Blazars are the most extreme class of active galactic nuclei (AGNs) characterized by a prominent jet pointing within a few degrees of our line of sight (Urry & Padovani 1995). They are known to be highly variable at gamma-ray wavelengths, as well as in the radio band, on timescales of days to months (von Montigny et al. 1995). Observationally they are divided into two main classes: BL Lac objects (BL Lacs) with an almost featureless spectrum and flatspectrum radio quasars (FSRQs) with strong broad emission lines in their spectrum (Urry & Padovani 1995).

The spectral energy distribution (SED) of blazars is characterized by two broad features. The first, peaking at lower energy, is generally explained in terms of synchrotron emission; the second feature, peaking at higher energies, is likely due to inverse Compton radiation (Sambruna, Maraschi & Urry 1996). Most of the observed radio to optical (and in some cases X-ray) emission from blazars is due to synchrotron radiation in the jet (Bregman et al. 1981; Urry & Mushotzky 1982; Impey & Neugebauer 1988; Marscher 1998).

Synchrotron emission is produced by relativistic electrons moving in a magnetic field (Rybicki & Lightman 1979). Inverse Compton photons originate from the interaction of the energetic electrons with seed photons. These seed photons could be produced by synchrotron emission, via synchrotron self-Compton (SSC) radiation (e.g. Konigl 1981; Marscher & Gear 1985), or they might originate from some external source - in this case it is external inverse Compton radiation (e.g. Sikora, Begelman & Rees 1994; Blandford & Levinson 1995).

The gamma-ray photons could originate through the SSC mechanism, and then one might expect a significant correlation between the gamma-ray and radio emission because of the same origin of the radio and gamma-ray photons. On the other hand, if there is no reliable proof for such a connection, that would support the independent origin of these emissions. Since the majority of AGNs identified with gamma-ray sources are also bright radio sources - about half of the 1400 gamma-ray sources from the first Fermi-LAT catalogue (Abdo et al. 2010b) were identified with radio sources - this provides the motivation to search for a correlation between radiation in the gamma-ray and radio bands.

By investigating such a correlation, one can study time delays between the different events in the gamma-ray and radio band light curves, the physical processes and the characteristics of the radiation in the AGN jets. The presence or absence of correlation also can help to determine more accurately the parameters of the models for the structure and processes in AGNs. The first dedicated studies of the gamma-ray-radio emission correlation in blazars were based on EGRET data (e.g. Stecker, Salamon & Malkan 1993; Padovani et al. 1993).

However, these results remain uncertain due to the use of observational data obtained non-simultaneously, and also because samples were flux limited (Muecke et al. 1997; Taylor et al. 2007). The search for significant correlation between gamma-ray and radio emission continued when Fermi-LAT telescope data became available. Using gamma-ray data from the EGRET and Fermi surveys, Ghirlanda et al. (2010) noted that the flux could change up to three times during the year. When average flux values are used for analysis, short-term variability (from one to several days or weeks) did not greatly affect the variability averaged over the year.

The F - Fr correlation has been studied by Ghirlanda et al. (2010) between the gamma-ray flux above 100 MeV (using 1FGL catalogue; Abdo et al. 2010c) and the 20 GHz flux density (using ATCA survey; Murphy et al. 2010). A statistically significant correlation (more than 3) was found both for the population of BL Lac and FSRQ sources. Also they considered selection effects (sensitivity limits for radio and gamma-ray telescopes) and the likelihood that some radio sources were not detected due to their variability in the gamma-ray band. Kovalev et al. (2009) also investigated the correlation between radio emission and gamma-ray flux (using Fermi-LAT data after the first 3 months of observations) for 135 AGNs.

The non-parametric Kendall tau test confirmed a positive correlation at a confidence level greater than 99.9 per cent between the gamma-ray flux (for the 100 MeV.1 GeV energy band) and radio flux density (measured by the Very Long Baseline Array at 15 GHz within several months of the Fermi-LAT observations). The same analysis for the second Fermi-LAT energy band, 1.100 GeV, also showed a positive correlation, but at a lower confidence level, 86 per cent. Ackermann et al. (2011) performed a detailed statistical analysis of the correlation between the radio and gamma-ray emission of AGNs detected by Fermi-LAT in its first year of operation. For the radio band, they used archival data at 8 GHz for 599 sources and concurrent measurements at 15 GHz for 199 sources, provided by the Owens Valley Radio Observatory monitoring programme (Richards et al. 2011).

One distinctive feature of that work was the study of not only the apparent, but also the intrinsic strength of the correlation, exploiting a new statistical method by Pavlidou et al. (2012). They found that the statistical significance of a positive correlation between the centimetre radio and the broad-band (E > 100 MeV) gamma-ray energy fluxes is very high for both FSRQs and BL Lac objects from their AGN sample.

Moreover, the correlation between high-frequency radio emission (at 37 GHz) and gamma-ray emission (100 MeV.100 GeV, Fermi-LAT data) for 249 northern AGNs was studied by Nieppola et al. (2011). They also found a significant correlation between both the flux densities and luminosities in the gamma-ray and radio bands and suggested that the gamma radiation is produced co-spatially with the 37 GHz emission, i.e. in the jet. On the basis of the work considered above, it can be concluded that homogeneous (derived from one instrument) and simultaneous observational data in the radio and gamma-ray bands are essential for detecting a possible correlation in radiation from blazars.

f0 (GHz)	f0 (GHz)	F (mJy beam-1)	BW (arcsec)
21.7	2.5	70	11
11.2	1.4	20	16
7.7	1.0	25	22
4.8	0.9	8	36
2.3	0.4	30	80