Spa Sparsi rsity-pr propor porti tional Ba Backhaul l an and Compute e fo for SDRs Moein Khazraee, Ye Yeswanth Gu Guddeti , Sam Crow, Alex C. Snoeren, Kirill Levchenko, Dinesh Bharadia, Aaron Schulman
Software Defined Radios have a lot of potential SDRs s are unive versa sal and and flexi xible They can decode any protocol on any band: (e.g., WiFi, Bluetooth, Zigbee, FM, LTE) SDRs s can enable exc xciting new new ap applicat cations ons - Universal IoT Gateway - Radios in the cloud
What makes SDRs so powerful? De Decouple pled d radio frontend from si signal-processi ssing backe kend Universal Capture Flexible Compute DATA Radio frontend Signal processor Downsample ADC Filter Process Bluetooth/ZigBee 2.4 GHz ISM band Paging 900 MHz band AM/FM VHF Aviation Band
Unfortunately, SDRs consume a lot of resources 100 MHz z (USRP 2) Serve ver-class ss processo ssor 800 M 800 Mbps 25 M 25 MHz DATA Radio frontend Signal processor Downsample ADC Filter Process s inefficient , SDR backh SD khaul and and comput compute is especially for low-bitrate frequency hopping protocols (e.g., 1 Mbps BLE)
Opportunity: the spectrum is sparsely occupied 512 4 Mbps Gbps None Occupied Bandwidth (MHz)
SO YOU’RE BUT THEY’RE TELLING ME MOSTLY BACKHAUL WASTED ON AND COMPUTE NOISE ?! ARE THE LIMITING FACTORS,
SDRs should be sparsity-proportional Threshold
SDRs should be sparsity-proportional Only backhaul and compute active signals
SparSDR : Sparsity Proportional Backhaul and Compute Requires only residential-class backhaul DATA Radio frontend Signal processor Downsample ADC Filter Process Sparsity-prop. Sparsity-prop. 2. Backend 1. Frontend Runs on embedded processors Fits in existing FPGAs A Raspberry Pi can handle 100 MHz bandwidth! (captured with a SparSDR-enabled USRP N210)
How can we do sparsity-proportional downsampling? Signal captured by SDR frontend 1. Divide the signal into discrete intervals 2. Perform a Discrete FFT 2 FFT 1 Fourier Transform (FFT) 3. Throw away unused freq. bins Frequency Frequency 4. Backhaul what’s left
FFT alone is not enough FFT spreads signals, erasing gains from sparsity
Improving backhaul efficiency with windowing Signal captured by SDR frontend Windowing focuses energy on active bins
Improving backhaul efficiency with windowing Cost of windowing 2x overhead
Now, the That’s the frontend. backend.
Reconstructing signals on the backend Naive ve method: Full length Inverse Fourier Transform Our m Ou meth thod: : P artial IFFT Frequency Domain 𝑔 0 $ Filtering Time Domain 0 0 𝑢 𝑢 Full length IFFT Partial IFFT Full IFFT compute is not sparsity proportional Needs additional phase correction
“Let m me s show y you t that w we c can m make existing S g SDRs s sparsity p proportional”
Frontend Implementation in FPGA SparSDR Frontend Sparsity-prop. 100 MHz Packet 1Gbit Downsampling ADC Framer Ethernet 100 MHz Configurable FPGA SDR SDR BRAM BRAM DSP DSP AD Pluto [$200] 116 (97%) 24 (30%) USRP N210 [$2K] 85 (67%) 95 (75%) USRP X310 [$10K] 772 (49%) 1417 (92%) Opportunity: Spare resources on SDRs
What did we manage to fit into SDR? US USRP N2 N210 • Max FFT length of 2048, or resolution of 48.8 KHz AD AD Pluto • Max FFT length of 1024, or resolution of 60 KHz SDR SDR DSP DSP AD Pluto 24 (30%) USRP N210 95 (75%) USRP X310 1417 (92%)
Increasing FFT length improves efficiency % of max backhaul
Challenge: There is an unpredictable data rate Data rate SDR SDR BRAM BRAM AD Pluto 116 (97%) USRP N210 85 (67%) USRP X310 772 (49%) Solution: FIFOs
Challenge: sample rate = FPGA clock rate SparSDR Frontend Sparsity-prop. 100 MHz 100 MHz Packet 1Gbit Downsampling ADC ADC Framer Ethernet 100 MHz 100 MHz 100 MHz Configurable FPGA Clock Clock I need a drink! New input sample at each clock cycle Solution: Use pipelined implementations of FFT and multiply
Case study: Universal IoT Gateway • USRP N210 and Rpi 3+ • SparSDR reconstructs and decodes Bluetooth captures in real time on Rpi 3+
Benefit of SparSDR for IoT
Is SparSDR compute sparsity-proportional? SparSDR’s backhaul and compute scale linearly with the rate of received BLE advertising packets.
Case study: .
Future Work • Porting to more platforms • Make cloud SDRs a reality • Low cost sensors for wide deployment: AD Pluto + Rpi ~ $200 • Make IoT Gateway more comprehensive • Multi-protocol decoding • Transmit side
Go get SparSDR for your N210 and AD Pluto! It’s open source and integrated into GNURadio: https://github.com/ucsdsysnet/sparsdr
Recommend
More recommend
