Microwave_Tower_AndrewIn the last “Back to Basics” blog post, Derren Oliver wrote about the ability of an antenna to focus the radio waves in the main beam, hence maximizing the signal level and reducing the interference to and from other links. Here, I’ll be explaining the microwave antenna’s ability to maintain radiated or received polarization purity between horizontally and vertically polarized signals. This is called cross-polar discrimination, or XPD.

In transmit mode, XPD is the proportion of signal that is transmitted in the orthogonal polarization to that which is required. In receive mode, it is the antenna’s ability to maintain the incident signal’s polarization purity.

For example, if a perfectly vertically polarized signal (containing no horizontal component) were incident upon a single polarized receive antenna, electrical and mechanical imperfections will introduce a small amount of ellipticity to the polarization of the signal. The signal can be thought of as having both vertical and horizontal components. The ratio of the consequential horizontal to vertical components is the XPD.

Here at CommScope we use the following formal definition of XPD: “Cross-Polarization Discrimination, in dB, is the difference between the peak of the co-polarized main beam, and the maximum cross-polarized signal over an angle twice the 3dB beamwidth of the co-polarized main beam.”

The angular region is included in the definition to account for any antenna movement caused by wind induced twist or sway.

We monitor these potential contributors to XPD in the factory by our Interport Isolation (IPI) measurement process. IPI has a relationship to XPD, and although it is not the only contributing factor, you can say that XPD will never be greater than IPI. As we have mentioned in previous blogs, to really understand and measure XPD you require a full range test, which CommScope performs in the design phase of all its antennas.

XPD is an important characteristic, particularly in dual-polarized systems, where cross-talk between polarizations can prevent the system’s quality objectives from being achieved. Radios can use cross-polar interference cancellers (XPICs) to isolate polarizations and compensate for any link or propagation induced coupling. However, good polarization purity from the antenna is important to allow the XPICs maximum flexibility to compensate for these dynamic variations.

Obviously, the subject of cross-polar discrimination is quite detailed. It is an important characteristic to understand when comparing microwave antennas. XPD can degrade network performance, meaning slower speeds, wasted spectrum and unhappy customers. We’ve got two more topics to introduce in the next blog post—return loss and voltage standing wave ratio (VSWR).

Any questions about XPD?

About the Author

Jim Syme

Jim Syme is product line manager for CommScope’s Microwave Systems division, responsible for business development in the Middle East, Africa and Asia-Pacific regions with additional global management responsibility for several major OEM customers. Jim began his 27-year career with Andrew Corporation in the microwave antenna design engineering group and is a regular participant at conferences on all matters related to microwave antenna systems. Prior to joining Andrew Corporation, Jim was a design engineer for the United Kingdom Atomic Energy Authority.

See all posts by this author


22 comments for "Back to Basics in Microwave Systems: Cross-Polar Discrimination"
OLATAYO OLALERE Monday, November 21, 2016 8:38 AM

Need to know what is the maximum and minimum value expected for a good link with XPD


Patrick Baffoe-korang Wednesday, January 03, 2018 8:58 AM

Can the XPD value of a given antenna change after it has been installed in the field?

Jim Syme Monday, January 08, 2018 4:32 PM

Hi Patrick

There is no reason for the XPD to degrade over time in an ideal world - the most likely scenario would be installation not being carried out as per the supplied bulletins and the antenna misaligning over time.

For example, if the antenna has been subjected to a SURVIVAL wind load occurrence then the antenna will almost certainly need realigned.

Environmental conditions can also degrade the performance (especially rain) but this should be cyclic as once back to clear, dry conditions the antenna should return to meet the RF spec.

The installation and interference from other new links can also impact the XPD.

I hope this answers your questions but, if not, please don't hesitate to contact us.

Best regards

Shyam Monday, January 15, 2018 3:36 PM

Pls any one tell me minimum and maximum xpd value, coupler details

Jim Syme Wednesday, January 17, 2018 2:58 PM

Hi Shyam

The minimum XPD of the antenna is what is specified in the antenna data sheets on www.commscope.com and this is measured under clear, dry conditions and assumes that the antenna has been installed and aligned correctly.

In an ideal world, the cross pol would be nulled in the azimuth and elevation planes but very small imperfections, that can't be avoided in antenna production, will start to fill these nulls (e.g. slight ellipticity in the circular waveguide, small misalignments during the assembly of feed components etc.).

Poor antenna installation (antennas not correctly aligned) can also degrade XPD as cross pol will increase in regions away from the antenna boresight.

Hopefully this answers your questions however please feel free to contact me if you have any further questions.

Best regards

ANUP JAIN Friday, April 27, 2018 7:48 PM


Here are some question regarding XPI against observations.


• XPI value is varying out of range in the day time and it is coming to Normal range in night time
• During these time , there is no variation in RSL level. But XPI varies out of range.
• Currently we are facing this issue in some MW Links.

Clarification / support required

• What may be the root cause of above mentioned issue
• Is it temperature sensitivity issue of any vendor radio? Because XPIC value is coming to normal range during night time.
• If the XPI value is coming to normal range, hope we should not take any action of OMT alignment / Re installation / cleaning ports etc.

