Are there any other methods to correct variation after background variation, normalization and dye effect correction in data handling of a microarray?

I have 2 slides. each of slides contains 2 channels scanned by Cy5 and Cy3 fluorescence excitation. I have handled data by background effect correction by aligning the mean of each channel in 4 slides. After that, I made scale normalization and dye effect correction to remove many variations. But, what are the next steps that are able to help me to remove more variation of microarray chip?

Can you give me any suggestion and useful keywords?