IMPROVEMENT OF WIRELESS NETWORK PERFORMANCE

BY POLARISATION DIVERSITY

Simulation from Measurement Results at only One Polarisation

Iñigo Cuiñas and Manuel G. Sánchez

Dept. Teoría do Sinal e Comunicacións, Universidade de Vigo, rúa Maxwell, s/n, 36310 Vigo, Spain

Keywords: Diversity Techniques, Polarisation, Radio Channel, Measurements, Wireless Network.

Abstract: The use of polarisation diversity techniques in reception could be adequate to improve the performance of

the wireless networks operating in deep multipath fading environments. This paper explores this possibility,

and presents a procedure to estimate the received cross-polarised power from wide band measurements

performed at just one polarisation. Three different strategies have been tested, and the results are presented

and analysed, detecting improvements even when the multipath is low in the channel. In highly multipath

rooms, the improvement in terms of received power reaches 21%.

1 INTRODUCTION

The deployment of wireless local area networks is

determined, in most situations, by the strong

multipath effect that could degrade the performance

of the complete communication system. This

multipath effect appears as frequency selective fast

fading events along the receiver path, when it is

moved; or as very low coverage at some locations

that could coincide with receiver position. In both

situations, the connection possibilities of the

network nodes may be reduced or even unavailable.

The multipath is present in most of the

environments where a radio communication system

is installed. However, it is at indoor scenarios where

the effects of multipath resulted to be more

hazardous for the performance of the system. This is

the reason why different indoor environments have

been used to check the proposal of this paper.

The work in this paper is centred in the 5.8 GHz

band, one of the assigned to wireless networks

(Dutta-Roy, 1999) (IEEE802.16, 2003) (Eklund et

al, 2002). Results of several measurement

campaigns performed in both line of sight (LoS),

non LoS (NLoS), and obstructed LoS (OLoS) indoor

environments have been used as a basis to check the

options of implementing a solution to reduce the

multipath consequences. These data are radio

channel responses, with transmitters at fixed

locations and receivers measuring along linear paths,

as explained in the second section. As we get

complex responses, we can analyse the multipath

effect, and we can try to reduce the influence of their

costs.

Among the various procedures that have been

tested to mitigate the multipath consequences,

different diversity techniques have been proposed.

The general solution of such techniques is to provide

two different propagation paths, with almost

uncorrelated received signals. Thus, the alternate use

of both signals, or the combination between them,

provide a final signal with better relation SNR than

that obtained with only one standard propagation

path.

Diversity techniques could be applied following

several strategies: frequency, space, time, angle,

polarisation, and hybrids that combine some of the

previously indicated (Dietrich et al, 2001)

(Turkmani et al, 1995).

Among the diversity techniques, the polarisation

diversity is analysed along this paper: two

orthogonally polarised signals at the same frequency

are used as inputs in the diversity receiver. In this

case, the pair of propagation paths is provided by the

pair of orthogonal polarisations. This technique has

been selected because the indoor environments

presents both multipath phenomenon and

depolarisation by transmission and reflection on the

walls, ceiling, floor, and so on. Thus, a certain

percentage of the transmitted signal would arrive the

115

Cuiñas I. and G. Sánchez M. (2010).

IMPROVEMENT OF WIRELESS NETWORK PERFORMANCE BY POLARISATION DIVERSITY - Simulation from Measurement Results at only One

Polarisation.

In Proceedings of the International Conference on Wireless Information Networks and Systems, pages 115-122

DOI: 10.5220/0002929001150122

 SciTePress

receiver orthogonally polarised. Moreover, the

spectrum consumption is reduced compared to

frequency diversity, as no new bands are occupied

by the second propagation path. Depending on the

amount of depolarised signal, the application of

polarisation diversity at reception could be more or

less advantageous.

The depolarisation indexes depends on the

building material of the wall, specifically on the

electromagnetic behaviour of the different

constructive elements, as previously studied in

(Cuiñas et al, 2009). Results presented in that paper

have been applied in the computation of the

improvement by polarisation diversity. The

information related to depolarisation is the aim of

the third section.

