3G Evolution An Overview of LTE for Mobile BroadBand James Kempf Research Area Packet Technologies Ericsson Research May 11 2010 with thanks to Dr Stefan Parkvall Ericsson Research Mobile Broadband A P S H G 3 o b r Tu LTE More Than 400 HSPA Devices 203 HSPA phones media players camera 50 161 PC with embedded HSPA PC cards USB modems 40 39 wireless routers 10 2010 3G iPad Commercially launched as of August 2007 Outline Introducing Ericsson Research Silicon Valley Basic Principles 3GPP Long Term Evolution LTE LTE Advanced Standardization Summary Introducing Ericsson Research Silicon Valley Ericsson Research Silicon Valley Part of our global research organization Silicon Valley Montreal Sweden Finland Germany Hungary Spain RTP Ericsson Research ER Silicon Valley Japan Italy China 600 people GSM 3G and LTE technologies were invented here Leading IP competence in IETF Delivers concepts and pre commercial prototypes Files over 50 of all Ericsson patents Packet Networking Open Application environment Radio Access Tech Signal processing Led by Jan S derstr m V P Research USA Research Focus Silicon Valley Applications Mobile Internet IP Routers Networks Service Layer Standard Services and IMS Multi Access Edge Wireline Access Wireless Access Internet Radio Access H1 P P1 PK Precoding Matrices H2 HK G1 Receiver Matrices G diag G1 GK G2 GK Energy heat Next Generation Internet Botnets and Spoofing Now an industry 2007 Signalling contr ol Billion subscr 2005 3 Mobile Fixed 2 1 2008 Mobile Nomadic Fixed BB 2013 Protocol X Fully deployed 10 years Ex IPv6 RSVP HIP New Technologies 50 Deployment Security Lagging Mobility Daily Spam 100 Billions IaaS PaaS Google Apps Gmail Google AppEngine Google Apps Gmail Google AppEngine Microsoft Azure MobileMe Microsoft Azure MobileMe Sun Project Caroline Windows Live Sun Project Caroline Windows Live Force com SalesForce Force com SalesForce Rollbase Gliffy Rollbase Gliffy Amazon EC2 Amazon EC2 GoGrid GoGrid Joyent Joyent Mosso Mosso FlexiScale FlexiScale NW Virtualization substra te link substrat e router virtual router slice virtual endsystem Mobile Broadband LTE Ev Tra ffi c LTE B width cost Trends Issues Cloud computing SaaS s Revenue Relative Price Band witdh pro st cess or ag ing e HSPA Ev 3G HSPA Time A new network paradigm without a clean slate deployment Basic Principles Trend Data is overtaking Voice Data is overtaking voice but previous cellular systems designed primarily for voice 70 WCDMA HSPA world average 400 3 350 60 300 Packet 50 200 RNCs GBit RNC h 250 40 30 150 20 100 Voice 10 50 0 0 sep 06 okt 06 nov 06 dec 06 jan 07 feb 07 mar 07 apr 07 maj 07 jun 07 jul 07 aug 07 sep 07 okt 07 nov 07 dec 07 Sum of Meas RNCs Sum of Total traffic GBit RNC h Sum of Speech traffic GBit RNC h Sum of Packet DCH HS traffic GBit RNC h Sum of DCH Packet traffic GBit RNC h Sum of HS Packet traffic GBit RNC h Sum of CS64 traffic GBit RNC h Sum of Others traffic GBit RNC h Ct r l Shi f t E Source NetQB History 30 kHz NMT AMPS Analog speech Research and standardization 200 kHz GSM Digital speech and lowlow rate data Research and standardization WCDMA HSPA 5 MHz HighHigh speed data up to 20 Mbps Research and standardization LTE HighHigh speed data up to 300 Mbps 1980 1990 2000 2010 20 MHz HSPA and LTE Mobile Broadband HSPA High Speed Packet Access Evolution of 3G WCDMA Gradually improved performance at a low additional cost Data rates up to 40 Mbit s in 5 MHz R99 Rel4 WCDMA Rel5 Rel6 HSDPA HSPA Rel7 Rel8 Rel10 HSPA evolution 4G LTE LTE Advanced LTE Long Term Evolution Significantly improved performance in a wide range of spectrum allocations Data rates up to 300 Mbit s in 20 MHz First step towards IMT Advanced 4G Extending broadband data networking from wired to wireless Wireless vs Wireline Many aspects are similar but there are some fundamental differences Wireline Cable No spectrum limitation over provisioning Relatively static channels No fading Congestion lost packets No mobility Wireless No cable Spectrum is scare Radio resource management Time varying radio channel Fast fading Fading lost packets Mobility Radio Channel Variations Path loss Given by Tx to Rx distance Slow Transmitted power PTx received power PRx PTx Log normal fading Due to random variations in terrain large scale Received signal strength in dB given by normal distribution Random variations in environment Often modeled by a Rayleigh distribution Fast Fast fading Radio Channel Variations Transmitted signal reflected in numerous objects multiple delayed signal copies received Large delay difference components can be processed separately Small delay difference components add constructively or destructively Large number of components central limit theorem Gaussian distributed amplitude Rayleigh distributed power Rayleigh fading fast fading Multi path fading Power fre qu en cy Radio channels rapidly varying signal quality ce spa Traffic Variations Traditional voice services Low 10 kbit s data rate Fairly constant during the call circuit switched ok Packet data services Behavior depends on type of service Typically rapidly and randomly varying rate requirements all or nothing resource requirement packet switched NW Packet data systems rapidly varying data rates TCP Basics TCP Internet s end to end transport layer protocol non real time Appl Appl TCP TCP IP IP IP IP IP IP Link Layer Link Link Layer Layer Link Link Layer Layer Link Layer Phy Layer Phy Phy Layer Layer Phy Phy Layer Layer Phy Layer Router Router Server Client Main responsibilities of TCP provide reliable data transport avoid congestion in the network Interaction with wireless links requires attention TCP Basics Error recovery and congestion control are intertwined lost packets are used as congestion signal by TCP radio link errors should be hidden from TCP Lost packets timeout slow start TCP congestion management RTT 30 ms 60000 RTT 60 ms 50000 Congestion avoidance phase Congestion avoidance phase 40000 Seq No bytes Window not yet ACKed packets in transmission Phase 1 Slow start Increase window by one on each received ACK window grows exponentially Phase 2 Congestion avoidance Increse window by 1 window size on each ACK window grows linearly 30000 20000 Slow start phase Slow start phase 10000 0 0 100 200 300 400 Time ms 500 600 700 800 TCP Basics TCP performance determined by data rate and latency High data rate alone not sufficient need low latency as well Delay bandwidth product Length of the pipe Latency Width of the pipe Data Rate