Raytheon BBN Technologies A Vision for Advanced Wireless Testbeds Chip Elliott Chief Scientist, BBN This document does not contain technology or Technical Data controlled under either the U.S. International Traffic in Arms Regulations or the U.S. Export Administration Regulations.
5G November 22, 2016 A Vision for Wireless Testbeds 2
10 9 G 5G November 22, 2016 A Vision for Wireless Testbeds 3
X 10 9 G 5G November 22, 2016 A Vision for Wireless Testbeds 4
X 10 9 G 5G November 22, 2016 A Vision for Wireless Testbeds 5
Outline • Wireless, the enabler . . . • . . . when everything is the cloud . . . • . . . opens up tomorrow’s apps • How to get started November 22, 2016 A Vision for Wireless Testbeds 6
A handy invention November 22, 2016 A Vision for Wireless Testbeds 7
If we built an RF equivalent . . . ? RF illumination patterns Fading and reflections Even select by modulations Probably feasible to build / demo November 22, 2016 A Vision for Wireless Testbeds 8
What will we see in 20 years ? • A world drenched in RF • Much fuller spectrum occupancy than today • Many, many illuminators with a wide variety of radiation patterns (lots of beams) • Many very agile illuminators (beam steering, spectrum, coding, distributed beam forming, etc etc) • Many flickering, mobile illuminators November 22, 2016 A Vision for Wireless Testbeds 9
Kind of like this, everywhere November 22, 2016 A Vision for Wireless Testbeds 10
Outline • Wireless, the enabler . . . • . . . when everything is the cloud . . . • . . . opens up tomorrow’s apps • How to get started November 22, 2016 A Vision for Wireless Testbeds 11
When everything is the cloud, Wireless is the Cloud’s Skin Edge datacenters Regional datacenters Core datacenters Ubiquitous high-capacity wireless “skin” November 22, 2016 A Vision for Wireless Testbeds 12
A first step into this new world Cloud Radio Access Networks (CRANs) November 22, 2016 A Vision for Wireless Testbeds 13
A more radical architecture Air “RF Hypervisors” Shared Sliced processing chains Shared Sliced processing chains with SDR data with SDR data Multi-tenant cloud Multi-tenant cloud • A fundamentally sliced (multi-tenant) architecture • SDRs w/ multiple back ends (pub sub) • Multifunction RF, all software-defined • Cloud-style, not telco-style November 22, 2016 A Vision for Wireless Testbeds 14
An example Air Web access Web access Drone control Drone control Radar Radar IM / Text IM / Text “RF Hypervisors” Shared Sliced processing chains Shared Sliced processing chains with SDR data with SDR data Multi-tenant cloud Multi-tenant cloud • A fundamentally sliced (multi-tenant) architecture • SDRs w/ multiple back ends (pub sub) • Multifunction RF, all software-defined • Cloud-style, not telco-style November 22, 2016 A Vision for Wireless Testbeds 15
Outline • Wireless, the enabler . . . • . . . when everything is the cloud . . . • . . . opens up tomorrow’s apps • How to get started November 22, 2016 A Vision for Wireless Testbeds 16
A very simple case Why not use the knowledge you have ? I know where I’m going Cloud scheduling wireless assets / handoff I know where I’m going November 22, 2016 A Vision for Wireless Testbeds 17
City-wide traffic orchestration? Why should cars be autonomous ? Even more so, why should drones be ? Surely they should leverage all existing knowledge about everything in the city … .. November 22, 2016 A Vision for Wireless Testbeds 18
Can every network access point help robots navigate ? From E911 -> wifi -> wallplug appliance ? November 22, 2016 A Vision for Wireless Testbeds 19
Can every city get, as a service, a realtime, distributed radar system ? Run radar slice in parallel with comms Leverage dense RF transmitters Maybe add tailored waveforms for weather, drones, etc etc (SDR … ) Multistatic transmitters & receivers November 22, 2016 A Vision for Wireless Testbeds 20
Urban transport / automobiles ? • DARPA GXV-T • Extreme mobility • (Why have 4 wheels?) • No windows • Augmented cloud-based reality instead • Looking forward . . . – Perhaps divorce “coach” from chassis, a personal “living room” with a leased-per-drive chassis ? – “Coach” as in trimodal containers, joining “buses” within city, parking bins, slide into hyperloop, etc – Chassis can service people, freight haulage, etc November 22, 2016 A Vision for Wireless Testbeds 21
Extremely fine-grain realtime sensing ? • Lidar on every car ? • Projected cost: $250 / unit • Current Velodyne stats – 120m range, < 2cm accuracy – 64 channels (emitter/receiver pairs) – 2.2 Million points per second – About 100 Mbps data stream • What if we harvested / pooled all those “data torrents”? – Many looks at everything near a road (many cars, many angles) – Continuous, realtime infrastructure monitoring – Measuring weather and climate change (watch leaves) – Counting cars at walmart – Realtime RF channel models (e.g. leaves again) – Synthetic data for those windows November 22, 2016 A Vision for Wireless Testbeds 22
Outline • Wireless, the enabler . . . • . . . when everything is the cloud . . . • . . . opens up tomorrow’s apps • How to get started November 22, 2016 A Vision for Wireless Testbeds 23
The best way to predict the future is to invent it November 22, 2016 A Vision for Wireless Testbeds 24
A “Deeply Programmable ” Citywide Testbed Some slices for Researchers, others for Service Providers The Internet SDX New apps / services Internet of Things IP Slices VM VM VM Sliced local cloud Citywide testbed(s) Next generation citywide wireless Low-latency, deeply coverage with many radio access points programmable, “sliced” local cloud November 22, 2016 A Vision for Wireless Testbeds 25
When everything is the cloud, Wireless is the Cloud’s Skin Edge datacenters Regional datacenters Core datacenters Ubiquitous high-capacity wireless “skin” November 22, 2016 A Vision for Wireless Testbeds 26
Let’s be ambitious ! November 22, 2016 A Vision for Wireless Testbeds 27
Recommend
More recommend
