ok, since every dataset of mine contains 60 values i count now the bytes of each dataset and write them into a separate file. this way i can access the separate datasets directly by seeking in the data file using the byte information out of the additional file.