The computation procedure begins with the data

provided by the wide band measurement campaign,

which represent the channel response in just one

linear polarisation. Based on the depolarisation

indexes induced by each reflection or transmission

phenomena, the cross-polarised signal could be

obtained by analysing separately each multipath

contribution. The actual copolar contribution and the

synthesised cross-polar one are then the inputs of the

diversity device, which provides the combination of

both contributions. All the procedure to obtain the

final signal is explained in detail along section

fourth.

Three different strategies are then applied to the

couple of signals to be combined into the final

result: sum, maximum and average diversity. The

performance of the application of such strategies are

related and analysed in section five.

The paper is organised, then, into six sections.

The second section is devoted of the measurements,

and the third one is focused on the fundamentals of

depolarisation indexes. Sections four and five are

centered in the results: the fourth on the way to

compute them, and the fifth in the analysis. Finally,

the sixth section summarises the conclusions.

2 MEASUREMENTS

A large wide band measurement campaign was

designed with the aim of obtaining the co-polar

response of the radio channel in several indoor

environments, both LoS, and OLoS or NLoS. The

campaign involved five different environment

configurations. The band of interest is centred in 5.8

GHz, as it is focused on propagation aspects for

wireless networks.

The measurement system was based on a vector

network analyser (VNA). The wide band condition

of the measurements is a key factor in this work, as

it allows the transformation to time domain and,

once moving to time, also the identification of

different multipath contributions. These

contributions could then be individually considered

when processing the cross-polar response of the

channel.

Along the following subsections, the setup used

during the measurements, the environments where

the experiments were performed, and the applied

procedure are described.

2.1 Measurement Setup

As previously commented, the measurement system

is based on a VNA Agilent 8510-C, which can

perform measurements up to 50 GHz, so far away

our needing. Both transmitting and receiving

antennas were connected to the ports 1 and 2 of the

VNA, which acts as signal generator and as receiver.

The transmitter end was placed in static locations

at each environment, whereas the receivin antenna

was installed on the top of a engineered mast, which

moved the receiver along a linear track, by means of

a neverended screw. Figure 1 shows the setup for the

receiver.

Figure 1: Receiver setup.

This positioning system, which consists of a 2.5

meter long linear table with a millimetre screw along

it, improves the precision of the positioning

compared to moving the antenna by hand.

Both antennas, Electro Metrics EM-6865,

present omni-directional radiation patterns. This

kind of pattern is interesting in order to get all the

multipath contributions with approximately the same

antenna gain, when dealing with receiver end. And

this pattern is also important in the transmitter, to

generate the maximum amount of multipath

components.

WINSYS 2010 - International Conference on Wireless Information Networks and Systems

116

The radiation pattern of such antennas was measured

within the anechoic chamber of the Radio Systems

Research Group, being the 3 dB beam width larger

than 50 degree around the horizontal plane, in

elevation, at 5.8 GHz. Whereas, the azimuth

radiation pattern resulted to be aproximately

omnidirectional. Figure 2 depicts the radiation

pattern in elevation.

Figure 2: Radiation patttern at 5.8 GHz, in elevation.

2.2 Indoor Environments

Various series of indoor radio channel frequency

response were measured in five different

environments, some in LoS condition, one in NLoS

and the other in OLoS situation. The measurements

were taken in research laboratories, with both

computers and electronic equipment. The furniture,

when it is present, is the typical of this kind of

rooms: office tables and chairs, and laboratory

benches. The positions of transmitter and receiver

are depicted at Figures 3, 4 and 5, at different rooms.

During the measurement campaign, the transmitter

was fixed at positions Txn, being n a natural number

between 1 and 5 devoted to the five environments,

and the receiver was moved along the lines labelled

as Rxn.

The five environments was selected trying to

represent a great variety of rooms: we can compare

results at large and small rooms, at furnished and

empty places, at square and rectangular spaces, in

LoS and NLoS conditions, and so on.

The points and paths labelled as "1" correspond

to LoS situation within a large room (more than 100

square meter) and the labelled as "2" to NLoS within

the previously commented room and an adjacent

saloon. Both plans can be observed at figure 3. The

wall that obstructs the propagation channel between

both antennas in the second situation is made of

bricks and concrete.

Figure 3: Map of the measured environments. LoS is

defined as situation 1, and OLoS as situation 2.

The walls at both rooms are built by bricks and

concrete, except the fine line parallel to Rx1, which

is made of chip wood, and the upper wall (opposite

to the place where the measurements took place) that

contains a large window.

