of an Electromagnetic signal when it propagates through space. Path loss is due to several effects such as free space path loss, refraction, diffraction, reflection, coupling and cable loss, and absorption. Path loss depends on several factors such as type of propagation environments, distance between transmitter and receiver, height and location of antennas. Also the signal from the transmitting antenna may take multiple paths (multipath) to reach the receiving side, which results in either increase or decrease of received signal level depending on the constructive or destructive interference of the multipath waves.
Path loss is usually expressed in decibels (db), as this method gives us an easy and consistent method to compare the signal levels at various points.
Where Lp is the path loss i.e. the ratio of power of received signal to that of transmitted.
Propagation models are used extensively in network planning, particularly for conducting feasibility studies and during initial deployment. They are also very useful for performing interference studies as the deployment proceeds. Numerous Experiments have been carried out all around the world, for checking the applicability of suitable path loss models in mobile communications specific scenario. In those experiments radio engineers carried out signal strength measurement for a specific area and compared the observed output with that of predicted outputs from different widely accepted propagation models so as to find out which model best predict the path loss for the given scenario. Also, some fine tuning for that model are done based on the difference.
Many propagation models are available for path loss predictions. Deterministic models are based on the laws of electromagnetic wave propagation and produce accurate predictions of the path loss, however they take high computational effort and require detailed and accurate description of all objects in the propagation path .Free space Model, and Plane Earth model, are easier deterministic propagation methods. Empirical models are based on extensive collection of data for specific case. They are not accurate but predict the most likely behavior the link may exhibit under specific conditions.
Okumura carried out a lot of field strength measurements in Tokyo, Japan with varying terrain, frequency antenna height and transmitted power. It states that, the signal strength decreases at much greater rate with distance than that predicted by free space loss .An empirical formula based on Okumura’s results has been developed by Hata in order to make the propagation loss prediction easy to apply .Hata gave no of path loss models for urban, suburban and open areas, often called Okamura-Hata model. The European Co-operative for Scientific and Technical Research (COST) extended the Hata model to be valid for PCS(personal communication System) Operating at 1800- 2000 MHz, which is named COST 231 Hata model .The European research group also developed an another model taking Diffraction effect into consideration, which is COST 231 Walfish Ikegami model. Actually, this model is a semi-deterministic model as the path loss calculations is made by combining empirical models with deterministic techniques .wave propagation models for planning of mobile communication network. There’s a North American model called Lee’s model named after W.C.Y. Lee, which is characterized by two parameters, power at a mile and path loss exponent.
Kathmandu is the capital city of our country Nepal with more than 1 million inhabitants. GSM Mobile launched by Nepal Telecom reached almost a decade now with total 3.5 million subscribers, and 1.5 million solely in Kathmandu. With such increase in no. of subscribers, the no of BTS has gone up to 1500 in whole Nepal. Operating with both the frequency 900 and 1800 MHz, there are nearly 500 BTS (BS) in Kathmandu only. With such increase in the no. of Bs in Kathmandu valley due to the growth in No of subscribers, proper planning methods are needed for placing BTS for acceptable field strength and Interference level in the coverage area. Also the Coverage holes are needed to be filled with proper planning. This Planning Requires appropriate propagation models that best predicts the path loss for our specific environment.
In this project, the feasibility analysis and fine tuning of Hata and Cost 231 Hata model has been done for propagation environments in Kathmandu. Taking three different areas of Kathmandu namely: Core city area, Ring-road area and Remote areas and five locations are selected for each of the terrain types and Field Strength Measurements offers a better means to understand what path loss model to use in certain propagation environments. Field strength measurements were conducted on the existing GSM 900 MHz Network of Nepal Telecom on the locations of interest.
The demand for increasing mobile subscribers needs efficient extension of a cellular network. For GSM like technology, there is additional complexity in making efficient Allocation of Base Stations and frequency planning. For this, proper path loss models should be used for coverage prediction and interference analysis. Much of the popular path loss models by renowned RF engineers are based on the observation data taken on their own country. So there is need of fine tuning of such model for applicability in our area. To my knowledge, there has not been any published work regarding such models feasibility analysis and modifications.This provides the impetus for this project to make an analysis of the observed data and necessary modification in Hata and Cost 231 Hata model for Kathmandu.
Also, the proposed re-farming of 900 MHz frequency spectrum for Future Generation of Mobile Communications provides another drive for this project. It is obvious that use of less frequency yields large coverage area than higher frequencies. Hence, for future generation mobile, Scientist have conceived of making use of 900 MHz spectrum and together with use of newer technologies, they could make efficient utilization of this spectrum.
This project can be used in realistic planning of GSM networks, with the predicted path loss in Kathmandu. For capacity enhancements of the network, GSM Engineers can use the path loss prediction models for intelligent placement of BTS’s with certain antenna height and power. Path loss also facilitates Link budget analysis and Design in a telecommunication system. In a GSM like Cellular network, Path loss is used for Frequency Re-use distance estimates so as to properly space the channels in Base stations (BS). Actual Frequency assignment plans for the Base Stations are also facilitated by the use of path loss. Better Coverage predictions and interference reduction is what planning engineers get by using better path loss model.