Thanks & Regards
Anup Jain

Jim Syme Tuesday, May 01, 2018 9:35 AM

Hi Anup

The antenna should be resilient to temperature fluctuations and the fact that the RSL is constant would lead me to believe the antenna is OK.

I think the radio is a more likely candidate for the changes and I'd suggest you contact them to discuss this.

Best regards

Shafiq .I .S.Marie Sunday, July 01, 2018 11:05 AM

Some times we do XPD test with a values more than 30 db , but some times we have 30db in one terminal between H&V and 34 db
Is there any thing called balance in the reading I Nmean it should read the same values or the reading should be more than 30 db and thats all

Ahmed Soboh Thursday, September 20, 2018 7:36 AM

Hi Jim

I have some cases for 4+4 XPIC_SD configuration where I'm facing link XPD value fluctuation. how to judge if the XPD is healthy. also what about the ATPC feature enabling effect with SD links on XPD value?

Jim Syme Thursday, September 27, 2018 8:12 AM

Hi Ahmed.

I assume you are talking about a Space Diversity set up here.

Fluctuation in XPD value on a space diversity link is almost inevitable as different Tx/Rx Antennas become more/less dominant in the link. Also, the fade/multipath effect is different from each polarisation so this will also cause fluctuation in XPD level not necessarily in line with co-polar fluctuation.

We don’t think the ATPC (Auto Tx Power Control) should cause XPD variation.

Hopefully this answers your questions.

Hanumant Wednesday, November 07, 2018 10:00 AM


Does this XPD value varies as modulation increases?

We have set 2048QAM utmost peak modulation in Double XPIC hop. where consistenly getting Hop utilization ~100% .
Raised the TT to vendor TAC , they have suggested us to crosscheck for installation of MW links.

Thanks in advance

Jim Syme Wednesday, November 07, 2018 3:55 PM

Hi Hanumant

XPD of the antenna is fixed and independent of the modulation rate and should not change.

We agree with the feedback to do a cross check.

Best Regards

A.Ham Monday, January 28, 2019 3:40 AM

Hi Experts , Is there any way to get the noise floor value of a particular receiver to help in estimating the SNIR value over one hop?

Jim Syme Monday, January 28, 2019 1:50 PM

Hi there...I'm afraid this is more of a question for the manufacturer of your radio transceiver than CommScope who are supplying only the antenna and transmision line systems. Hopwfully they can help you out.

Best regards

Mohammad Abdullah Sunday, April 21, 2019 9:11 PM

Hello Gents,
Can any one plz tell me what woldbe the max value of XPD and Why?

Jim Syme Tuesday, April 23, 2019 1:59 PM

Hi Mohammad

CommScope specify the minimum XPD level in our data sheets (i.e. the worst case that could be achieved). XPD is the difference between the peak of the co-pol beam and the maximum xpol signal over an azimuth angle twice the 3dB beamwidth of the co-pol.

The XPD could be much greater - in theory, it would be in the noise floor of the measurement with perfect components but antenna/feed builds are not perfect due to manufacturing tolerances so this impacts the XPD.

I'm sorry but we couldn’t put a maximum value on XPD...normally the minimum value is of much more importance than the maximum value.

Hopefully this addresses your question in a satisfactory way.

Best Regards

Fernando Mendes Thursday, May 02, 2019 4:46 PM

Dear Sir:
Is the XPD value of the antenna under test dependent on the model of the measurement antenna, assuming both are linearly polarized? For example, would the XPD of the antenna under test be different between the 2 measurement scenarios below:
measurement 1: antenna under test is model X and the measurement antenna is also a model X
measurement 2: antenna under test is model X and the measurement antenna is a model Y (maybe a simple dipole)

Thank you.

hossam Saturday, August 17, 2019 10:58 PM

Hi ,

there is any relation between XPD and channel spacing , if we incrase the channel spaing we should increase the XPD value ?

Jim Syme Monday, August 19, 2019 12:46 PM

Hi Hossam

The XPD is given as a minimum spec over the entire operational band.

A Channel will take up a subset of the operational band and the XPD could be better but no worse than our spec.

The spacing will not influence the XPD requirement.

Best Regards

Ismael Thursday, October 03, 2019 5:28 AM

Hi, I have a question, how is a discrimination test performed on a microwave link? Is it true that there should be a difference of 30 db between polarities? and if there is not this difference is when there is data loss

Joe Depa Thursday, October 03, 2019 1:08 PM

Hi Ismael, Jim is out of the office until Monday. He will respond to your question when he returns.

Jim Syme Monday, October 07, 2019 9:37 AM

Hi Ismael.

The XPD of a link is very much dependent on the performance of the antenna being used in any link.

What is very important to remember however is that if you have two 30dB XPD antennas then you should not expect 30dB XPD over the link - you will probably only get around 27dB XPD.

There are several other factors that can have a major factor in the link XPD in the field including installation/alignment accuracy, interference from other antennas, multi-path depolarization (ground/object reflections can create an alternate path to the link) and measurement accuracy of the radio equipment.

If you'd like a more detailed conversation with our RF Engineering team then I can put you in contact with them.

I hope this helps.

Best regards

Add Your Comment

Please submit your comment using the form below