Figure 4 depicts the third environment, which is

again a LoS situation in a smaller square room, with

approximately 45 square meters.

The walls of such room are also brick made,

except the wall opposite to the transmitter, which

contains a large window.

Figure 4: Map of the third measurement environment

(dimensions in centimeter).

Finally, figure 5 depicts the situation for both fourth

and fifth environments. Both are placed in a long

room of 36 square meters. Whereas the fourth

environment consists of a completely empty room,

with perfect LoS conditions, the fifth consists of a

furnished office room, in OLoS conditions. When

furnished, office elements were placed within the

radio channel: desks, chairs, closets, etc.

IMPROVEMENT OF WIRELESS NETWORK PERFORMANCE BY POLARISATION DIVERSITY - Simulation from

Measurement Results at only One Polarisation

117

Figure 5: Map of the fourth and fifth measurement

environment.

This environment is more complicated: both North

and South walls are made of brick, the West wall is

made of chip wood in its North half, and brick the

South half, and the East wall is brick constructed,

but it contains a window.

2.3 Measurement Procedure

The transmitting antenna was kept stationary at a

height of 1.8 meter. This location guaranteed that the

radiating element was approximately equidistant

from floor and ceiling.

The receiving antenna was moved along 2.5

meter long linear paths by means of the automatic

positioner. Data were taken every one-eighth of a

wavelength (Dossi et al, 1996), which represent a de

facto standard when measuring radio channels, as

adjacent samples are far enough to be uncorrelated

and they are near enough to keep all fade event.

At each position, complex frequency responses

have been measured in a 160 MHz band around 5.8

GHz, with a resolution of 200 kHz, due to the 801

points in the frequency scan. As a consequence, the

sounder resolution in the delay domain is 6.25 ns,

while the maximum measurable delay is 5 μs.

The measurements were taken following a

procedure “measure-move-stop-measure-...” in order

to avoid Doppler effects within the data.

3 DEPOLARISATION

Indoor environments typically present large

multipath phenomena, which are the main trouble

when planning a wireless network. However, the

depolarisation induced by transmission of waves

across the walls (or by reflection on the walls) is not

commonly taken into account during the planning

procedure. And it could be useful to improve the

performance of the receiving signal if a polarisation

diversity technique is implemented at the reception

end.

This section deals with the fundamentals of

depolarisation and the depolarisation indexes used to

processing the results.

3.1 Depolarisation Phenomenon

A phenomenon associated to reflection, the

depolarisation that could be generated when a wave

beats a flat obstacle, appears to be not so fine

defined and modelled as the reflection itself. This is

probably because typical planning tools, as ray-

tracing, were initially created to be used at

frequencies corresponding to cellular phone or

television broadcasting, at which the typical

obstacles (walls) are electrically flat enough to

provide strong specular reflections.

At higher frequencies, the electrical size of a

given obstacle becomes larger. At 5.8 GHz, as an

example, some simulation tools could not work as

well as expected, because when a wave reaches an

obstacle, several reflection paths are generated in

any directions, not only the specular direction

(Cuiñas et al, 2007). And moreover, the obstacle

depolarises the wave in a certain percentage, which

is not commonly considered in such prediction tools.

3.2 Depolarisation Indexes

The depolarisation index, for any material, at any

angle of incidence and any polarisation of the

transmitted waves is the fraction of the power of this

wave that is received in the orthogonal polarisation.

From this definition, depolarisation indexes may be

computed by means of a matrix procedure (Cuiñas et

al, 2009).

The depolarisation indexes for the reflection

mechanism, computed in the specular direction, are

summarized in table 1. The values depend on the

polarisation of the incident wave, which is denoted

by “h” when it is horizontal and “v” when vertical.

The specular situation is adequate to define very

good reflectors, which reflect most of the incident

wave towards the opposite direction, being the

normal to the surface the axis of symmetry.

The results of table 1 indicate that brick wall

provides reduced depolarised waves compared to the

co-polar reflected waves in the specular direction.

The other considered materials provide depolarised

waves up to 9.8% compared to the co-polar one.

But when a more complex analysis is expected,

as it is the situation of the present paper, all

scattering directions have to be considered, and not

just specular one, because the reflector could be

randomly located and oriented.

WINSYS 2010 - International Conference on Wireless Information Networks and Systems

118

Table 1: DI (%) induced by reflection, in the specular

direction of observation.

