Self-Management for Unified Heterogeneous Radio Access Networks Twelfth ISWCS International 2015 Symposium on Wireless Communication Systems Brussels, Belgium August 25, 2015
Dynamic Spectrum Allocation (DSA) in LTE macro networks Nils Dreyer Technical University Braunschweig
SEMAFOUR Vision NETWORK OPERATOR General Network- Oriented ObjecGves SERVICE PROVIDER Network Status related to General Network- Oriented ObjecGves UNIFIED SELF-MANAGEMENT SYSTEM Integrated SON Management Decision Support System Monitoring & Diagnosis Policy- Based SON Management OperaGonal SON CoordinaGon MulG- RAT / MulG- layer SON WLAN Mgmt. 2G / 3G SON LTE SON Physical Network Resources Physical Network Resources Physical Network Resources HETEROGENEOUS NETWORK 3 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Introduction 4 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Dynamic Spectrum Allocation (DSA) Mobile traffic demand is continuously increasing: From Q1 2014 to Q1 2015: data traffic increased by 55% Voice traffic remains constant In 2020: 3.7 Billion devices are expected to support LTE [1] Ericson Mobility Report June 2015 Scarce frequency spectrum that is not optimally used Between the Radio Access Technologies LTE frequency allocation in Germany 5 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Dynamic Spectrum Allocation (DSA) Dynamic Spectrum Allocation (DSA) Approach: Spectrum reallocation depending on the Network load With DSA spectrum can be used more optimally DSA can be used as intermediary step towards GSM spectrum refarming Traffic Intensity Map with Hot Spot in the center 6 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Solutions and Simulation Results 7 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Dynamic Spectrum Allocation (DSA) Dynamic Spectrum Allocation Intra-RAT DSA LTE <> LTE Change of Spectrum Usage for LTE cells in case of an overload situation Inter-RAT DSA LTE <> GSM Share the GSM spectrum with LTE spectrum by activating a second LTE carrier 8 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Dynamic Spectrum Allocation (DSA) Dynamic Spectrum Allocation Intra-RAT DSA LTE <> LTE Change of Spectrum Usage for LTE cells in case of an overload situation Inter-RAT DSA LTE <> GSM Share the GSM spectrum with LTE spectrum by activating a second LTE carrier 9 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Intra-RAT DSA Intra-RAT DSA Adapt the bandwidth of neighboring LTE cells to reduce interference Different DSA configurations were investigated: 20, 17.2, 14.4 or 10 MHz as bandwidth of surrounding cells Unequally distributed traffic request 10 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Intra-RAT DSA Intra-RAT DSA Key Results Significant gains only in spatially highly non-uniform traffic distributions / highly asymmetric traffic ratio between LTE cells Load decrease in Hotspot cell if bandwidth of neighboring cells is reduced Resource utilization gain of up to 25% for a realistic scenario Intra-RAT DSA leading to a reduction of utilized resources 11 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Dynamic Spectrum Allocation (DSA) Dynamic Spectrum Allocation Intra-RAT DSA LTE <> LTE Change of Spectrum Usage for LTE cells in case of an overload situation Inter-RAT DSA LTE <> GSM Share the GSM spectrum with LTE spectrum by activating a second LTE carrier 12 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Inter-RAT DSA Inter-RAT DSA Activate a second LTE carrier in the GSM frequency domain Find the optimal frequency position of the secondary carrier Interference increase Hardly interference from LTE to GSM (proven in simulation) Strong interference from GSM to part of the LTE carrier possible 13 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Inter-RAT DSA LTE interfered by GSM Optimal centre Frequency for 2 nd LTE carrier is based on 1. Its expected Coverage 2. Ranking based on requested traffic per pixel and its corresponding (worst case) SINR LTE carrier placed in GSM spectrum: SINR for different spectrum ranges of 200kHz 14 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Inter-RAT DSA Algorithm (Small Demo) Traffic Intensity Map LTE Bin Load 2pm Frequency LTE GSM Number of Carrier LTE Cell Load Overload 15 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Inter-RAT DSA Algorithm (Small Demo) Traffic Intensity Map LTE Bin Load 3pm Frequency LTE GSM Number of Carrier LTE Cell Load Overload 16 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Inter-RAT DSA Algorithm (Small Demo) Traffic Intensity Map LTE Bin Load 4pm Frequency LTE GSM Number of Carrier LTE Cell Load 17 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Inter-RAT DSA Inter-RAT DSA Results LTE can handle the traffic request in case of an hotspot formation Cell area throughput increase of 100% based on real network assumptions (gain highly depends on GSM network load) Additional throughput gain of up to 10% when GSM carrier is disabled It is assumed that the network can be reconfigured once an hour Network Load 18 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Conclusion 19 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Dynamic Spectrum Allocation (DSA) Intra-RAT DSA Bandwidth reconfiguration can be a measure to react on overloaded LTE cells DSA Gain highly depends on an asymmetric throughput request: 25% spectrum gain for a real scenario Inter-RAT DSA High throughput gain is possible for LTE cells: GSM is hardly interfered through LTE Can slowly close the gap displacing the GSM spectrum through a LTE spectrum GSM and LTE RAT must be coordinated: more complex when implemented in a real network 20 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Don t miss! DEMONSTRATOR on Dynamic Spectrum Allocation Presented during the next break! POSTERS on the SEMAFOUR use cases Presented during the next break 21 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015
Thank you for your attention www.fp7-semafour.eu 22 Workshop on Self-Management for Unified Heterogeneous Radio Access Networks, August 25, 2015