Backscatter In WiFi Networks:The PHY Design

To meet the ever-exploding IoT access requirements,more and more new wireless technologies are being developed.Among them,backscatter is especially popular because of its simple structure,battery-free and wireless working mode,and ultra-low maintenance cost.It conveys data by passively reflecting and modifying radio frequency signals in the environment,such as WiFi,Bluetooth,LTE,and 5G.Reflection is a natural phenomenon that consumes no energy.The system only needs to switch the reflection states to modulate the data,and the corresponding power consumption is in the order of μW.Many backscatter systems utilizing ambient signals and commercial radios have been designed to avoid the need for specialized readers,so as to reduce deployment cost.But their compatibility with those wireless infrastructures and networks focuses on the modulation and demodulation in the physical layer.Those systems just use ambient signals,devices,and networks blindly,passively,and inefficiently:they modulate the excitation signals without verifying whether they are the target carriers and whether there are corresponding receivers;they wait for the excitation signals without trying to order ambient devices to generate them;they support only low-rate transmissions even when the excitation resource is abundant and the channel is clean.To tackle those drawbacks,this thesis considers how tag can interact with WiFi networks to obtain stable and controllable target carriers for high-speed transmission.Specifically,this thesis provides the physical layer techniques required to embed backscatter nodes into WiFi networks,which mainly includes three aspects:how to distinguish WiFi signals from the complex spectrum,how to decode and reply to WiFi,and how to stream data to WiFi at a high rate.Correspondingly,the works of this thesis are summarized as follows:(1)This thesis introduces the carrier signal identification method.There are various wireless signals in the environment.We focus on the most common WiFi(802.11b/n),Bluetooth,and ZigBee signals in the 2.4GHz ISM band.The backscatter node needs to identify the signal before reflection and modulation.Otherwise,it will cause serious failures.For this reason,Multiscatter that uses the envelope to identify carrier protocol is proposed.It mainly includes a baseband envelope extraction circuit,an envelope-based protocol recognition algorithm,and an overlay modulation module.For the accurate envelope waveform and the acceptable strength,the bandwidth of the baseband envelope extraction circuit is set to 40MHz,and then a low-power ADC is used to convert the envelope into digital signal.Protocol identification is achieved mainly through cross-correlation.After that,the backscatter node can modulate the carrier signal accordingly.In addition,the system also greatly simplifies the identification algorithm to reduce computation burden and power consumption.Experiment results show that Multiscatter can achieve an average identification accuracy of more than 95%.The idea of identifying and then modulating also performs well.(2)This thesis introduces how backscatter tag interacts with WiFi network.The ability to understand what WiFi is talking about and give response correspondingly is necessary to obtain controllable carriers.For this,Chameleon,which keeps compatibility with native WiFi,is introduced.The key insight is that bit ’1’ and bit ’0’ in 802.11b(1Mbps mode)have distinguishable envelope pulse widths at symbol boundaries.With this observation,we introduce how the tag can decode WiFi signal based on the envelope.Still,the starting part of a symbol is enough for decoding.The remaining can still work as backscatter carrier.The tag modulates each symbol immediately after decoding.Then the carrier is completely known to the tag so that it can reshape the carrier to create a WiFi packet containing any information.Combining those aspects,Chameleon has the ability to use the Beacon signal from ambient WiFi AP to create its own Beacon.Experimental results show that Chameleon realizes a bi-directional throughput of about 1Mbps with native WiFi.Still,Beacon packets generated by Chameleon can be received by computers,smartphones,and iPads,and displayed as a WLAN access point.That means the tag is able to interact with native WiFi devices at the physical layer.So that it is feasible for the tag to interact with COTS WiFi for stable carrier.(3)This thesis introduces a high-throughput backscatter modulation using WiFi signal.Existing backscatter systems either have low data rates or need receivers customized in software or hardware.For this,SubScatter which uses sub-symbol modulation to keep high throughput and compatibility with COTS radios at the same time is introduced.Its carrier is CCK mode 802.11 b packets.The node imitates CCK modulation of standard 802.11 b to carry independent information in their 4 concurrent streams.The backscattered packets equally go through two rounds of independent CCK modulation.It can be proven that they are still standard WiFi packets.Experimental results show that SubScatter realizes a throughput of 10.9Mbps using commercial WiFi devices.