I would assume that there are no house-keeping genes in bacteria. Transcription in bacteria is influenced by competition of sigma factors for the RNA polymerase, so even genes under control of sigma-A are not transcribed evenly under all conditions. Throw in things like the stringent response and the dependence of mRNA stability on ribosome availability and the concept of house-keeping genes goes down the drain.

Still, depending on your experimental setup, there might be markers that work. But that depends on the conditions you want  to use.

I would advise to look at the MLST genes, they are usually selected for their high conservation troughout the species: http://mlst.warwick.ac.uk/mlst/dbs/Ecoli.

If their expression is even, is another matter and should be investigated before proceding.

Since E. coli has fast cell division cycles (duplication rate) and the bacteria do not grow in synchrony,  I doubt that it is possible to define any gene that fits the text-book definition of housekeeping gene in eukaryotes. If you are examining your cells in log-growth phase, you might want to try to use genes for the metabolism of the carbohydrate source you are using as "reference" genes, as on average, they should be uniformly expressed at a high levels under such circumstances.