If you provide HDOP with precision in meters, as you pointed out in the comments, you are providing completely false fantasy value.
As an alternative answer related to the real question, I just wanted to provide the appropriate link to the GIS Stackexchange website, where this question was also asked: https://gis.stackexchange.com/questions/97774/how-can -i-reversed-horizontally -dilution-of-position to a-radius-of-68-confidence
In short: this is not trivial, and one HDOP is not enough for reliable accuracy. As user30184 commented, some simply use, for example, 3-5 meters as the accuracy of the device, and then multiply it by HDOP to get accuracy.
But in general, let me quote claudegps from this link also provided :
You can't. DOP is not an indicator of "error" or "accuracy." Poor DOP, for example, does not always mean poor accuracy. Moreover, “accuracy” should take into account not only DOP: imagine that you are indoors (very low signal, multi-channel equalizer), but with good DOP: you can have very poor accuracy even with good DOP ... So, your DOP based readings will be incorrect. Unfortunately, if you only have NMEA sentences, you usually do not have enough data to evaluate the accuracy (this can be done internally for the recipient, since there is much more information inside).
You can also see this link: https://www.gps-forums.com/threads/estimating-accuracy-from-raw-nmea-data.46273/ And finally, at this link :
6 sources of errors (and additional factors) affect the accuracy of GPS positions