1. Overview
With the development of the network, in view of the advantages of 900M coverage, in order to enhance the depth of coverage and improve competition, the current FDD 900M has been accelerated deployment, but it also brings interference problems. At present, interference troubleshooting has become a large number of problems in the deployment process of FDD 900M. Due to the difficulty of interference investigation and the long period required for investigation, how to avoid FDD 900M interference through optimization means is an urgent problem to be solved at this stage.
Impact of FDD 900M interference:
(1) Index degradation, high interference brings high packet loss and high call drop.
(2) User complaints are on the rise. Since voice services are more sensitive than data services, user perception in areas with strong interference decreases significantly, and complaints also increase.
2. Optimization principles
A provincial company conducted detailed analysis and innovatively proposed a "three-dimensional and six-category" interference analysis model. First, according to the location of the interfered frequency of the FDD 900, it is divided into two types: control channel interference and traffic channel interference. Secondly, based on the size of the interference and the number of interfered RBs, the interfering cells of the service channel are subdivided into four sub-categories in "two dimensions and four images", and thus the "three dimensions and six categories" interference avoidance classification is proposed, laying the foundation for the subsequent precise deployment of avoidance technologies .
3. Classification of interference cells
(1) Business channel
If the number of the interfered RB is between 3 and 22, the cell is defined as traffic channel interference, and the four-quadrant cell is classified according to the degree of interference and the number of interfering RBs, where the RB threshold is 20% (5), and the RIP threshold is - 100dbm.
Class A: Interference with RB exceeds 20%, RIP>-100dbm;
Class B: Interference with RB exceeds 20%, RIP<-100dbm;
Category C: RB less than 20%, RIP>-100dbm;
Category D: RB less than 20%, RIP<-100;
(2) Control channel
If the number of the interfered RB is 0~1 or 23, 24, and RIP (interference power)>-100dbm, then the cell is designated as a cell with strong control channel interference. According to whether there is strong interference with the PUSCH channel next to the PUCCH, it is divided into two categories:
Class A: Strong interference from PUSCH next to PUCCH, RIP>-100dbm;
Class B: PUSCH next to PUCCH has no or weak interference, RIP<-100dbm;
4. Classification optimization strategy
For different types of interference, take different measures, as follows:
Network-wide FDD 900 Cell Interference Classification
A cell may have control channel interference and traffic channel interference at the same time. For example, there may be strong control channel interference and traffic channel_type A interference at the same time. Since the control channel has a greater impact, this kind of situation is preferentially attributed to control channel interference. Therefore, statistically, the number of interfering cells of traffic channel_A and traffic channel_B will be less than that of control channel interference. There are 1,820 FDD900 cells in the existing network of a certain city, among which there are 226 cells with control channel and traffic channel_type A interference, accounting for 12.42%.
Optimize implementation
1. PUCCH Blanking
(1) Principle introduction
In the uplink channel, the PUCCH occupies 2 RBs at both ends, and some manufacturers may reserve 3 RBs. When the RB interference at both ends of the system is obviously stronger than that of other RBs, the available uplink RBs can be indented from both ends to the middle, that is, the RBs with strong interference at both ends are discarded, and the two ends of the PUCCH channel resources are symmetrically indented to the middle by PUCCH_Blanking_NUM/2 RBs , the available resources of the PUCCH channel remain unchanged, and the PUSCH channel resources are reduced by PUCCH_Blanking_NUM RBs. If PRACH is set to self-adaptive, it will move synchronously. If self-adaptive is not set, the starting position of PRACH needs to be manually set.
(2) Network management settings
The parameter path on the network management is E-UTRAN FDD cell –> uplink and downlink physical channel configuration –> PUCCH Blanking function switch, number of RBs lost by PUCCH Blanking.
(3) Field test
The PUCCH frequency domain occupies both ends of the bandwidth, followed by the PRACH. In the case of self-adaptive PRACH frequency domain position in the background, before the PUCCH Blanking function is enabled, PRACH occupies 6 RBs starting from RB16 (as shown in the left figure); after the function is enabled, the number of RBs dropped by PUCCH Blanking is set to 2, and PRACH occupies RB15 The first 6 RBs (as shown on the right). It can be seen that both ends of the PUCCH move toward the bandwidth center by 1 RB.
(4) Index analysis
Select 44 eligible cells for optimization. From the perspective of sky-level indicators, there are large fluctuations. The downlink packet loss rate has a tendency to improve, but there is no obvious improvement in the uplink. Since the PUCCH is used to feedback downlink data packets, it mainly affects the downlink packet loss rate. Overall, the downlink packet loss rate drops from 0.314 to 0.243 after optimization, which is a significant improvement.
2 Raising the minimum access level + VoLTE switching back to TDD pilot
(1) Principle introduction
VoLTE voice packet loss due to wireless link quality will seriously affect user experience. Therefore, when the voice quality of the serving cell where the user is located is worse than the threshold, the VoLTE voice perception experience can be improved by triggering inter-frequency switching in time. When there is a large amount of interference, the uplink and downlink scheduling capabilities are limited and voice quality is affected. However, at this time, the coverage-based inter-frequency triggering conditions may not be met. In this scenario, inter-frequency handover based on voice quality is required. For example, in an area with strong FDD900 interference, users are migrated back to TDD through inter-frequency handover based on voice quality, and at the same time, switch strategies are set on the TDD side to distinguish between QCIs to prevent users from ping-pong switching between TDD and FDD900.
The overall strategy is as follows:
| interoperability |
FDD side |
Surrounding and this station TDD->FDD |
| re-election |
The minimum access level is changed to -100 |
For the reselection of the 948.3 frequency point, the minimum access level is changed to -100, and the XLOW is changed to -100 |
| to switch |
Enable quality-based inter-frequency handover for voice services (A4 threshold is set to -110) |
Distinguish between QCI setting switching strategies Voice service: A5 threshold 1=-120, A5 threshold 2=-105; Data service: A5 threshold 1=-106, A5 threshold 2=-105; |
(2) Network management settings
i) FDD side: raise the minimum access level in idle state
ii) FDD side: quality-based handover
The quality-based switching mechanism of ZTE equipment is triggered by downlink MCS and uplink SINR. When one of the "uplink SINR threshold for poor voice quality" or "downlink MCS threshold for poor voice quality" is met, an inter-frequency handover based on voice quality is triggered. Before the quality-based measurement report is reported, if both the "uplink SINR threshold with good voice quality" and "downlink MCS threshold with good voice quality" meet the conditions, the base station will not initiate inter-frequency handover based on voice quality. The voice quality-based switching defaults to the 140 configuration number A4 event, which cannot distinguish the measurement configuration numbers with different frequency point configurations.
iii) TDD side: reselection of surrounding and local TDD sites -> FDD
For the reselection of the 948.3 frequency point, the minimum access level is changed to -100, and the XLOW interface value is changed to 0.
iv) TDD side: Differentiate the QCI setting switching strategy
Step 1: Refer to the perQCI inter-frequency measurement configuration index group ID.
Step 2: A5 threshold 1 for service QCI number 1 and service QCI number 2 is changed to -120, A5 event RSRP threshold 2 (dBm) is changed to -105). Remarks: It is a modification for PERQCI measurement configuration index group ID 2.
Step 3: Determine the handover strategy from the current TDD cell to FDD900
PUCCH Blanking and QCI1 NI frequency selection can improve network quality, but the effect of frequency selection for all services (without distinguishing voice data) is not obvious. Quality-based inter-frequency handover (FDD to TDD switching) is mainly an interference avoidance method. When user perception is degraded, users will be migrated out in time. This will not improve the network itself, but it will improve user perception.
END