Free-Space Optical Communications SNC May 8, 2008 Thomas Little tdcl@bu.edu 617-353-9877 In collaboration with J. Carruthers and the ERC Team (BU--UNM--RPI)
Illumination and Communication Three key messages 1. Solid State Lighting is coming It brings an opportunity to embed networking 2. Pervasive Computing is upon us Mobile wireless devices, embedded networked sensors, ID, WiFi hotspots Never far from the Internet (a) indoors and (b) outdoors 3. LED-based communication and networking has important advantages: Bandwidth, bandwidth density Privacy--security Ubiquity if piggybacked on lighting Unregulated spectrum HF2 Stick, www.sylvania.com Incredible opportunity to affect new medium for ubiquitous network access 2
Free-Space Optical Networking: Where it Matters Deliver HD video to individual seats Airbus holds > 500 people; HD requires 13 Mb/s; short range personal lighting/communication for channel isolation; copper is heavy. High bandwidth density (>10 Mb/m 3 ) Localized communication between vehicles Emerging safety-oriented technology: active braking, traffic monitoring; warning message propagation. Directional transmission, P < 1%, < 100ms latency Indoor localization Finding roaming patients and doctors in a hospital; techniques can be problematic; lights can be uniquely modulated with ID; tagging bats; security in downlink channel. Data trickle. Providing opportunistic mobile access Hotspots wherever there is illumination. Ubiquity. Moving vehicles. Internet access Mesh networks From Airbus (www.airbus.com) Courtesy of Thomas Kunz 3
(802.11) vs. Free-Space Visible Spectrum (LED) Attribute @ 2.4 GHz LED Advantage Security/privacy Promiscuous, known limitations, penetrates walls Transmission range Attribute typically < 100m @ 2.4 GHz Opacity, LOS, locality LED Limited by container, typically < Advantage Point-to-point Interference Aesthetics Security/privacy Available capacity Cost of spectrum Available BW capacity Safety -- intensity Data rates at 1 m Very high when available Seldom supported Cost of additional spectrum Interference Relatively safe at 2.4GHz. Problems at higher frequ. (e.g., 60GHz) interference from self, multipath, ISM sources Additional item in room Promiscuous, penetrates walls Limited by multiaccess, degrades from peak (< 50 Mb/s) Limited by contention, degrades from peak Very high when available Self, multipath, ISM sources Near peak w/o contention (< 50Mb/s). Rapidly contention limited Opacity, LOS, locality LOS permits spatial reuse. Limited by noise None (yet) LOS permits spatial Common (e.g., IR, laser) reuse Problems with narrow beams in visible spectrum None (yet) Visible natural and man made light Built into lighting Visible natural and man Near peak w/o contention. Not limited by contention made light with multiple transceivers. Data rates at 100 m Aesthetics Contention and noise/interference limited typically < 10 Mb/s Additional item in room Noise limited in diffuse case, LOS rates sustainable with Built intensity into lighting Data rates at 1 km Zero Installation cost Installation cost Moderate -- separate item Symmetry Multipath Typically symmetric Moderate - separate item Destructive interference Low - piggybacked LOS rates Mb/s to Gb/s (laser) Low -- piggybacked Noise at receiver Strategic positioning of lighting/nic Multipath Picocell deployment Point-to-multipoint Mature Destructive interference Difficult to control trans. range Noise Directional, at receiver controllable intensity Limited pilots Pico cell deploymt Estimated comp. cost Antenna arrays Est. comp. cost < $20 Performance under contention Perf. w/ contention Antenna alignment Acquisition delay only Path redundancy Volume coverage Layer interaction Irregular coveratge pattern, difficult to control transmission range Specialty item, wired to room Decreases with transmission footprint Irregular, but typically complete TPC bridges NET and PHY < $20 Decreases with transmission footprint Achieved with multiple access points Lensed, directional, intensity modulation < $2 (based on IrDA) Built into lighting. < $2 (based on IrDA) Limited by no. sources, not a function of footprint Limited by no. sources, not a function of footprint Acquisition delay unless diffuse. Issue w/ mobility Achieved with multiple LEDs LOS and reflective, not complete under LOS only. TPC and PAT bridge NET and PHY Path redundancy Achieved with multiple access points Achieved with multiple LEDs 4
Research Focus--Benefits 1. An opportunity to embed networking Leverage power line modems, WiFi, lower energy use Networking where there is illumination 2. Ubiquitous communication is an enabler Mobile wireless devices, embedded networked sensors, ID, WiFi hotspots, transportation Never far from the Internet (a) indoors and (b) outdoors Better data and control from/to the physical world 3. LED-based communication and networking has important advantages: Bandwidth, bandwidth density Privacy--security, bypassing Ubiquity if piggybacked on lighting Unregulated spectrum peer-to-peer prototype 5
Demonstration of Communications using COTS Components (PHO 445) Transmitter Receiver N. Barraford, P. Dib B. Gallagher K. Shah