Material

Incidence

angle (deg)

DIh DIv

Brick wall

10 0.31 % 0.48 %

20 0.66 % 0.19 %

30 0.67 % 0.65 %

40 0.72 % 0.75 %

Chip

wood

10 4.57 % 4.66 %

20 9.80 % 8.20 %

Stone and

concrete

facade

10 2.12 % 1.57 %

20 5.31 % 1.27 %

30 1.80 % 3.60 %

40 5.27 % 3.74 %

50 9.27 % 8.38 %

With this aim, median depolarisation indexes for

each material at all pair of angles of incidence and

observation are provided in table 2.

Table 2: Median DI (%) induced by reflection.

Material DIh DIv

Brick wall 23% 30%

Chip wood 18% 18.5%

Stone and concrete facade 4.5% 4%

Once the complete (180 degree) observation arc, not

just the specular angles, is introduced, the

depolarisation indexes grow, and differences

between incident polarisations also appear in the

brick wall case. The brick wall is the more non

isotropic material among the considered, as it

presents a clearly oriented structure, whereas the

chip wood panel and the stone and concrete facade

are the result of the solidification of a mass, which is

expected to present a more isotropic behaviour. The

high median values of depolarisation indexes

indicate that high depolarised waves could be

generated when several scatterers are present in an

environment, which is the case of indoor scenarios.

Furthermore, the transmission mechanism across

walls induces depolarisation. In this case, and

focused on the environments under test, the interest

is mainly the depolarisation by transmission across a

brick wall with normal incidence. The measured

value for DI at such situation was 9.4 %, considering

transmission with vertical polarisation.

4 PROCESSING

As measurements have been done following a wide

band scheme, information about the multipath

components can be obtained from the outcomes.

Knowing the multipath scheme, or the power-delay

profile (PDP) in the co-polar installation, it is

possible to compute a synthetic mirror (another

PDP) in the cross-polar domain.

Then, each couple of PDPs (co-polar and cross-

polar) could be the input of a diversity block, which

provides a new received signal with better

performance than just the co-polar one.

The following subsections contain the

computation of these cross-polar PDPs as well as the

results of applying different diversity techniques at

the reception end.

4.1 Effect of the Multipath in the Total

Received Power

The receiving antenna at each measuring location is

reached by the direct ray, which links the

transmitting and the receiving antennas following

the shortest path. But that antenna is also reached by

several contributions coming from paths generated

by reflections on the walls and, perhaps,

transmissions across some wall. The received power

from each contribution, associated to the time delay

relative to the direct ray arrival time, construct the

PDP. This profile defines the multipath environment

at each reception location.

Commonly, this PDP is shaped by the antenna

pattern. In this case, with azimuth omnidirectional

antennas, most of the PDP is due to the multipath,

and only a few part could be defined by the

elevation pattern of the antenna. Consequently, the

PDP used along this work could be assumed as the

product of the environments where the

measurements were performed.

4.2 Computation of Cross Polar

Received Power

The measurement outcomes are complex frequency

responses between 5.72 and 5.88 GHz, and they

have a shape as depicted in figures 6 and 7, which

contain the amplitude and the phase respectively.

If only waves following the direct path arrived

the receiving antenna, the amplitude of the

frequency response would be approximately flat. In

fact, it would be locally flat, but it would be smaller

at higher frequencies than at lower. The behaviour of

the phase would be expected to be linear. Evidently,

if we observe figures 6 and 7, the amplitude is not

flat and the phase is not linear, which indicates the

presence of multipath components.

IMPROVEMENT OF WIRELESS NETWORK PERFORMANCE BY POLARISATION DIVERSITY - Simulation from

Measurement Results at only One Polarisation

119

Frequency (GHz)

(

)

Figure 6: Example of the amplitude of the complex

frequency response.

Frequency (GHz)

(

)

Figure 7: Example of the phase of the complex frequency

response.

Applying an inverse fast Fourier transform to each

complex frequency response, this can be turned to

the time domain, with a resolution of 6.25 ns

between adjacent samples. Figure 8 depicts an

example.

At this time response, the different contributions

after one, two, three, or more reflections can be

identified, using an inverse ray tracing procedure.

