Genomic DNA contamination: The band near the well that resembles genomic DNA suggests that there might be genomic DNA contamination in your RNA samples. This could occur if the RNA isolation process didn't adequately remove genomic DNA. To mitigate this, you could try using DNase treatment during RNA extraction to degrade any remaining genomic DNA.
Low molecular weight RNA: one band looks bright at the end of the gel. It is difficult to predict as such without comparing it with the ladder.
RNA fragmentation: The presence of smaller fragments in the gel suggests some RNA fragmentation (as far as I know, saliva samples contain a high amount of RNAse, so please cross-check).
To check the integrity of RNA, you should run RNA sample on denaturing agarose gel. Have you used denaturing agarose gel?
If you have used non-denaturing gel, then the results will be difficult to interpret because the secondary structure of RNA alters its migration pattern in non-denaturing gel so that it will not migrate according to its true size. Moreover, you will also observe multiple bands representing different structures of a single RNA species.
Please note that non-denaturing gel will not give you sharp bands as they would be seen in denaturing gel, a clear 28S and 18S rRNA bands seen for eukaryotic sample.
Please find attached below the protocol for preparing denaturing agarose gel for RNA samples, if you need one.