PATH LOSS MODELS
In this chapter, description of various Path Loss model will be discussed. The model characteristics, along with mathematical formulae, will be shown that is useful for further calculations, analysis in this project. In addition, three kinds of dependency factors of path loss will be noted that will be analyzed later.
FREE SPACE PATH LOSS
This model is the most simple and primitive path loss model where the influence of all objects and obstacles in the propagation environment is ignored. Here, the Received signal is inversely proportional to the square of distance between the transmitter and receiver. So, the free space Path loss is given by
As Decibel (db) method gives us easy and consistent method to observe and analyze the signal levels at various points, the free space path loss can also be expressed in logarithmic format as
This model is a RF propagation model that was developed based on the data collected in the Tokyo city, Japan. The model served as a base for all other empirical propagation models. In this model, the propagation area is divided into terrain categories: open area, suburban area, and urban area. Urban area is used as a reference area and Correction factors are applied in it to calculate the path loss for other terrains.
It is also known as the Okumura-Hata model for being a advanced version of the Okumura Model, is the most widely used model in radio frequency propagation for predicting the behavior of cellular transmissions in city outskirts and other rural areas. This model incorporates the graphical information from Okumura model and develops it further to better suit the need . Hata Model predicts the total path loss along a link of terrestrial microwave or other type of cellular communications.
Operating frequency, Base Station Antenna height. This model is suited for both point-to-point and broadcast transmissions.
COST 231 HATA MODEL
It is also called the Hata Model DCS Extension, which is a RF propagation model that extends the Hata Model to cover a more range of frequencies. Also, this model is applicable to Open, Suburban and Urban Areas .
PATH LOSS DEPENDENCY FACTORS
Path loss usually depends on operating frequency, Base station Antenna Height, and distance of the Mobile stations (MS) from the Base Station (BS). For each of the popular empirical models discussed above, those dependency factors are observed separately and lastly combined together to give a formula for that path loss model.
PATH LOSS EXPONENT
We know that received signal at a distance from a BS or a transmitter is inversely proportional to some power of the distance, i.e. farther we go from the BS or transmitter, the received level will degrade by some factor.
For empirical models, Field strength measurements gives an easy and better way for feasibility analysis and fine tuning of the models for certain propagation environments. In our case, I intend to check the applicability and the modification of Hata and Cost 231 Hata model for propagation environments in Kathmandu. Three different terrains are taken into consideration, namely: Core city area, Ring-Road area, and remote area. Field strength measurements are carried out in fifteen different locations, where five of them belong to one terrain types. Location of BS, Transmitted power, antenna height is taken for the areas of interest in Kathmandu. Field strength measurements were conducted on the existing GSM network of Nepal Telecom and the receiver used was TEMS tool with a mobile set.
DRIVE TESTING USING TEMS
TEMS is an air interface test tool for real time diagnostic of different parameters for RF optimization. Also, all the data can be saved for later analysis. It consists of a laptop with TEMS software, a mobile set with a connector, GPS for location, and a hardware lock key. As for this project, only the received level is concerned, so the received level is noted as we drive farther from a BS.
Above Figure shows the trail of the drive test path through a vehicle, in the location of interest. Here, the distance of a sample point in the trail can easily be noted using a scale in the map. Before performing drive test in all of those sites, proper Information regarding Site locations, their Carrier Frequencies, BS antenna Height, power transmitted from the BS are taken and used in the data collection and Calculations part.
In this project, several assumptions are considered to ease data collection and analysis phase, which are listed as:
- Though the signal strength measurements are done for different BS with different Carriers (ARFCN), operating frequency is taken to be fixed at 900MHz.
- The effective power radiated from the BS is taken to be fixed at 50 dbm.
- Also height of mobile () taken to be fixed at 1.5 meters as the data is measured through TEMS inside a vehicle.
- Core city area, Ring-road Area and Remote areas in Kathmandu are assumed to be Urban, Sub-urban and Rural area respectively for simplification in analysis and comparison. If that assumption gives different result, then it can be rectified easily based on the integrated analysis of the path loss plot vs distance.
MEASURED PATH LOSS
For each of the propagation environments which have been selected and drive tested for data collection, the path loss can easily be calculated using the formula:
So, Path loss can easily be observed from the collected data, to be increasing as we go farther from the BS. Data Collected for Fifteen different Locations can also be shown in tabular format as:
PATH LOSS FROM MODELS
After the observation of measured path loss in the propagation environments in Kathmandu, we need to calculate the path loss from the formula for the two models Hata and COST 231 Hata Model, with the distance d of MS from Bs in km, Frequency f in MHz, BS height hb in meters, Antenna height in meters taken from the actual scenario. Calculations are done using Excel sheet and the data are shown as:
PATH LOSS PLOTS:
As we get all the measured data and predicted data from the two models, the average data of path loss as shown in the table below are plotted with the increasing distance d of MS from BS as shown in the graph below for each propagation environments.
