A.3.2. Synchronization Recovery
As previously discussed, OTD measurements can be easily used for geo-location once the synchronization reference of the different sites is known. However, such synchronization is not often available in a real network, so that it is necessary to develop a practical methodology to estimate it. The presented algorithm follows the guidelines suggested in [9] and consists of an iterative approach that can be divided into the following steps:
Estimation of an initial synchronization reference for each site, which can be obtained by assuming that the initial location of the UE is the position of the sector with strongest signal.
Selection of a set of very well-conditioned events (reporting a high number of different sites, without multipath propagation, etc.) that are expected to speed up the convergence to the actual synchronization reference.
Geo-location of those selected events using Equation A.5 with the previously estimated synchronization reference.
Update of the synchronization reference according to the latest geo-location.
Repetition of the procedure until the synchronization reference converges.
A.3.3. Filtering of Events
There are multiple sources that can distort the OTD measurements of certain events such as strong multipath propagation. Those events should be clearly detected beforehand and not geo- located because they can worsen the estimated location. On the other hand, this kind of analysis could be also beneficial for the operator to identify certain configuration problems in the network. The main filtering criteria are listed below:
Wrong Scrambling Code (SC) assignment: decoding the proper site based on the reported SC, so that the site coordinates are correct, is essential for an accurate geo-location. The probability of failing on the SC assignment depends on the SC planning and the area in which the MRs are collected. For instance, this probability increases in very dense areas.
Multipath propagation: due to building and terrain reflections the signal can arrive with different delays, degrading the reported OTD measurements.
Handover effects: according to 3GPP specifications [6], OTDs reporting the serving sector must be a multiple of 256 chips if there is good synchronization. However, after a handover process some time is needed for recovering such synchronization to the new serving sector.
Events collected during this period may not be reliable.
Do'stlaringiz bilan baham: |