Load Balancing in Downlink LTE Self- Optimizing Networks TD - PowerPoint PPT Presentation

FP7 ICT-SOCRATES Load Balancing in Downlink LTE Self- Optimizing Networks TD (10)10071 COST 2100, 10 th MCM Athens, Greece February 3 rd – 5 th NSN, Wroclaw, Poland NSN, Munich, Germany TUBS, Braunschweig, Germany IBBT, Ghent, Belgium

Outline 1. Introduction 2. Simulation metrics 3. Load balancing algorithm 4. Load estimation for the target eNodeB 5. Simulation scenarios 6. Simulation results 7. Conclusion WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 2/24

Introduction  Problem SeNB TeNB – Users concentrate in the area served by one cell – Unequal load distribution causes an overload – Users can not be served with required quality level due to lack of resources SeNB TeNB  Main Idea Received signal strength – Reallocate some users from the overloaded cell to less loaded neighbour cell(s) LB HO offset Hysteresis – Overloaded (SeNB) cell must find neighbour cell(s) (TeNB) which may accommodate additional load Distance – SeNB adjusts the HO offset of the TeNB and Overlaped area increase forces users to HO to the TeNB SeNB TeNB  Result – TeNB increases the overlapping area and takes over some users previously served by SeNB – LB operation sets free resources at SeNB – SeNB is able to serve remaining users with the required QoS WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 3/24

Virtual load Virtual cell load can be expressed as the sum of the required resources of all users u connected to cell c by connection function X(u) which gives the serving cell c for user u. 1 D ˆ u c ( ) M R SINR PRB u u | X ( u ) c  Du is the average data rate requirement per user u  R ( SINRu ) is the average throughput data rate per physical resource block (PRB) for user u  M PRB is the number of available PRBs ˆ c ˆ c  All users in a cell are satisfied as long as . In a cell with we will 1 1 1 have a fraction of satisfied users ˆ c WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 4/24

Throughput mapping DL throughput DL PRBs for 512 kb/s 5 9 8 Throughput bps/Hz 4 7 required PRBs 6 3 5 4 2 3 2 1 1 0 0 -6.5 -4.5 -2.5 -0.5 1.5 3.5 5.5 7.5 9.5 11.5 13.5 15.5 17.5 19.5 21.5 23.5 SINR [dB]  Throughput mapping bases on the concept of a truncated Shannon-Gap mapping curve ( ) log 1 Thr SINR SINR 2  The necessary number of PRBs for the required throughput Du and the transmission bandwidth of one PRB BW = 180 kHz can be obtained from the following equation D u N PRB ( ) Thr SINR BW WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 5/24

LB performance evaluation z metric  Unsatisfied users due to resources limitation z load – the total number of unsatisfied users in the whole network (which is the sum of unsatisfied users per cell, where the number of users in cell c is represented by Mc ) 1 max 0 , 1 z M load c ˆ c c z  Unsatisfied users due to power limitation (applies to UL transmission) power D 0 u for M max, u ( ) R SINR u z power D 1 u c u X u c for M max, u ( ) R SINR u – Where Mmax,u denotes the maximum number of PRBs that can be granted to user u WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 6/24

LB algorithm  The load level of all cells is permanently monitored  If the load level exceeds a certain threshold the load balancing is initiated 1. Sort all users by their SINR 2. Split the users in groups (according to the best suited target eNodeB) 3. Decide on the users to be handed over to other cells (based on the remaining cell load target for the source eNodeB) 4. Assure that the target eNodeBs are not overloaded after the load balancing activity 5. Modify the HO thresholds of the target eNodeBs 6. Send HO command to the selected users WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 7/24

List of target eNodeBs for load balancing  We propose a decentralised load balancing solution – all neighbouring eNodeBs are potential targets for load balancing – the decision depends on the reported load situation from all eNodeBs  An decision made by one individual eNodeB cannot take the larger network environment into account (e.g. the neighbour of the neighbouring eNodeB of the overloaded cell may be also be overloaded) – the central load balancing entity can report the cell load of the 2 nd neighbours  The central entity provides guidelines on LB priorities – cell 1 can obtain information about its neighbour cells over X2 – best target cells for cell #1 seem to be cells # 3 and # 7 3 3 – based on the overall load distribution available in the central 60% 60% SON entity we generate a priority list for the load balancing 4 4 2 2 85% 85% event: 1 1 – 4 5 5 7 7 – 5, 6 75% 75% 60% 60% 6 6 65% 65% – 3 – Cell # 7 is not on the list (LB to this cell is not allowed) – Cell #3 has the lowest priority due to the load situation in cell # 2  The central entity has been presented by NSN in the SA5 meeting in Vancouver (pseudo CR for TS 32.521) WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 8/24