Once the contributions have been classified in terms

of the number of reflections on the walls before they

reach the receiving antenna, the orthogonal

contributions at each delay can be computed.

Firstly, the path followed by each multipath

contribution has to be identified, and the walls that

generated each reflection or transmission mechanism

have to be categorised. Depending on the material

that constitutes the wall, the angle of incidence and

the frequency, a total depolarisation index (TDI) can

be obtained. Mean values, used along computation,

are summarised in table 3.



Delay (microseconds)

Figure 8: Example of time response.

The direct path contribution TDI needs a

supplementary comment. In LoS conditions, this

TDI depends on the depolarisation induced by both

antennas. In NLoS conditions, and additional

depolarisation due to the transmission across the

wall separating both rooms has to be considered.

When dealing with OLoS (furnished) environments,

we decided to take into account only the effect of the

antennas.

Table 3: Mean TDI (%) at different environments.

Environment Multipath contribution

Direct 1 ref 2 ref 3 ref

1, LoS 1.8 0.75 1.5 24.6

2, NLoS 9.4 15.8 24 32.2

3, LoS 1.8 0.75 1.5 24.6

4, LoS 1.8 8.2 16.4 24.6

5, OLoS 1.8 8.2 16.4 24.6

If the total power ariving the receiver antenna at

each time delay represents 100%, the co-polar

contribution would be (100-TDI)%. The collection

of co-polar contributions represent the measured

PDP, which is the basis to compute the cross-polar

PDP. Obviously, the cross-polar contributions at

each delay would be TDI %.

Consequently, the procedure to compute the

cross-polar contribution is:

1. Compute the co-polar PDP from the measured

frequency response.

2. Calculate the total power at each delay, based on

the correspondent TDI.

3. Obtain the cross-polar power at each delay.

4. Combine the collection of delays to obtain the

cross-polar PDP.

4.3 Polarisation Diversity at Reception

Once the co-polar and cross-polar PDPs have been

computed at each receiving location, different

polarisation diversity technique schemes have to be

WINSYS 2010 - International Conference on Wireless Information Networks and Systems

120

applied to each couple of data.

The three tested combination schemes are:

1. Sum, where the final signal could be the sum of

both input signals.

2. Mean, where the final signal is the average of

both inputs.

3. Switching, where the diversity device swicths

between both channels in order to get the

maximum at each instant.

After application of these diversity schemes at all

the receiving locations, the results are prepared to be

analysed.

5 RESULTS ANALYSIS

The application of diversity techniques provides

improvements in the received power signal. Among

the three considered schemes, the combination by

sum appears to perform better than the other pair,

and it offers enhancements as summarised in table 4.

Table 4: Estimation of the mean improvement by

polarisation diversity at different environments.

Environment Improvement (%)

1, LoS 17.89

2, NLoS 23.29

3, LoS 10.18

4, LoS 18.51

5, OLoS 21.23

Considering each environment separately, the

maximum improvement receiving location, the

minimum, the mean and the range of improvements

along the receiving path can be analysed.

Table 5 contains the data for the environments 1

and 2, which allows the comparison between LoS

and NLoS situations within similar rooms. The

improvement in terms of received power appears to

be larger when there is no line of sight between

transmitter and receiver, with a mean of 23.29%.

Analysing the maximum and minimum

improvements, the application of polarisation

diversity appears to be more advantageous in NLoS

conditions: the enhancement is, at some points, only

2% in LoS conditions.

Table 6 makes available the data to compare the

effect of polarisation diversity at reception as a

function of the size of the indoor environment, in

LoS conditions (environments 1, 3, and 4). The first

comment is that improvements are detected at every

room, but the performance of the networks installed

in large, and even long, rooms appears to be more

enhanced than in small rooms. Besides, the receiving

locations where less enhancement has been detected

present values under 4%.

Table 5: Comparison between LoS and NLoS.

Environme

Improvement (%)

Mean Max. Min. Range

1, LoS 17.89 32.28 2.00 30.28

2, NLoS 23.29 37.23 10.38 26.85

Table 6: Comparison among different size rooms.

Environment Improvement (%)

Mean Max. Min. Range

1, large 17.89 32.28 2.00 30.28

3, small 10.18 31.94 1.60 30.34

4, long 18.51 56.04 3.46 52.58

Table 7 contains the comparison between furnished

(OLoS) and empty (LoS) environments

performance. Both data come from the same room,

but changing the contents. In presence of furniture,

the polarisation diversity technique works better, as

it provides larger signal enhancements, even

although no obstacles block the line of sight between

transmitting and receiving antennas.

