generally, co-dominant marker locations on the chromosome are known. 

An SSR that's co-dominant means for both AA and aa genotypes, you can amplify the DNA sequence of each (such as using PCR) and distinguish between them based on their band size (DNA length  that you amplified).  SSRs are DNA sequence repeats, so in co-dominant SSR marker, AA genotype has a certain DNA length when amplified, while aa has a different length. If you cross parent 1 (AA genotype for example) with parent 2 (aa genotype), then if the location of that genotype is segregating in Mendelian fashion, then you will expect the children to have 25% chance of AA, 50% chance of Aa or aA, and 25% chance of aa - since there is equal chance of you getting AA and aa, and you can tell them apart in your SSR based on the difference in band lengths.  Some SSRs do not behave in a co-dominant fashion - one reason why they are not co-dominant is because in these SSRs, the DNA length when you amplify AA and aa genotypes are the same or indistinguishable.  The genotypes may actually be different between two organisms with different alleles at these SSRs (they have different sequences or the length of sequences between two alleles are too similar to one another to tell apart), but since you cannot tell them apart based on your PCR amplification and analysis, these SSRs are not considered co-dominant.  Another reason why an SSR marker is not co-dominant is if you can amplify one allele (AA for example) but you cannot amplify the DNA of the other allele (aa).  This may be because this specific SSR DNA sequence doesn't exist or exist in a partial sequence in the aa genotype, so that you cannot amplify it using PCR, or you will need to adjust your PCR primers to make this co-dominant.  You can check this article for more explanations: https://www.researchgate.net/publication/226925645_An_introduction_to_markers_quantitative_trait_loci_QTL_mapping_and_marker-assisted_selection_for_crop_improvement_The_basic_concepts

Article An introduction to markers, quantitative trait loci (QTL) ma...

Main reason behind codominant is repeat based allelic variation 

Check the attached pic..for clear idea

I disagree with Andika's answer above.

Co-dominant markers can distinguish hets (Aa) from fixed (AA or aa) genotypes. If the two genotypes are not different at that locus (e.g. both are AA), then they are not polymorphic.

Hi Annie, not sure what you're disagreeing on, but to me your answers just elaborates further my points and introduced the concept of non-polymorphism.  My answer doesn't disagree at all with your comment on distinguishing hets from fixed genotypes. To further elaborate on your answer, SSRs may "appear" to be not polymorphic based on PCR band that you run on a gel (based on size) but in some cases they may actually be polymorphic if you sequence them (different sequences but similar lengths).  In this case, conventional SSR analysis using size difference does not tell the full story of the actual genotype - this was what I meant in my previous answer.

Hi Andika, so what should we categorize SSR markers as: (A) Dominant marker, (B) Co-dominant marker, (C) maybe both, depending on the experiment results?

Hi Annie, for the statement "If the two genotypes are not different at that locus (e.g. both are AA), then they are not polymorphic":

No polymorphic results (from PCR) doesn't mean that the two alleles are the same (or 'not different'). One of the alleles can be has 1-bp deletion (for example) and cause frame-shift mutation, but both alleles can be PCR-amplified with  'same' sized bands. In this case, they are different, but on the gel there is no polymorphism between them.

Hi Sandeep Kumar Soni,

Attached is a good (2016) SSR marker review paper for you.

Yes I agree with Yuan-Yeu, depending on the experimental results and the parental materials for testing the population of samples, we can determine whether a specific SSR is co-dominant or not.

yes i agree with the content provided