Load estimation in the downlink  The load estimation at the TeNB has to be computed before the LB  It is based on a SINR estimation for the time after the LB – Required UE measurements RSRP – We assume that the UE does not change its position during the LB operation load estimation error I – interference from other eNB 1 a) SeNB TeNB S 1 S 2 0.9 0.8 0.7 0.6 cdf 0.5 0.4 I – interference from other eNB 0.3 SeNB TeNB b) S 1 0.2 S 2 0.1 0 0 10 20 30 40 50 60 70 80 90 100 [%]  Load estimation error measured for moving hotspot  Accuracy of load estimation could be higheer if users are fixed S 2 SINR TeNB S 1 S S 1 2 SINR SeNB WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 9/24

Load estimation in the uplink load estimation error in UL a) I – interference from UEs from other cells SeNB TeNB 1 0.9 S 1 S 2 0.8 0.7 0.6 cdf 0.5 0.4 SeNB I – interference from UEs from other cells TeNB b) 0.3 0.2 S 2 0.1 S 1 0 0 5 10 15 20 25 30 load estimation error [%]  Load estimation error measured for a moving hotspot  Accuracy of the load estimation could be SINR SINR S IoT higher if users would be static TeNB SeNB This equation is valid for users not  IoT cannot be predicted, we assumed limited in power small changes during the LB WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 10/24

Scenarios for evaluation studies Study 1: Impact of environment Scenario proposal: In a network setting with multiple cells consider different site to-site distances and cell types – Hexagonal network grid ISD 500m – Hexagonal network grid ISD 1700m – Nonregular network grid Study 2: Impact of service type Scenario proposal: Consider scenarios with a high and/or low rate of broadband service users. The load situation in the surrounding cells should also be varied for this study for the sake of service type or link direction. – Like VoIP service, UL/DL 30 kbps – Like video service, DL 512kbps, UL 256 kbps Study 3: Impact of user mobility Scenario proposal: The speed of the users should be varied to create scenarios with high/low user mobility – User mobility speed 3 km/h – User mobility speed 30 km/h Study 4: Impact of traffic load Scenario proposal: The amount of traffic load, load balance, size and shape of the overloaded area and location of the overloaded area should be varied in the scenarios. – Users move through cells – Users move along the cell borders WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 11/24

Network layouts and hotspot routes ISD = 500m; hotspot is moving from cell 10 to 20 ISD = 1700m; hotspot is moving along the cell borders Base Station Base Station 28 31 34 28 31 34 1000 shotspot route shotspot route 3000 30 33 36 30 33 36 800 29 32 35 29 32 35 600 2000 25 7 10 37 25 7 10 37 27 9 12 39 27 9 12 39 400 26 8 11 38 1000 26 8 11 38 200 22 4 1 13 40 22 4 1 13 40 Y [m] Y [m] 24 6 3 15 42 0 24 6 3 15 42 0 23 5 2 14 41 23 5 2 14 41 -200 55 19 16 43 55 19 16 43 -1000 -400 57 21 18 45 57 21 18 45 56 20 17 44 56 20 17 44 -2000 -600 52 49 46 52 49 46 -800 54 51 48 54 51 48 -3000 53 50 47 -1000 53 50 47 -4000 -3000 -2000 -1000 0 1000 2000 3000 4000 -1000 -500 0 500 1000 X [m] X [m] non-regular network; hotspot is moving from cell 27 to 13 1) Regular network layout, 2000 Base Station 2) Regular network layout, shotspot route hotspot moving through the 23 25 1500 26 22 hotspot moving along the cells 24 1000 cell borders 27 5 4 20 19 500 28 8 6 7 29 1 2 21 Y [m] 0 9 3 11 10 17 -500 30 16 12 13 14 -1000 18 32 31 15 3) Non regular network layout, -1500 34 35 hotspot moving through the 33 -2000 36 cells -2000 -1500 -1000 -500 0 500 1000 1500 X [m] WWW.FP7-SOCRATES.EU Dipl.-Ing. Thomas Jansen, TU Braunschweig, Institut für Nachrichtentechnik 12/24