Table 7: Comparison between empty (LoS) and furnished

(OLoS) situations.

Environment Improvement (%)

Mean Max. Min. Range

4, LoS 18.51 56.04 3.46 52.58

5, OLoS 21.23 32.01 5.28 32.01

6 CONCLUSIONS

The improvement in the performance of wireless

networks by polarisation diversity has been

estimated from radio channel measurements in the

5.8 GHz band. Measurements at only one

polarisation were carried out. The cross-polar

responses of the channels have been computed from

these co-polar data, using depolarisation indexes.

The newness of the proposal is the analysis of

polarisation diversity at several scenarios based in

just one polarisation measurements.

Although three possible strategies for

implementing the polarisation diversity have been

taken into account, the main improvements were

provided by combination by sum scheme.

The presence of obstacles (when dealing with

obstructed line of sight situations), and mainly the

absence of line of sight (NLoS situations), leads to

IMPROVEMENT OF WIRELESS NETWORK PERFORMANCE BY POLARISATION DIVERSITY - Simulation from

Measurement Results at only One Polarisation

121

larger improvements in the performance provided by

the polarisation diversity. Even in LoS situations, the

improvements are noticeable. They are over 23% in

terms of power at NLoS environments, but the

minimum mean improvement has been estimated in

10%, which are interesting values for planning.

The proposal of using polarisation diversity in

reception could be of interest for network designers,

mainly in such environments where fading due to

multipath is especially deep.

ACKNOWLEDGEMENTS

This work has been supported by the Autonomic

Government of Galicia (Xunta de Galicia), Spain,

through Project PGIDIT 08MRU002400PR.

The authors would also like to acknowledge

Eduardo Cebrián Martínez de Lagos, Mr. Juan

Aguilera, and Mr. José Manuel Prado for their help

during the measurement campaigns, as well as Mr.

José Carlos Fernández Ribao for his help during the

data processing.

REFERENCES

Cuiñas, I., Martinez, D., Sanchez, M.G., Alejos, A.V.,

2007. “Modelling and measuring reflection due to flat

dielectric surfaces at 5.8 GHz”, IEEE Transactions on

Antennas and Propagation, vol. 55, no. 4, pp. 1139-

1147.

Cuiñas, I. Sanchez, M.G., Alejos, A.V., 2009.

“Depolarisation due to scattering on walls in the 5

GHz band”, IEEE Transactions on Antennas and

Propagation, vol. 57, no. 10, pp. 2934-2939.

Dietrich, C.B., Dietze, K., Nealy, J.R., Stutzman, W.L.,

2001. “Spatial, Polarization, and Pattern Diversity for

Wireless Handheld Terminals”, IEEE Transactions on

antennas and propagation, vol. 49, no. 9.

Dossi, L., Tartara, G., Tallone, F., April 1996. "Statistical

Analysis of Measured Impulse Response Functions of

2.0 GHz Indoor Radio Channels", IEEE Journal on

Selected Areas in Communications, vol. 14, no. 3,

pp.405-410.

Dutta-Roy, A., December 1999. "Networks for Homes",

IEEE Spectrum, pp.26-33.

Eklund, C., Marks, R.G., Stanwood, K.L., Wang, S., 2002.

“IEEE Standard 802.16: A technical overview of the

WirelessMAN air interfaces for broadband wireless

acces”, IEEE Communications Magazine, vol. 40, no.

6, pp.98-107.

IEEE Standard 802.16a-2003, IEEE Standard for Local

and metropolitan area networks — Part 16: Air

Interface for Fixed Broadband Wireless Access

Systems — Amendment 2: Medium Access Control

Modifications and Additional Physical Layer

Specifications for 2-11 GHz.

Turkmani, A. M. D., Arowojolu, A. A., Jefford, P. A.,

Kellett, C.J., 1995. “An Experimental Evaluation of

the Performance of Two-Branch Space and

Polarization Diversity Schemes at 1800 MHz”, IEEE

Transactions on Vehicular Technology, vol. 44, no. 2.

WINSYS 2010 - International Conference on Wireless Information Networks and Systems

122