COMPARISION AND ANALYSIS
From the above plot for each propagation environments, comparing the average path loss values with that of Hata and COST 231 Hata model plot gives the result that Hata model best fits the propagation characteristics in Core city area and Ring-road area whereas the COST 231 model best predicts the propagation characteristics in Remote area. Also the Minimum Square Error (MSE)calculated for each of the model is within acceptable limits, which can be calculated using the formula:
From the MSE analysis also, I find that Hata model best fits the Kathmandu Core city area and is also suitable for Kathmandu Ring-road Area. Whereas in the case of Kathmandu Remote Area, Cost 231 Hata model best predicts the path loss characteristics.
FINE TUNING OF THE BEST FIT MODEL
Due to the least and acceptable MSE in both Ring road area and remote area, the best fit models need not be modified for such propagation environments in Kathmandu. Solely based on the Minimum Square Error in path loss, the Best fit model for Core city area in Kathmandu i.e. Hata model can be modified accordingly.
This formula is the modified version of the Path loss in urban area as given by Hata model for application in Kathmandu.
By the comparison of the actual path loss with that of the predicted path loss from the two models Hata and COST 231 Hata model, and integrated analysis of all the plots, the feasibility analysis of the two models for three different propagation environments in Kathmandu is done. From which we get, Hata model is not feasible for the application in remote area due to much high minimum square error. Also the Hata model best fits the core city area in Kathmandu with MSE slightly greater than the acceptable limits as given by . So the new modified Hata model is given based on the MSE. And for Ring-road area both model is applicable but the COST 231 Hata Model best fits it due to least MSE.
In this project, the popular empirical path loss models for mobile communications are studied. Among them two models, Hata and COST 231 Hata model is compared with actual path loss in three different propagation environments in Kathmandu, for applicability in macro-cellular Base Stations (BS). Comparison and Analysis of the integrated plot of the path loss from two models and actual path loss in those environments, along with mathematical calculations gave us the result. And the best fit model based on MSE calculations is either modified or left as it is, for those three propagation environments in Kathmandu. Also the assumptions for the three areas are consistent with the actual model which made the analysis and comparison easy. The choice of which model to use actually depends on the terrain type and the surrounding environments and objects around it, the selection of that model in planning a mobile communication network will give the best result and better network is what will be the output. So, Service providers must use efficient planning tools to better model their propagation environments along with the best predicted path loss. Hence, this project is useful for such realistic planning for GSM 900 MHz in Kathmandu. Further enhancements in this project can yield similar results for future mobile technologies and other spectrum too.
PROBLEMS / LIMITATIONS:
The receiver used for this project is a mobile set which has a sensitivity of -105 dbm. Hence, during field strength measurement, I can only measure signals up to 600m from the Base station in core city area. If there was separate equipment for signal measurement, it would have yield much better result, maybe that’s the reason why there is much difference between the measured and predicted value due to lesser no of sample points. Also, Effective power radiated from the antenna is assumed to be 50 dbm, (same for all BS) due to unavaibility of measuring actual power radiated from the antenna and the losses in the RF cables and connectors. Also the antenna azimuth and down tilt not taken into consideration and sometimes, the measurement of received signal is done in locations that may be outside of the main lobe of the antenna radiation, i.e. outside of the half-power beam width (HPBW).
This work can be extended to apply for other mobile technologies in other frequency spectrum such as 1800 MHz DCS, 2100 MHz WCDMA. Also high sensitive field strength measurements tools can be used for future generation Mobile technologies for better result. As this project only dealt with the selection and modification of Path loss models for applicability in Kathmandu areas, the result of this project can be used for interference estimation and frequency assignment planning for new network or extension of existing network.
- Z. Nadir, N. Elfadhil, F. Touati, “Pathloss Determination Using Okumura-Hata
Model And Spline Interpolation For MissingData For Oman”, Proceedings of the World Congress on Engineering 2008 Vol I WCE 2008, July 2 – 4, 2008, London, U.K.
- Popescu, I.et al., “Applications of neural networks to the prediction of propagationpath loss for mobile communications systems”, Applied Electromagnetism, 2000. Proceedings of the Second International Symposium of Trans Black Sea Region, 2000.
- K.Ayyappan et. al., “Propagation Model for highway in Mobile Communication System”, from http://www.ubicc.org/files/pdf/Highway_249.pdf
- F.M. Landstorfer, “Wave Propagation Model for the Planning of Mobile Communication Network”, from http://amsacta.cib.unibo.it/1427/1 /GAAS_99_065.pdf
- J. Wu and D. Yuan, “Propagation Measurements and Modeling in Jinan City”, IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, Boston, MA, USA, Vol. 3.
- M. Hata, “Empirical Formula for Propagation Loss in Land Mobile Radio Services”, IEEE Transactions on Vehicular Technology, Vol, VT-29, No, 3, 1980.
- P K DALELA, et al., “A New Method of Realistic GSM Network planning for Rural Indian Terrains”, IJCSNS International 370 Journal of Computer Science and Network Security, VOL.8 No.8, August 2008
- V.S. Abhayawardhana, et al., “Comparison of Empirical Propagation Path Loss
Models for Fixed Wireless Access Systems”, from http://www.cl.cam.ac.uk/ research/dtg/lce-pub/public/vsa23/VTC05_Empirical.pdf