LAUSR

With the ever-growing number of networked devices and a higher likelihood of having line-of-sight communication most of the time, location estimation of a blind node in the 5th generation Ultra-Dense Networks (UDNs) system has gained considerable attention in recent years. One of main factors for accurate location estimates in 5G UDN is node randomness. Several location methods and their performance analyses have been addressed for localization in 5G UDN. Although the distribution of reference nodes (RNs) is considered in the literature, the information on spatial node distribution is only used to evaluate the average performance and is not utilized in the location methods. In this paper, a Cramer–Rao lower bound (CRLB) and three location estimators including the iterative, closed-form, and hybrid algorithms are proposed for localization in 5G UDN with randomly distributed RNs. Both range measurements and prior information on spatial node distribution are utilized for the proposed location methods and CRLB. Moreover, some characteristics of the CRLB for 5G UDN localization are derived in this paper. Detail comparison between the proposed CRLB and the previous performance study on CRLB for 5G UDN is given. Theoretical analysis proves that the proposed CRLB for the case with randomly distributed RNs is smaller than the average CRLB for the case with fixed location RNs. The top and bottom bounds of the proposed CRLB in the cases with low and high signal noise ratios are also given. Performance evaluation shows that the proposed methods perform better than the conventional methods only based on range measurements and can asymptotically attain the CRLB.