LTE Direct Overview. Sajith Balraj Qualcomm Research

Transcription

1 MAY CONTAIN U.S. AND INTERNATIONAL EXPORT CONTROLLED INFORMATION This technical data may be subject to U.S. and international export, re-export, or transfer ( export ) laws. Diversion contrary to U.S. and international law is strictly prohibited. LTE Direct Overview Sajith Balraj Qualcomm Research

2 Exploding Adoption of Mobile App Universe Source: Internet Trends Kleiner Perkins Caufield & Byers

3 The Rising Need For Ambient Awareness Intersecting market forces creating urgent industry appetite am bi ent a ware ness [am-bee-uhnt uh-wair-nis] noun the state or condition of one s mobile app continuously and passively monitoring for relevant value in one s proximity.

4 Obstacles To Scale Despite opportunity, still no mainstream consumer deployment GPS and network calls place onerous burden on ambient awareness Power drain #1 reason consumers unsubscribe from PBS Battery Continuous location tracking entails high consumer privacy cost Stubborn majority cite privacy costs as higher than perceived benefits Privacy Signal load and server processing hard to handle at high density May present ceiling to scale for most applications High Density

5 LTE Direct A New Wireless Sense Efficient, Proximate, Device-to-Device Communication Technology Continuous Awareness of Device Services Low-overhead discovery through synchronous operation Autonomous, Always ON discovery of what s around Relevant Discovery Across 1000s of services Power efficient, broadcast based design Range of several 100s of meters* LTE Direct No location tracking Potential app controls over cloud sharing Privacy Sensitive High spectral efficiency and reuse Distributed Interference management High Density of Direct Data Connections * For typical outdoor scenario with maximum path loss of 135dB

6 Devices Advertise their Services Sushi Palace- Today s Specials StarBucks on 4 th Ave Abe s Books: Book Signing in the next HOUR! Joe: Looking for new Extinct Warriors game I have tickets to sell for today s game Friend Finder Ticket Finder

7 All Services Identified; Relevance Determined Sushi Palace- Today s Specials StarBucks on 4 th Ave Abe s Books: Book Signing in the next HOUR! Joe: Looking for new Extinct Warriors game I have tickets to sell for today s game Friend Finder Sea Cafe Ticket Finder Seller: 4:00 pm game seat 45B

8 Communicating with the Desired Service Sushi Palace- Today s Specials StarBucks on 4 th Ave Abe s Books: Book Signing in the next HOUR! Joe: Looking for new Extinct Warriors game I have tickets to sell for today s game Ticket Finder Seller: 4:00 pm game seat 45B

9 What is LTE-Direct? Licensed spectrum framework Synchronous system Device transmits at nominal mobile power levels Proposed as a Rel 12 3GPP feature Study Item approved in 3GPP SA1 in Sept 11 Proposing RAN Study Item for Rel 12 LTE-Direct proposal includes two subfeatures Device to Device Peer Discovery (PD) Device to Device Data Communications (Comms) LTE-D as an integrated feature of the LTE Ecosystem

10 Expressions enable efficient discovery An Expression(s) is used by peers in discovery of proximate services, apps and context. Expression(s) are also used by proximate peers to establish direct communications.

11 Expression Basics Expressions at the application/service layer are referred to as Expression Names Names are mapped to 128 bits at the physical layer and referred to as Expression Code Expressions are classified as Public or Private Expressions based on the type of mapping

12 Highly Efficient Discovery Design LTE Direct discovery resources configured by the RAN Resources time-synced within the geography UEs transmit and receive Expressions within the discovery resources UE 1 UE 2 UE 3 LTE D Discovery LTE D Discovery WAN qualitative illustration; all UEs on the same carrier frequency Band displayed is Uplink of FDD or TDD Resource Assignment FDD Uplink or TDD resources are assigned for Discovery. Resource semi-statically allocated (typically <1% ) enb(s) assign part of discovery resources via SIB to authorized LTE D devices 2. Resource (SIB) 3.Discovery

13 Discovery Design: An Example 10 MHz FDD system: 44 PUSCH RBs used for discovery Sub-frame allocation are variable and controlled by enb Example: 64 sub-frames every discovery period of 20sec Number of direct discovery resources (DRIDs): 44*64 = sec Uplink D2D Peer Discovery D2D Peer Discovery WAN

