Case study on a e MBMS implementation on

Case study on a e. MBMS implementation on a real mobile network 08/10/2013 Sebastien GRIMOUD grimoud@anfr. fr 08/10/2013 Case study on a comparison between e. MBMS and DVB-T 2 on a high tower high power network 1

To. C • Problem statement • Simulation framework • Evaluation metric • Results and interpretation 08/10/2013 Case study on a comparison between e. MBMS and DVB-T 2 on a high tower high power network 2

Problem statement • Mobile networks are sometimes foreseen for broadcasting linear TV • Mobile networks count numerous sites and are therefore expensive • But network infrastructure is already rolled-out for mobile communications • Question is therefore to know whether the mobile infrastructure is suitable for broadcasting linear TV. • The contribution focus on the air interface. It does not prejudge of the capability of the backhaul. • Several reception modes are investigated. Suitability of mobile technologies for linear-tv broadcasting targetting different reception mode is a key question for evaluating the benefits of network convergence. 08/10/2013 Case study on a comparison between e. MBMS and DVB-T 2 on a high tower high power network 3

Simulation framework • 1150 NBs, corresponding to some real network sites. • Restricted service area. • H-Omnidirectional 4° downtilt (set by mobile service req. ), 2500 W ERP. • Fixed rooftop antenna (BT 419), mobile outdoor and handheld indoor reception investigated. • Propagation FD 15000 km. • Cyclic Prefix is 16µs for e. MBMS (max available value). • A C/I margin is applied to meet a 70% of location service area criteria. • The network is assumed to be a national SFN. • SFN interference are derived according to EBU formula. Antenna configuration is constrained by the mobile service requirements. 08/10/2013 Case study on a comparison between e. MBMS and DVB-T 2 on a high tower high power network 4

Evaluation metric • Broadcast is different from unicast transmission because: • retransmission mechanisms are not available; • MCS must be the same for the whole network. • Therefore, service is lost when the SINR is below the PR threshold. • The result is given in term percentage of the population located in the service area for different PR. A higher PR means a more efficient MCS and therefore a higher throughtput. • The service area is the set of pixels in the coverage area where the probability of service is higher than 70%. (SINR above threshold). • The coverage area are the set of pixels where the probability of reception in the presence of noise only is higher than 95%. The SFN interference reduces the coverage area. In order to mitigate the effect of SFN interference, robust MCS must be used. However this decreases the spectral efficiency. 08/10/2013 Case study on a comparison between e. MBMS and DVB-T 2 on a high tower high power network 5

Results and interpretation PR 5 d. B Fixed Mobile Coverage Service area (% pop) 99. 9 99 Indoor (area restricted to building raster) 89. 29 85 85 87 e. MBMS does not appear to be ready for ubiquitous service, especially for rooftop and mobile reception modes. 08/10/2013 Case study on a comparison between e. MBMS and DVB-T 2 on a high tower high power network 6