14 Comparison Study Schemes are compared based on Power consumption of UEs OTA system resources used for comparable discovery Schemes are normalized with respect to discovery radius and latency Parameters Default Values Range Number of People* /cell Latency Cell radius Discovery radius Mobility Number of RBs available for discovery per sub-frame 20 sec 1000m 500m 10 km/hr 44 Fraction of mobile People* 20% 1%-100% Probability of interest* * * we study sensitivity with respect to these parameters ** probability that UEs are interested in other UEs

15 LTE Direct Discovery Significantly Outperforms OTT Discovery 150 Power consumption increase wrt legacy 10 3 Resource usage comparison Percentage increase Over the top discovery Direct discovery RBs per second Over the top discovery Direct discovery Number of UEs per cell Direct discovery performs better less than 7% reduction in stand-by time due to direct discovery, compared to up to 150% reduction in stand-by time due to over the top discovery Number of UEs per cell Direct discovery uses less resources Broadcast vs. Multiple Unicast messages 12x more resources used at 2000 UEs

16 WFD Message Based Discovery vs LTE Direct Broadcast Discovery Wi-Fi Direct Discovery Two step asynchronous message based discovery WFD Device exchanges device discovery request response messages followed by service discovery with every WFD in range LTE Direct Discovery Devices broadcast their services at physical layer using 128 bits Devices wakeup periodically synchronously to discover all devices within range

17 LTE Direct Enables Discovery of ~ 16X More Devices Message-based discovery limits Wi-Fi Direct discovery capacity Low duty cycle discovery in LTE Direct leads to significant power improvement vs Wi-Fi Direct LTE Direct WiFi Direct Profile 1 WiFi Direct Profile 2 Number of devices discovered Total Active Duration (s) Assumptions Times Square (outdoor dense deployment scenario) Ped A channel model ITU-1411 NLOS model; Wi-Fi Direct 10dB shadowing and 20dB wall penetration loss BW Assumptions: LTE Direct 10 MHz, Wi-Fi Direct 20 5 GHz 23dBm tx power, -5dB antenna gain, 7dB noise figure WiFi Direct Profile % Service Discovery WiFi Direct Profile 2 10 % Service Discovery

18 LTE Direct Communication LTE Direct Communication follows from discovery as a LTE system feature Network authorizes devices to communicate directly For system capacity optimization enbs Needed for Public Safety use case

19 LTE Direct Communication: Network Assisted Connection Setup Direct connection setup via network Operator Network MME MME Allow an operator to authorize and control the direct connection setup between UEs. Allow an operator to determine the user traffic routing between Direct and NW paths. enb 2. Direct connection context setup via Network enb 1. Direct discovery 3. Direct Alert & Direct radio bearer configuration 4. Direct communication

20 Spectrally Efficient D2D Communications Distributed link scheduling maximizes resource utilization Distributed link scheduling protocol enables maximal spatial reuse depending on the interference environment Orthogonalize in time if interference is too high Otherwise reuse traffic resources for maximum spectral efficiency Resources Reused Resources Orthogonalized

21 Low Priority Connections Yield To High Priority Connections in Proximity Transmitter yielding A transmitter will not transmit if this action will result in deteriorating (SIR) higher priority scheduled links Receiver yielding The receiver to which a transmitter is trying to communicate with will only accept the transmission over the link if its link has minimum quality (minimum SIR) A B High Priority Traffic Channel C D Low Priority Traffic Channel Transmitter may yield to A-B Receiver may yield due to A-B

22 Peer to Peer Traffic Optimization Capacity* per Km 2 (10 MHz) ,260 Mbps < 50m < 100m < 200m Link distance uniform up to x meters Lots of Spatial Re-use with Distributed Scheduling *Preliminary results 900MHz, ITU-R 1411 LOS model. Assumes fully loaded system.

23 research.qualcomm.com MAY CONTAIN U.S. AND INTERNATIONAL EXPORT CONTROLLED INFORMATION This technical data may be subject to U.S. and international export, re-export, or transfer ( export ) laws. Diversion contrary to U.S. and international law is strictly prohibited.